Video yükleniyor...

Video Yüklenemedi

Bu video yüklenirken bir sorun oluştu. Bu geçici bir ağ sorunundan kaynaklanıyor olabilir veya video kullanılamıyor olabilir.

Ana Sayfaya Dön

𝗥𝗼𝗯𝗼𝘁𝗶𝗰 𝗧𝗔𝗥𝗨𝗣 𝘃𝘀. 𝗟𝗮𝗽𝗮𝗿𝗼𝘀𝗰𝗼𝗽𝗶𝗰 𝗜𝗣𝗢𝗠 𝗥𝗲𝗽𝗮𝗶𝗿 𝗼𝗳 𝗩𝗲𝗻𝘁𝗿𝗮𝗹 𝗛𝗲𝗿𝗻𝗶𝗮 Defect closure: 🔵Standard of care in both approaches 🔵Robot facilitates closure of larger defect up to a maximum size of 5cm for TARUP repair Comparison: 🔴Retromuscular (rTARUP) vs. Intraperitoneal (IPOM) mesh placement 🔴No tacks (rTARUP) vs. Use of tacks (IPOM)... 🔴Minimal risk of bowel adhesions (rTARUP) vs. Possibility of bowel adhesions (IPOM) 🔴Extensive dissection (rTARUP) vs. Minimal dissection (IPOM) 🔴LOS and analgesic requirements less for rTARUP Less pain in rTARUP due to: 🟢Better and precise haemostasis in dissection plane facilitated by robotic platform 🟢No use of tacks 🟢Robotic port remote centre technology with less muscle trauma at port sites #FOAMed #SoMe4Surgery #hernia #surgery #roboticsshow more

Derby Pancreaticobiliary & Robotic AWR Unit

10,603 subscribers

11,606 görüntüleme • 2 yıl önce •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

1 Yorum

Hatem Elbernawi profil fotoğrafı

Hatem Elbernawi2 yıl önce

@Innov_Medicine @SAGES_Updates @MISIRG1 @BehindTheKnife @MedicalAdv @imedverse @rbarbosa91 @altaf_awan12 @ib9994 @javlatif Both are done beautifully and excellent skill

Benzer Videolar

𝗘𝗺𝗲𝗿𝗴𝗲𝗻𝗰𝘆 𝗥𝗼𝗯𝗼𝘁𝗶𝗰 𝗥𝗲𝗽𝗮𝗶𝗿 𝗼𝗳 𝗟𝗲𝗳𝘁 𝗜𝗻𝗴𝘂𝗶𝗻𝗮𝗹 𝗛𝗲𝗿𝗻𝗶𝗮 𝘄𝗶𝘁𝗵 𝗜𝗖𝗚 Surgical Approach: 🔴Initial assessment of small bowel with ICG 🔴Small bowel placed in pelvis with warm wet swabs overlying 🔴TAPP approach to hernia repair ➡️Critical view of MPO ➡️ICG to help delineate key structures ➡️Progrip mesh ➡️3-0 absorbable V-lock to close peritoneum 🔴Reassessment of small bowel after hernia repair ➡️ICG present in bruised area demonstrating viability ➡️Aided decision making to not resect or oversew Robotics in EGS: 🟢Case selection key 🟢MDT approach with anaesthetic and scrub teams to coordinate 🟢Minimally invasive approaches in EGS should be encouraged #FOAMed #GITwitter #MedEd #Surgery #Hernia #EGS

𝗘𝗺𝗲𝗿𝗴𝗲𝗻𝗰𝘆 𝗥𝗼𝗯𝗼𝘁𝗶𝗰 𝗥𝗲𝗽𝗮𝗶𝗿 𝗼𝗳 𝗟𝗲𝗳𝘁 𝗜𝗻𝗴𝘂𝗶𝗻𝗮𝗹 𝗛𝗲𝗿𝗻𝗶𝗮 𝘄𝗶𝘁𝗵 𝗜𝗖𝗚 Surgical Approach: 🔴Initial assessment of small bowel with ICG 🔴Small bowel placed in pelvis with warm wet swabs overlying 🔴TAPP approach to hernia repair ➡️Critical view of MPO ➡️ICG to help delineate key structures ➡️Progrip mesh ➡️3-0 absorbable V-lock to close peritoneum 🔴Reassessment of small bowel after hernia repair ➡️ICG present in bruised area demonstrating viability ➡️Aided decision making to not resect or oversew Robotics in EGS: 🟢Case selection key 🟢MDT approach with anaesthetic and scrub teams to coordinate 🟢Minimally invasive approaches in EGS should be encouraged #FOAMed #GITwitter #MedEd #Surgery #Hernia #EGS

Derby Pancreaticobiliary & Robotic AWR Unit

23,164 görüntüleme • 23 gün önce

𝗟𝗮𝗽𝗮𝗿𝗼𝘀𝗰𝗼𝗽𝗶𝗰 𝗙𝗲𝗻𝗲𝘀𝘁𝗿𝗮𝘁𝗲𝗱 𝗦𝘂𝗯𝘁𝗼𝘁𝗮𝗹 𝗖𝗵𝗼𝗹𝗲𝗰𝘆𝘀𝘁𝗲𝗰𝘁𝗼𝗺𝘆 Patient Presentation: 🔵History of severe cholecystitis 🔵CT on index admission showed: ➡️Cholecysto-colic fistula ➡️Air locules on GB ➡️Figure 1 below 🔵Follow up CT (~4 months later): ➡️Good improvement with apparent resolution of fistula as absence of air in GB ➡️Figure 2 below 🔵Symptomatic gallstones 🔵Listed for planned biliary surgery Operative Approach: 🔴Colonic adhesion to GB fundus 🔴Fused hepatocystic triangle ➡️Decision to perform subtotal cholecystectomy 🔴Dissection plane (window) identified between GB and colon 🔴Presumed area of fistula tract detached with cuff of GB wall 🔴Stone retrieved and GB opened to identify cystic duct ostium on opening GB ➡️Closed with 2/0 Vicryl 🔴Gallbladder dissected off as far as safe to do so 🔴Detached colonic attachment with GB cuff ➡️No obvious fistula tract ➡️The area was under run to achieve haemostasis 🔴Cystic artery transfixed with figure of 8 suture 🔴Drain x 1 ➡️Removed after 24 hours 🔴Uneventful postop course #FOAMed #MedTwitter #GITwitter #SoMe4Surgery #HPB

𝗟𝗮𝗽𝗮𝗿𝗼𝘀𝗰𝗼𝗽𝗶𝗰 𝗙𝗲𝗻𝗲𝘀𝘁𝗿𝗮𝘁𝗲𝗱 𝗦𝘂𝗯𝘁𝗼𝘁𝗮𝗹 𝗖𝗵𝗼𝗹𝗲𝗰𝘆𝘀𝘁𝗲𝗰𝘁𝗼𝗺𝘆 Patient Presentation: 🔵History of severe cholecystitis 🔵CT on index admission showed: ➡️Cholecysto-colic fistula ➡️Air locules on GB ➡️Figure 1 below 🔵Follow up CT (~4 months later): ➡️Good improvement with apparent resolution of fistula as absence of air in GB ➡️Figure 2 below 🔵Symptomatic gallstones 🔵Listed for planned biliary surgery Operative Approach: 🔴Colonic adhesion to GB fundus 🔴Fused hepatocystic triangle ➡️Decision to perform subtotal cholecystectomy 🔴Dissection plane (window) identified between GB and colon 🔴Presumed area of fistula tract detached with cuff of GB wall 🔴Stone retrieved and GB opened to identify cystic duct ostium on opening GB ➡️Closed with 2/0 Vicryl 🔴Gallbladder dissected off as far as safe to do so 🔴Detached colonic attachment with GB cuff ➡️No obvious fistula tract ➡️The area was under run to achieve haemostasis 🔴Cystic artery transfixed with figure of 8 suture 🔴Drain x 1 ➡️Removed after 24 hours 🔴Uneventful postop course #FOAMed #MedTwitter #GITwitter #SoMe4Surgery #HPB

Derby Pancreaticobiliary & Robotic AWR Unit

17,864 görüntüleme • 2 yıl önce

𝙇𝙖𝙥𝙖𝙧𝙤𝙨𝙘𝙤𝙥𝙞𝙘 𝙍𝙚𝙥𝙖𝙞𝙧 𝙤𝙛 𝙋𝙚𝙧𝙛𝙤𝙧𝙖𝙩𝙚𝙙 𝘿𝙪𝙤𝙙𝙚𝙣𝙖𝙡 𝙐𝙡𝙘𝙚𝙧 𝙬𝙞𝙩𝙝 𝙋𝙧𝙞𝙢𝙖𝙧𝙮 𝘾𝙡𝙤𝙨𝙪𝙧𝙚 & 𝙊𝙢𝙚𝙣𝙩𝙖𝙡 𝙋𝙖𝙩𝙘𝙝 Presentation: 🔵1 day history of sever upper abdominal pain 🔵Smoker and excess ETOH 🔵CXR ➡️Pneumoperitoneum 🔵CT ➡️Perforated duodenal ulcer 🔵Time from ED➡️theatre 3 hours Operative Approach: 🔴Thorough peritoneal lavage ➡️Lap washout allows for thorough removal of intra-abdominal contamination ➡️Modern OR table tilting can provide better view then open in select cases 🔴If fragile tissue then omental patch only ➡️This approach is a modified version of Grahams patch repair ➡️When suitable, we prefer primary closure with patch after 🔴For large perforations kocherisation of duodenum may be indicated 🔴Convert to open if: ➡️Large ulcer ➡️Associated ulcer bleed 🔴H. Pylori eradication therapy on discharge 🔴OGD in 6-8 weeks to confirm healing Key Points: 🟢Timely intervention critical 🟢Lap approach in EGS should be encouraged 🟢Compared to open, advantages of lap are: ➡️Reduced pain ➡️Reduced LOS ➡️Reduced incisional hernia rate 🟢Trainees should be exposed to lap EGS early to develop skills and confidence #FOAMed #GITwitter #MedEd #SurgEd #EGS #SoMe4Surgery

𝙇𝙖𝙥𝙖𝙧𝙤𝙨𝙘𝙤𝙥𝙞𝙘 𝙍𝙚𝙥𝙖𝙞𝙧 𝙤𝙛 𝙋𝙚𝙧𝙛𝙤𝙧𝙖𝙩𝙚𝙙 𝘿𝙪𝙤𝙙𝙚𝙣𝙖𝙡 𝙐𝙡𝙘𝙚𝙧 𝙬𝙞𝙩𝙝 𝙋𝙧𝙞𝙢𝙖𝙧𝙮 𝘾𝙡𝙤𝙨𝙪𝙧𝙚 & 𝙊𝙢𝙚𝙣𝙩𝙖𝙡 𝙋𝙖𝙩𝙘𝙝 Presentation: 🔵1 day history of sever upper abdominal pain 🔵Smoker and excess ETOH 🔵CXR ➡️Pneumoperitoneum 🔵CT ➡️Perforated duodenal ulcer 🔵Time from ED➡️theatre 3 hours Operative Approach: 🔴Thorough peritoneal lavage ➡️Lap washout allows for thorough removal of intra-abdominal contamination ➡️Modern OR table tilting can provide better view then open in select cases 🔴If fragile tissue then omental patch only ➡️This approach is a modified version of Grahams patch repair ➡️When suitable, we prefer primary closure with patch after 🔴For large perforations kocherisation of duodenum may be indicated 🔴Convert to open if: ➡️Large ulcer ➡️Associated ulcer bleed 🔴H. Pylori eradication therapy on discharge 🔴OGD in 6-8 weeks to confirm healing Key Points: 🟢Timely intervention critical 🟢Lap approach in EGS should be encouraged 🟢Compared to open, advantages of lap are: ➡️Reduced pain ➡️Reduced LOS ➡️Reduced incisional hernia rate 🟢Trainees should be exposed to lap EGS early to develop skills and confidence #FOAMed #GITwitter #MedEd #SurgEd #EGS #SoMe4Surgery

Derby Pancreaticobiliary & Robotic AWR Unit

390,695 görüntüleme • 12 gün önce

Mass Effect fans sometimes call Drew Karpyshyn's departure one of Bioware's biggest mistakes. But a far less-known exit, that of Chris L'Etoile, had an arguaby worse impact. His vision for the Reapers was dropped last-minute, and it could've made Destroy a very unpopular ending. 🔴 In Reaper lore, there's often mention of "Harvesting". On rare occasions, Reapers, the Catalyst, and EDI use vague descriptors like "processing", "essence", and "preserving". But beyond that, the nature and details of Harvesting are never fleshed out, despite its key role in Reaper cycles. 🔵 After all, what exactly Reapers are and how they function has always been a debated topic in the fandom, due to the conflicting writing in Mass Effect 1, 2, and 3. 🟡 Chris L'Etoile, who was in charge of writing AI, had a more specific idea. See attached media for proof. Reapers are strictly machines, but they are merely a platform - a "bank" built with the purpose of hosting billions of people whose consciousness was digitalized and uploaded onto a neural network, complete with their personalities, experiences, and memories. Think Mikoshi from Cyberpunk, or the "Virtual Aliens" from Mass Effect's own lore. This was explained by EDI upon finding the Human-Reaper. Voice acting and WIP cutscenes were also done. After Chris left, his work was discarded except for Legion and EDI. 🟢 Going into Mass Effect 3, anyone can see how Chris' original vision could've added tons of depth and complexity to our antagonists. Each time a Reaper is destroyed, billions of uploaded minds are also killed. Whether these people are alive or not lets Mass Effect explore similar transhuman themes as Cyberpunk does with its engrams. The concept can be reinforced by having the Catalyst emphasize what he only hints at in the final game: That Reapers preserve all science/culture/technology. With digitalized people, they're total civilizational repositories. This all throws a wrench into the morality of Destroy. There've been tens of thousands of Reaper cycles. The Catalyst's form/voice can even change based on which uploaded civilization "holds the mic" at the moment. As a cherry on top, Chris' Reapers would've played nicely into ME3's existing exploration of whether AI is alive or not (e.g. EDI's character arc, and Geth during Rannoch). #MassEffect #bioware

Mass Effect fans sometimes call Drew Karpyshyn's departure one of Bioware's biggest mistakes. But a far less-known exit, that of Chris L'Etoile, had an arguaby worse impact. His vision for the Reapers was dropped last-minute, and it could've made Destroy a very unpopular ending. 🔴 In Reaper lore, there's often mention of "Harvesting". On rare occasions, Reapers, the Catalyst, and EDI use vague descriptors like "processing", "essence", and "preserving". But beyond that, the nature and details of Harvesting are never fleshed out, despite its key role in Reaper cycles. 🔵 After all, what exactly Reapers are and how they function has always been a debated topic in the fandom, due to the conflicting writing in Mass Effect 1, 2, and 3. 🟡 Chris L'Etoile, who was in charge of writing AI, had a more specific idea. See attached media for proof. Reapers are strictly machines, but they are merely a platform - a "bank" built with the purpose of hosting billions of people whose consciousness was digitalized and uploaded onto a neural network, complete with their personalities, experiences, and memories. Think Mikoshi from Cyberpunk, or the "Virtual Aliens" from Mass Effect's own lore. This was explained by EDI upon finding the Human-Reaper. Voice acting and WIP cutscenes were also done. After Chris left, his work was discarded except for Legion and EDI. 🟢 Going into Mass Effect 3, anyone can see how Chris' original vision could've added tons of depth and complexity to our antagonists. Each time a Reaper is destroyed, billions of uploaded minds are also killed. Whether these people are alive or not lets Mass Effect explore similar transhuman themes as Cyberpunk does with its engrams. The concept can be reinforced by having the Catalyst emphasize what he only hints at in the final game: That Reapers preserve all science/culture/technology. With digitalized people, they're total civilizational repositories. This all throws a wrench into the morality of Destroy. There've been tens of thousands of Reaper cycles. The Catalyst's form/voice can even change based on which uploaded civilization "holds the mic" at the moment. As a cherry on top, Chris' Reapers would've played nicely into ME3's existing exploration of whether AI is alive or not (e.g. EDI's character arc, and Geth during Rannoch). #MassEffect #bioware

Orikon

31,634 görüntüleme • 3 ay önce

How to Identify Agents with Sound Alone: Audio Profiles Each agent in VALORANT has a unique walk sound. We can use this to identify who is moving around without seeing them AND without them using abilities. Each agent has unique tells, here's Proof: Key: (How Easy it is to Identify who is running) 🟢 - Extremely Easy to identify without trying 🟡 - Somewhat difficult until you know what to listen for 🔴 - Difficult, and needs practice In order of their names Alphabetically (up to Miks in release date) Astra - 🟢. Astra has golden, loose jewelry. This makes it easy to identify her because she is one of the only agents that you can hear it, and of those agents, it's the most consistent. She also has lighter steps as she is meant to be smaller than the other agents 5'6 (170cm) Breach - 🟢. Breach is easily identifiable by the heavy boots that he wears. He is one of the loudest agents in terms of walking as his boots as well as his radianite arms that are also quite loud, being made of metal. Brimstone - 🟢. Similar to Breach, Brim is also easily identifiable due to how heavy his footsteps are. What separates him from Breach is his boots are leather but stomps, likely coming from his military background. Chamber - 🟡. Chamber is what started this entire thing for me. You can identify chamber by his loose gold watch, which chimes as he walks. The reason this is yellow circle and not green is because it is not as frequent as Astra and he wears sneakers, which the majority of the cast wear. So besides the watch it can be somewhat difficult to identify, if you're not actively listening for it. Clove - 🟡. Clove is somewhat difficult, but what makes them unique is how soft / quiet they walk. They are very light, wears shoes that have a lot of cushion and is meant to feel like a smaller character, canonically 5'6 / 168cm. Cypher 🔴 - Despite playing a lot of Cypher, I have a really hard time differentiating him from other characters with similar patterns like Viper and Fade. Comapred to Fade, I would say his have more bass. Deadlock 🟢 - The only person you could mix Deadlock up with is Sova. However, Deadlock's walk has a much more "wet" sound to it, as if she were crushing in soft snow with her snow boots. Fade 🔴 - Too similar to other agents, especially Cypher makes it very hard to tell. The only thing that makes her unique is her consistency, the decibel spread is almost nothing. Gekko 🟡- Also pretty hard, as he is one of the many average weight sneaker users. He is very similar to Killjoy, but with more bass which is how i remember. He's wearing vans, it sounds like rubber. Harbor 🟢 - Extremely easy. You can hear his braclet from a mile away. One of the more unique ones. Iso 🔴- Another middleweight flat shoe wearer. I think he is meant to be heavier than agents like Gekko and Miks who sound similar. This is one wear you have to look at who is alive to differentiate between him and other agents. Jett 🟢- Extremely easy, maybe because I've heard it so much with the Jetts in my games, but I think the idea is she is meant to feel light and dainty, as if the wind is carrying her. That's why she walks so soft. She is also meant to be quite small at 5'4 / 164cm which corroborates the quiet steps. KAY/O 🟢- The easiest. Even people who don't play VALORANT could tell you that's the sound of a robot. The metal makes it very easy to tell. Killjoy 🟡 - Somewhat difficult, I think the giveaway are the boots. Like I said about Gekko, hers in comparison create softer sounds. Miks 🔴 - Really disappointed that Miks, the sound guy doesn't have anything unique to his walk. It is similar to Gekkos, in that it is rubbery but also muddy. Neon 🟢- Iconic. Being a runner, she is wearing running shoes. Not much said, I think anyone could hear the difference. Omen 🟡 - Somewhat difficult. You are listening for a "wet" sound when he walks. He has been out for long enough where I can hear it distinctly enough. Phoenix 🟡 - Somewhat difficult. You are listening for the same rubbery sound as Miks, but with a lot more bass, he drags his feet somewhat. Raze 🟡 - Hard to hear at first, but you are listening for a "mud" / squish sound. Similar to Omen, but not "stompy" Reyna 🟢- Maybe just me but I have an easy time hearing Reyna's. This is the one I can identify the quickest in game. I listen for a lot of bass towards the front of her step, as she has raised shoes which put more impact on her toes / balls of her feet. The sound is unique enough where you can her pretty clearly. Sage 🟡 - Somewhat difficult. Can be confused with Raze sometimes, as you are also listening for a "muddy sound" Skye 🟢- Very easy, she has trail shoes. Makes sense, she's from Australia. Sova 🟢- Again, similar to deadlock, Sova has boots made for winter, and there are feint sounds of him walking on ice. This one is somewhat similar to Skye, but not enough to mistakes the two. His has more bass. Tejo 🟡 - I didn't play as much when Tejo was popular, and this could be why I don't hear much differences in Gekko and Miks as well, but I find Tejo easier because of his heavty duty boots. Veto 🟡 - Somewhat hard to hear, only because I haven't played enough but there is dampness and a muddy sound to his walk similar to Raze, Omen etc. with that rubbery sound Viper 🟡 - somewhat hard to hear. Again, I think she sounds very similar to Fade and Cypher. But I think she is easier to identify because of the bass in each step. Vyse 🟢 - Very easy once you start listening. She has a lot of metal attached to her, which you can hear. Waylay 🔴 - Another middleweight rubber user. Hard to hear, but I haven't played with/against enough Waylay to really know her pattern. Yoru 🟢 - Iconic sound, you have been conditioned to hear this sound over and over with the Fake Footsteps and the reworked Yoru Clone + his fake clone of himself, he is most commonly walking around rather than sneaking, so you will probably become very accustom to him. Things to note: 1. These noises are NOT client / team sided. At the end of this video I show my friend walking as Chamber and Phoenix, you can hear a clear difference. He is on the ENEMY team, not mine. 2. You are NOT relying on audio alone. Agents all have their unique abilities which are significantly more telling of who is nearby compared to footsteps. The point of this exercise is to help identify WHO is there if there is no other way of knowing. 3. It is not meant to be used every round, hell even every game. I have really good hearing, so I can make good use of it. But as you can see by the red circles, it is not universal. Again, there are other more obvious ways of telling who is nearby. 4. But how is this useful? Well for me, I like to memorize the Controllers the most, because if I can identify that the sound of a controller's footsteps. I know they are not lurking, which allows for a little more peace of mind for my teammates rotating onto my site. 5. I used the same spot on purpose to show that they have different sound profiles. HOWEVER, it is important to note that with the different surfaces that you can run across, some of the noises that I described could sound different or harder to hear.

How to Identify Agents with Sound Alone: Audio Profiles Each agent in VALORANT has a unique walk sound. We can use this to identify who is moving around without seeing them AND without them using abilities. Each agent has unique tells, here's Proof: Key: (How Easy it is to Identify who is running) 🟢 - Extremely Easy to identify without trying 🟡 - Somewhat difficult until you know what to listen for 🔴 - Difficult, and needs practice In order of their names Alphabetically (up to Miks in release date) Astra - 🟢. Astra has golden, loose jewelry. This makes it easy to identify her because she is one of the only agents that you can hear it, and of those agents, it's the most consistent. She also has lighter steps as she is meant to be smaller than the other agents 5'6 (170cm) Breach - 🟢. Breach is easily identifiable by the heavy boots that he wears. He is one of the loudest agents in terms of walking as his boots as well as his radianite arms that are also quite loud, being made of metal. Brimstone - 🟢. Similar to Breach, Brim is also easily identifiable due to how heavy his footsteps are. What separates him from Breach is his boots are leather but stomps, likely coming from his military background. Chamber - 🟡. Chamber is what started this entire thing for me. You can identify chamber by his loose gold watch, which chimes as he walks. The reason this is yellow circle and not green is because it is not as frequent as Astra and he wears sneakers, which the majority of the cast wear. So besides the watch it can be somewhat difficult to identify, if you're not actively listening for it. Clove - 🟡. Clove is somewhat difficult, but what makes them unique is how soft / quiet they walk. They are very light, wears shoes that have a lot of cushion and is meant to feel like a smaller character, canonically 5'6 / 168cm. Cypher 🔴 - Despite playing a lot of Cypher, I have a really hard time differentiating him from other characters with similar patterns like Viper and Fade. Comapred to Fade, I would say his have more bass. Deadlock 🟢 - The only person you could mix Deadlock up with is Sova. However, Deadlock's walk has a much more "wet" sound to it, as if she were crushing in soft snow with her snow boots. Fade 🔴 - Too similar to other agents, especially Cypher makes it very hard to tell. The only thing that makes her unique is her consistency, the decibel spread is almost nothing. Gekko 🟡- Also pretty hard, as he is one of the many average weight sneaker users. He is very similar to Killjoy, but with more bass which is how i remember. He's wearing vans, it sounds like rubber. Harbor 🟢 - Extremely easy. You can hear his braclet from a mile away. One of the more unique ones. Iso 🔴- Another middleweight flat shoe wearer. I think he is meant to be heavier than agents like Gekko and Miks who sound similar. This is one wear you have to look at who is alive to differentiate between him and other agents. Jett 🟢- Extremely easy, maybe because I've heard it so much with the Jetts in my games, but I think the idea is she is meant to feel light and dainty, as if the wind is carrying her. That's why she walks so soft. She is also meant to be quite small at 5'4 / 164cm which corroborates the quiet steps. KAY/O 🟢- The easiest. Even people who don't play VALORANT could tell you that's the sound of a robot. The metal makes it very easy to tell. Killjoy 🟡 - Somewhat difficult, I think the giveaway are the boots. Like I said about Gekko, hers in comparison create softer sounds. Miks 🔴 - Really disappointed that Miks, the sound guy doesn't have anything unique to his walk. It is similar to Gekkos, in that it is rubbery but also muddy. Neon 🟢- Iconic. Being a runner, she is wearing running shoes. Not much said, I think anyone could hear the difference. Omen 🟡 - Somewhat difficult. You are listening for a "wet" sound when he walks. He has been out for long enough where I can hear it distinctly enough. Phoenix 🟡 - Somewhat difficult. You are listening for the same rubbery sound as Miks, but with a lot more bass, he drags his feet somewhat. Raze 🟡 - Hard to hear at first, but you are listening for a "mud" / squish sound. Similar to Omen, but not "stompy" Reyna 🟢- Maybe just me but I have an easy time hearing Reyna's. This is the one I can identify the quickest in game. I listen for a lot of bass towards the front of her step, as she has raised shoes which put more impact on her toes / balls of her feet. The sound is unique enough where you can her pretty clearly. Sage 🟡 - Somewhat difficult. Can be confused with Raze sometimes, as you are also listening for a "muddy sound" Skye 🟢- Very easy, she has trail shoes. Makes sense, she's from Australia. Sova 🟢- Again, similar to deadlock, Sova has boots made for winter, and there are feint sounds of him walking on ice. This one is somewhat similar to Skye, but not enough to mistakes the two. His has more bass. Tejo 🟡 - I didn't play as much when Tejo was popular, and this could be why I don't hear much differences in Gekko and Miks as well, but I find Tejo easier because of his heavty duty boots. Veto 🟡 - Somewhat hard to hear, only because I haven't played enough but there is dampness and a muddy sound to his walk similar to Raze, Omen etc. with that rubbery sound Viper 🟡 - somewhat hard to hear. Again, I think she sounds very similar to Fade and Cypher. But I think she is easier to identify because of the bass in each step. Vyse 🟢 - Very easy once you start listening. She has a lot of metal attached to her, which you can hear. Waylay 🔴 - Another middleweight rubber user. Hard to hear, but I haven't played with/against enough Waylay to really know her pattern. Yoru 🟢 - Iconic sound, you have been conditioned to hear this sound over and over with the Fake Footsteps and the reworked Yoru Clone + his fake clone of himself, he is most commonly walking around rather than sneaking, so you will probably become very accustom to him. Things to note: 1. These noises are NOT client / team sided. At the end of this video I show my friend walking as Chamber and Phoenix, you can hear a clear difference. He is on the ENEMY team, not mine. 2. You are NOT relying on audio alone. Agents all have their unique abilities which are significantly more telling of who is nearby compared to footsteps. The point of this exercise is to help identify WHO is there if there is no other way of knowing. 3. It is not meant to be used every round, hell even every game. I have really good hearing, so I can make good use of it. But as you can see by the red circles, it is not universal. Again, there are other more obvious ways of telling who is nearby. 4. But how is this useful? Well for me, I like to memorize the Controllers the most, because if I can identify that the sound of a controller's footsteps. I know they are not lurking, which allows for a little more peace of mind for my teammates rotating onto my site. 5. I used the same spot on purpose to show that they have different sound profiles. HOWEVER, it is important to note that with the different surfaces that you can run across, some of the noises that I described could sound different or harder to hear.

ROCKET

424,367 görüntüleme • 27 gün önce

I find this explanation of the Chinese system by Prof Keyu Jin (in a recent lecture at Harvard’s Fairbank center) absolutely fascinating. Keyu Jin is a professor of economics at LSE (London School of Economics) and serves on the board of companies like Credit Suisse. She’s also the daughter of Jin Liqun, former Vice Minister of finance of China so she’s a rare West-based academic (maybe even the only one) who actually has insight into the Chinese system from the inside. Essentially what she’s explaining is that a key reason why China was so successful economically is because of its decentralized nature, which creates two mutually compounding loops of competition, as opposed to one loop in the West. What does that mean? Well, contrary to popular belief that imagines China as being this centrally planned economy where almost everything is decided in Beijing, the inverse is actually true: China is actually one of the most decentralized countries in the world. To illustrate this, a metric that’s always amazed me is the fact that in China local governments (provinces, cities, villages, etc.) control a crazy 85% of the country's expenditures. On average that same metric for OECD countries is 33% (as in 64% of the expenditures are controlled at the federal/national level to China’s 15%). In the US for instance, which is already more decentralized than most given it’s a federation with states, only 45% of the country’s expenditures happen at the state and local level: almost twice less than in China! The effect of this, as Keyu Jin explains, is that provinces and larger municipalities in China have an immense degree of autonomy over the way they run their respective economies and fiercely compete with each other. This is the first loop. And then of course the second loop is that you have companies competing with each other in the market. As a result what constantly evolves in China is not only companies themselves but the environment in which they evolve: you constantly have this or that province running a new policy that proves very effective, making them gain an advantage vs other localities, initiative which is then copied by other localities. This makes the economic environment incredibly dynamic as it allows the state to move in unison with the economy, as opposed to slowing it down as is often the case in other countries. So what’s the role of the central government in all this? The key role, Keyu Jin argues, is setting broad objectives as well as personal management and promotion. And this is what makes the whole system work as therein lies the incentive for localities to compete with each other: because local officials know that if they do a better job than their peers, they’re on track for promotion by the central government. In “China Inc”, the central government is the board of directors and HR, presiding over an army of local CEOs with immense degrees of autonomy over their own “companies”. Keyu Jin gives the example of the solar industry. There was at some point (around 2005) a directive by the central government to develop the solar industry. The graph she shares in her talk is incredible: within a few years you had solar companies as well as patents related to research on solar technology pop up literally everywhere in China. With the result we all know about today: China today completely dominates the solar industry and solar technology (according to the International Energy Agency China's share in all the manufacturing stages of solar panels exceeds 80%). As she explains, this makes the Chinese system somewhat paradoxical as it is at the same time incredibly decentralized but also incredibly effective at mobilizing the country for centrally-decided objectives, in fact she goes as far as comparing this effectiveness to the country being in a constant state of “wartime mobilization”. An interesting comparison would be if you had all the countries in North America, the EU and North Africa (altogether roughly the population of China) all united under a common leadership deciding on common objectives and on the career path of all these countries’ officials, based on how well they achieve these objectives in their respective countries. We’re seeing this system being mobilized in its full strength today on leading edge semiconductors after US sanctions, and this is why these sanctions will undoubtedly ultimately prove so self-defeating: once the Chinese “wartime mobilization” machine is given an objective - and you can be sure this objective is prioritized very highly - the fight is essentially over, you can consider it done. Once you have hundreds of thousands of PhDs, companies and officials all at the same time competing and working within the same broad “China Inc” roof to make something happen, it will ultimately get done. If you want China NOT to develop a technology, the very last thing you want is to make them mobilize the full strength of the machine on it. With the sanctions the U.S. effectively told China: “please we beg you, do dedicate your formidable economic mobilization power to becoming a semiconductors powerhouse as fast as possible” 🤦 Another particularity of the system that Keyu Jin highlights - and I’ll end on this - is that this system also allows China to “allocate losses to certain groups of people, interest groups and sectors” in order to “enact system-level changes'', something she says is “very difficult for other governments with more political constraints to do”. For instance we’re seeing this play out in real-time with the real-estate industry: China recognized there was a housing bubble and Xi issued its “houses are for living in, not for speculation” directive. We’re since witnessing an engineered deflating of the bubble, ensuring to the extent possible that the losses are borne out by real estate developers and speculators, and not too much by society as a whole. This is part of the reason why China has never suffered a recession in the modern era: it does controlled demolition when necessary but tries to ensure it doesn’t suffer massive crises like we’ve repeatedly witnessed in the U.S. for instance. Of course no system is perfect. Weaknesses of the Chinese system include for instance local protectionism: there’s a perverse incentive for local officials to protect their local companies in order to give them a leg up vs companies from other provinces, which ultimately comes at the detriment of everyone. Another weakness is corruption, a sempiternal problem in China, where local officials - who are extremely powerful due to the nature of the system - will decide that getting promoted isn’t incentive enough and will try to cash in on their position of power. Cracking down on this is also a key remit of the central government and of course one of the major initiatives of Xi since he came to power. Lastly, another clear weakness is obviously that everything ultimately relies on the wisdom of what the system gets mobilized for, on the wisdom of these broader objectives coming from the central government. If they’re ill-thought, you effectively have a whole country working towards the wrong objectives… On this we’re often told that this problem doesn’t happen in countries where what the economy works towards is set more organically by the “invisible hand of the market” but if you think about it, it actually happens just the same as the “invisible hand of the market” actually equates “what’s good for shareholders” and what’s good for shareholders isn’t exactly always a perfect proxy for what’s good for society, to say the least... For instance it’s absolutely insane that we’ve just had 2-3 generations in the West where the best and brightest went to work for the finance industry to engineer ever more convoluted schemes to make money out of nothing, simply because it’s insanely profitable to do so. Anyone looking at this rationally can see it’s not exactly the best use of our precious human resources as a society… So all things considered, if I had to choose I’d much rather have our broad societal objectives set by human beings rather than by the theoretical concept of “what makes the most money deserves the most focus”. And as it turns out the Chinese system actually fares decently well against capitalism: human beings aren’t evidently too bad at deciding what human beings should work on if they’re being thoughtful and strategic about it.

I find this explanation of the Chinese system by Prof Keyu Jin (in a recent lecture at Harvard’s Fairbank center) absolutely fascinating. Keyu Jin is a professor of economics at LSE (London School of Economics) and serves on the board of companies like Credit Suisse. She’s also the daughter of Jin Liqun, former Vice Minister of finance of China so she’s a rare West-based academic (maybe even the only one) who actually has insight into the Chinese system from the inside. Essentially what she’s explaining is that a key reason why China was so successful economically is because of its decentralized nature, which creates two mutually compounding loops of competition, as opposed to one loop in the West. What does that mean? Well, contrary to popular belief that imagines China as being this centrally planned economy where almost everything is decided in Beijing, the inverse is actually true: China is actually one of the most decentralized countries in the world. To illustrate this, a metric that’s always amazed me is the fact that in China local governments (provinces, cities, villages, etc.) control a crazy 85% of the country's expenditures. On average that same metric for OECD countries is 33% (as in 64% of the expenditures are controlled at the federal/national level to China’s 15%). In the US for instance, which is already more decentralized than most given it’s a federation with states, only 45% of the country’s expenditures happen at the state and local level: almost twice less than in China! The effect of this, as Keyu Jin explains, is that provinces and larger municipalities in China have an immense degree of autonomy over the way they run their respective economies and fiercely compete with each other. This is the first loop. And then of course the second loop is that you have companies competing with each other in the market. As a result what constantly evolves in China is not only companies themselves but the environment in which they evolve: you constantly have this or that province running a new policy that proves very effective, making them gain an advantage vs other localities, initiative which is then copied by other localities. This makes the economic environment incredibly dynamic as it allows the state to move in unison with the economy, as opposed to slowing it down as is often the case in other countries. So what’s the role of the central government in all this? The key role, Keyu Jin argues, is setting broad objectives as well as personal management and promotion. And this is what makes the whole system work as therein lies the incentive for localities to compete with each other: because local officials know that if they do a better job than their peers, they’re on track for promotion by the central government. In “China Inc”, the central government is the board of directors and HR, presiding over an army of local CEOs with immense degrees of autonomy over their own “companies”. Keyu Jin gives the example of the solar industry. There was at some point (around 2005) a directive by the central government to develop the solar industry. The graph she shares in her talk is incredible: within a few years you had solar companies as well as patents related to research on solar technology pop up literally everywhere in China. With the result we all know about today: China today completely dominates the solar industry and solar technology (according to the International Energy Agency China's share in all the manufacturing stages of solar panels exceeds 80%). As she explains, this makes the Chinese system somewhat paradoxical as it is at the same time incredibly decentralized but also incredibly effective at mobilizing the country for centrally-decided objectives, in fact she goes as far as comparing this effectiveness to the country being in a constant state of “wartime mobilization”. An interesting comparison would be if you had all the countries in North America, the EU and North Africa (altogether roughly the population of China) all united under a common leadership deciding on common objectives and on the career path of all these countries’ officials, based on how well they achieve these objectives in their respective countries. We’re seeing this system being mobilized in its full strength today on leading edge semiconductors after US sanctions, and this is why these sanctions will undoubtedly ultimately prove so self-defeating: once the Chinese “wartime mobilization” machine is given an objective - and you can be sure this objective is prioritized very highly - the fight is essentially over, you can consider it done. Once you have hundreds of thousands of PhDs, companies and officials all at the same time competing and working within the same broad “China Inc” roof to make something happen, it will ultimately get done. If you want China NOT to develop a technology, the very last thing you want is to make them mobilize the full strength of the machine on it. With the sanctions the U.S. effectively told China: “please we beg you, do dedicate your formidable economic mobilization power to becoming a semiconductors powerhouse as fast as possible” 🤦 Another particularity of the system that Keyu Jin highlights - and I’ll end on this - is that this system also allows China to “allocate losses to certain groups of people, interest groups and sectors” in order to “enact system-level changes'', something she says is “very difficult for other governments with more political constraints to do”. For instance we’re seeing this play out in real-time with the real-estate industry: China recognized there was a housing bubble and Xi issued its “houses are for living in, not for speculation” directive. We’re since witnessing an engineered deflating of the bubble, ensuring to the extent possible that the losses are borne out by real estate developers and speculators, and not too much by society as a whole. This is part of the reason why China has never suffered a recession in the modern era: it does controlled demolition when necessary but tries to ensure it doesn’t suffer massive crises like we’ve repeatedly witnessed in the U.S. for instance. Of course no system is perfect. Weaknesses of the Chinese system include for instance local protectionism: there’s a perverse incentive for local officials to protect their local companies in order to give them a leg up vs companies from other provinces, which ultimately comes at the detriment of everyone. Another weakness is corruption, a sempiternal problem in China, where local officials - who are extremely powerful due to the nature of the system - will decide that getting promoted isn’t incentive enough and will try to cash in on their position of power. Cracking down on this is also a key remit of the central government and of course one of the major initiatives of Xi since he came to power. Lastly, another clear weakness is obviously that everything ultimately relies on the wisdom of what the system gets mobilized for, on the wisdom of these broader objectives coming from the central government. If they’re ill-thought, you effectively have a whole country working towards the wrong objectives… On this we’re often told that this problem doesn’t happen in countries where what the economy works towards is set more organically by the “invisible hand of the market” but if you think about it, it actually happens just the same as the “invisible hand of the market” actually equates “what’s good for shareholders” and what’s good for shareholders isn’t exactly always a perfect proxy for what’s good for society, to say the least... For instance it’s absolutely insane that we’ve just had 2-3 generations in the West where the best and brightest went to work for the finance industry to engineer ever more convoluted schemes to make money out of nothing, simply because it’s insanely profitable to do so. Anyone looking at this rationally can see it’s not exactly the best use of our precious human resources as a society… So all things considered, if I had to choose I’d much rather have our broad societal objectives set by human beings rather than by the theoretical concept of “what makes the most money deserves the most focus”. And as it turns out the Chinese system actually fares decently well against capitalism: human beings aren’t evidently too bad at deciding what human beings should work on if they’re being thoughtful and strategic about it.

Arnaud Bertrand

193,688 görüntüleme • 2 yıl önce

Here’s my written & video review of the new 2026 Tesla Model Y Performance after driving it for a week. This is the best-value new Tesla you can buy. Crazy performance, no real drawbacks, and all for just $57,490. Let’s dive in. Price: I haven’t seen many others mention this: every option on the Model Y Performance is included at no extra cost in the US (except FSD). So if you spec a Model Y Premium AWD with an upgraded paint color, tow package, white interior, and upgraded 20" wheels, a fully loaded Model Y Premium AWD ends up only about $2,500 less expensive than a fully loaded Model Y Performance. Ride Quality: I thought I might feel worse ride quality vs my Premium AWD Model Y with 20" wheels, but I struggled to find any real difference, despite the larger 21" wheels, firmer suspension setting, and 0.6" lower ride height on the Performance trim. A true testament to Tesla's engineering magic on this thing. Exterior Design: Unlike the previous Model Y Performance, Tesla made some exterior design tweaks with a new front and rear fascia to spice things up a bit. The result, in my opinion, is the best-looking Model Y trim you can buy. It definitely has a more aggressive presence in person, even if it's subtle. The carbon-fiber spoiler boosts high-speed stability and cuts aerodynamic drag by 10%. The new 21’’ Arachnid 2.0 wheels look fantastic in person, one of my favorite designs ever from Tesla. Staggered wheel and tire fitment provides better grip and steering. The beefier 275mm rear tires (255mm in the front) give the vehicle a better stance from behind. Vehicle-to-Load (V2L): For the first time on a Model Y in North America, you can now plug in anything you want to the exterior charge port with an adapter, even a campsite! It provides up to 2.4 kW of power (120V at 20A) from two household outlets. It's a great feature. Interior: It’s what you know and love, but with a few changes that elevate the ownership experience. New with the Performance is a larger 16" center screen (vs. 15.4" on non-Performance models), with thinner bezels and higher resolution. It’s not a huge difference on paper, but you definitely notice it in daily use. The carbon-fiber décor on the door cards and dash is a nice touch, though I would like to see it extended to the center console. The new performance seats are the best seats of any Tesla I've ever experienced. While they retain aggressive bolstering in the torso area, the bottom seat cushion has less aggressive bolstering than on the Model 3 Performance seats, making it easier to get in and out. It also doesn’t squeeze your thighs too tightly. The powered thigh extenders add comfort on longer drives, especially for taller people who want extra support. And of course, they’re heated and ventilated. The headrests also feel more comfortable than the ones in my Model Y. I want these seats. Unlike the old Model Y Performance, the new one has no Track Mode. Why? Because nobody used it lol. No point in putting engineering resources into something people won’t use. The refreshed Model 3 Performance still has it, though. Cabin Quietness: Despite the thinner-profile tires, there is no noticeable difference vs my Premium Model Y. Decibel reading results at highway speeds were visually the same compared to my 2026 Model Y Premium (65-66). Driving Impressions: It’s amazing. Sharp, precise, and agile. The vehicle feels stable at all times. Acceleration is blistering (3.3s 0–60 mph), with plenty of punch even at higher speeds. The tires offer good grip, and cornering is fantastic for an SUV. The suspension setup is great. More steering wheel feedback would be nice, though. Cruising around traffic is a joy. The brakes are much improved over the previous Model Y Performance and are far better suited for spirited driving. There are three acceleration modes: Chill, Standard, and Insane. Just stay in Insane. You’d be insane not to lol. The car also lets you switch between two ride and handling modes: Standard and Sport. The difference isn’t huge, but Standard is better if you’ve got passengers. FSD: I unfortunately wasn’t able to get FSD V14 on this car. It had V13.2.9, so I didn’t use it much. But in the little time I did, it was smooth and comfortable. It didn’t bother me because V14 will perform just as well here as it does on my 2026 Model Y Premium AWD. Conclusion: You won’t find another new SUV today that offers this level of performance for the price. Back in 2022, when Tesla couldn’t build Model Ys fast enough, a fully loaded Model Y Performance cost over $90,440. Today, the refreshed and far more capable 2026 Model Y Performance is just $57,490 fully loaded, and you can simply subscribe to FSD for $99/month. The 2026 Model Y Performance delivers utility, great performance, comfort, tech, self-driving and everything else people love about the Model Y. It’s a no-brainer purchase. I want one badly, but I'll need to show restraint, as I’m saving up for a house lol.

Sawyer Merritt

225,461 görüntüleme • 9 ay önce

**MH370x Breaking Update** My Letter to Congress Dear Congress, My name is Ashton Forbes and I am currently disclosing the most important videos in the history of the world. All of this information is publicly available. I would like to state that I have a Top Secret US Government clearance as part of my job as a contractor. My job has nothing to do with advanced technology, I am only stating this to establish my credibility. I am not bound by an NDA. This is not a hoax, disinformation, or misinformation. You don't need to believe me because everything is verifiable. I am requesting a public hearing. I am willing to testify in front of congress as to the authenticity of these videos and explain every aspect of them. I would also invite physics experts to validate the science on display. The videos in question are that of the true fate of Malaysian Airlines Flight 370, on March 7th, 2014 at 18:40UTC at the Nicobar Islands. One is of an MQ-1C Gray Eagle with a thermal layer added by the leaker, and the other is a 3D battlespace produced by the SBIRS (Space Based Infrared System), via SIGINT (Signals Intelligence) using data from Spy Satellite USA-229 which has a sister satellite next to it classified as debris. This allows for the proven 3D stereoscopic imagery we see in the Satellite video. The oldest archived versions of the videos we could find come from "RegicideAnon" a UFO video uploader, who uploaded unrelated videos previously, indicating they are not the source. Higher quality versions were released by other UFO uploaders on youtube later on, indicating neither is the original source. The dates are damning. Satellite Video (Received: March 12, 2014, Uploaded May 19, 2014) - MQ-1C Gray Eagle (Received: June 5, 2014, Uploaded June 13, 2014) - So why now, many ask? Because only in 2023 do we have the basis to understand these videos to be real. We needed the 2017 DoD Navy UAP videos to understand what FLIR footage looks like, the 2019 Trump satellite leak to understand those capabilities, 2020 scientific papers that show 'wormholes' are humanely traversable, LK-99 that shows the emergence of superconductivity and finally AI in daily use as ChatGPT. Without all these things the MH370 videos seem like magic. We know the location of the videos because the investigative group I started, MH370x, satellite experts used amateur historical trajectories to identify the correct satellite in the correct position to take the 3D stereoscopic video we see. We can see six sets of coordinates in the satellite video which we had incorrectly thought were in the South Indian Ocean, until we were able to realize the only possible location was the Nicobar Islands because the plane is turning left in both videos and due to the coordinate shifts. This means the plane is turning south and to the east. The satellite and witness (Katherine Tee) indicate the time is 18:40UTC, March 7th, 2014. Proof the Satellite video is 3D Stereoscopic - SBIRS - USA-229, the smoking gun - The Witness - I have been investigating and writing about these videos as a Citizen Journalist for the past 10 weeks. My X Corp (Twitter) following has gone from 30 followers to 8500+ of multidisciplinary backgrounds from the strength of the evidence alone. My handle is . We have definitively proved every aspect of the videos to be authentic. We know the assets, the time, the location, we have a witness, we know that there was a fire on the plane likely from the lithium ion batteries which broke containment, causing the Halon gas to permeate throughout the plane. Lithium Ion Battery Fire - There was no debris field, which is impossible for a 777 crashing into the ocean. The small amounts of debris found are consistent with the fire scenario, and some debris had burn marks. A fire suppression device from a B777 washed up in the Maldives and was not investigated despite having visible serial numbers. The Maldives were intentionally excluded from the search, despite witnesses on one of the small islands seeing the plane flying low and identifying the red/blue stripe of Malaysian Airlines early in the morning on March 8, 2014. All the Witnesses - We know that the 'official' narratives are a lie to cover up that this technology was deployed to either save the plane, or as espionage to prevent the 20 Freescale Semiconductor scientists onboard from going to China. I suspect they are integral to the technology we see. We know the plane didn't crash into the ocean because the SOSUS system didn't hear the acoustic sound. The same system that heard the Titan sub pop and the Navy lied about it for 5 days while oxygen counters were on every major news channel. The Diego Garcia hydrophones and Western Australia hydrophones also didn't hear it. Missing Hydrophone Data - I wish that this being MH370 was the most important part of these videos, because ultimately I'm doing this for the families of the victims, the witnesses, the leaker of these videos, and the world that has been lied to. However, the most important part of these videos is that they prove conclusively that the US Government is hiding Superconductivity, Teleportation, and 'Free Energy' from the world. Everything we see in these videos can be explained by science. This is not 'aliens' in these videos. This is our technology. The assets are filming the plane before the orbs even show up. The drone cannot catch a 777-200, it must have intercepted it. This is a US Government operation. I do not believe that we could be this secretly advanced without a reverse engineering program. The orbs in this video are ignoring gravity, being pulled forward by some 'gravity engine' indicated by the dark lines in front of them, and upon intercept are traveling at estimated Mach3 speeds. The explanation for their pattern is artificial intelligence, a computer program. Superconductive Harmonic Orbs - Traversable Wormholes - The videos do not show annihilation because E=MC^2 and the 'zap' would be much larger. It's not an explosion because it's cold in the thermal, not hot. It's a black hole. It's also not 'cloaking' because the smoke stops when the plane disappears. It has to be teleportation based on science. A wormhole. I found out, to my own surprise, that humanely traversable wormholes are theoretically possible. This singularity is causing a transitional phase state change in the plane where it reverts to a wave function and obtains a probabilistic nature. This is only possible with superconductivity and ‘free energy’ technology. Macroscopic Decoherence (how the plane was teleported) - I was also able to identify the leaker of the videos as being Lieutenant Commander Edward C. Lin. He checks every box to be the leaker. Our Government attempted to put him in prison for life as a traitor, but he is no spy. Edward C. Lin is a hero. He took a plea deal after he was convinced he damaged national security, but he never revealed this information to our enemies. He simply wanted to do the right thing and tell the world the truth of this technology and what happened to MH370. He was sentenced to 9 years in prison and likely is bound by his plea deal to never speak about the videos again. We can vindicate this man as well as everyone else who was lied to or discredited. Evidence that the leaker is Edward C. Lin - I also got a tip from a source that told me that the nephew of Retired General Joseph F. Dunford has seen the videos and indirectly confirmed their authenticity. He was likely in charge of the operation as the commander of the International Security Assistance Force in 2014. He then served as the 19th Chairman of the Joint Chiefs of Staff, the nation’s highest-ranking military officer, and the principal military advisor to the President, Secretary of Defense, and National Security Council from Oct. 1, 2015, through Sept. 30, 2019. He is now on the board of directors at Lockheed Martin. While the destination of the plane is speculative, the most likely place is Diego Garcia. The witnesses in the Maldives indicate that, as well as the American passenger Phillip Wood EXIF data photo, that points to Diego Garica where he claimed to be held prisoner. I'm not sure what happened to the passengers. It is scientifically possible that some survived. None of the families of the victims have reached out to me. I am operating under the assumption that they were returned to their various countries in exchange for their silence, and Phillip Wood may be in Witness Protection. Phillip Wood EXIF photo is not 'fake' - I know the videos seem impossible, but they are authentic. They are not CGI. We were able to show that the satellite video is a Citrix session logged into the actual spy satellite database due to the framerate discrepancy between the mouse we see, and the background. (24fps vs 6fps) Also, hundreds of community VFX experts have analyzed the footage frame by frame and not a single discrepancy can be found. The list of requirements to ‘hoax’ the videos is practically impossible. What it would take to 'hoax' the MH370 videos - It's not CGI (see replies) - These videos implicate the US Government in a black budget reverse engineering program alla whistleblower David Grusch's sworn testimony to congress. This information will shock you and I hope you will seek to expose the truth. This surpasses petty politics. Please help for the good of the country. This is a verifiable conspiracy, and while I believe that it is being done because the forces that be think that society will collapse if this technology is made publicly available, I disagree. We can handle this information and the technology has the capability to change the circumstances of millions if not billions of people. What matters in this world is the time we spend in it, with the people we care about. That will not change. The reason I am requesting a public congressional hearing is because I believe the weight of the evidence can convince enough people in congress that the true events of MH370 were covered up by our own government and that there is a black budget advanced technology program, most likely based on reversed engineered Unidentified Aerial Phenomenon (UAP). Sincerely, -Ashton Forbes #MH370x #MH370

MH370x Breaking Update My Letter to Congress Dear Congress, My name is Ashton Forbes and I am currently disclosing the most important videos in the history of the world. All of this information is publicly available. I would like to state that I have a Top Secret US Government clearance as part of my job as a contractor. My job has nothing to do with advanced technology, I am only stating this to establish my credibility. I am not bound by an NDA. This is not a hoax, disinformation, or misinformation. You don't need to believe me because everything is verifiable. I am requesting a public hearing. I am willing to testify in front of congress as to the authenticity of these videos and explain every aspect of them. I would also invite physics experts to validate the science on display. The videos in question are that of the true fate of Malaysian Airlines Flight 370, on March 7th, 2014 at 18:40UTC at the Nicobar Islands. One is of an MQ-1C Gray Eagle with a thermal layer added by the leaker, and the other is a 3D battlespace produced by the SBIRS (Space Based Infrared System), via SIGINT (Signals Intelligence) using data from Spy Satellite USA-229 which has a sister satellite next to it classified as debris. This allows for the proven 3D stereoscopic imagery we see in the Satellite video. The oldest archived versions of the videos we could find come from "RegicideAnon" a UFO video uploader, who uploaded unrelated videos previously, indicating they are not the source. Higher quality versions were released by other UFO uploaders on youtube later on, indicating neither is the original source. The dates are damning. Satellite Video (Received: March 12, 2014, Uploaded May 19, 2014) - MQ-1C Gray Eagle (Received: June 5, 2014, Uploaded June 13, 2014) - So why now, many ask? Because only in 2023 do we have the basis to understand these videos to be real. We needed the 2017 DoD Navy UAP videos to understand what FLIR footage looks like, the 2019 Trump satellite leak to understand those capabilities, 2020 scientific papers that show 'wormholes' are humanely traversable, LK-99 that shows the emergence of superconductivity and finally AI in daily use as ChatGPT. Without all these things the MH370 videos seem like magic. We know the location of the videos because the investigative group I started, MH370x, satellite experts used amateur historical trajectories to identify the correct satellite in the correct position to take the 3D stereoscopic video we see. We can see six sets of coordinates in the satellite video which we had incorrectly thought were in the South Indian Ocean, until we were able to realize the only possible location was the Nicobar Islands because the plane is turning left in both videos and due to the coordinate shifts. This means the plane is turning south and to the east. The satellite and witness (Katherine Tee) indicate the time is 18:40UTC, March 7th, 2014. Proof the Satellite video is 3D Stereoscopic - SBIRS - USA-229, the smoking gun - The Witness - I have been investigating and writing about these videos as a Citizen Journalist for the past 10 weeks. My X Corp (Twitter) following has gone from 30 followers to 8500+ of multidisciplinary backgrounds from the strength of the evidence alone. My handle is . We have definitively proved every aspect of the videos to be authentic. We know the assets, the time, the location, we have a witness, we know that there was a fire on the plane likely from the lithium ion batteries which broke containment, causing the Halon gas to permeate throughout the plane. Lithium Ion Battery Fire - There was no debris field, which is impossible for a 777 crashing into the ocean. The small amounts of debris found are consistent with the fire scenario, and some debris had burn marks. A fire suppression device from a B777 washed up in the Maldives and was not investigated despite having visible serial numbers. The Maldives were intentionally excluded from the search, despite witnesses on one of the small islands seeing the plane flying low and identifying the red/blue stripe of Malaysian Airlines early in the morning on March 8, 2014. All the Witnesses - We know that the 'official' narratives are a lie to cover up that this technology was deployed to either save the plane, or as espionage to prevent the 20 Freescale Semiconductor scientists onboard from going to China. I suspect they are integral to the technology we see. We know the plane didn't crash into the ocean because the SOSUS system didn't hear the acoustic sound. The same system that heard the Titan sub pop and the Navy lied about it for 5 days while oxygen counters were on every major news channel. The Diego Garcia hydrophones and Western Australia hydrophones also didn't hear it. Missing Hydrophone Data - I wish that this being MH370 was the most important part of these videos, because ultimately I'm doing this for the families of the victims, the witnesses, the leaker of these videos, and the world that has been lied to. However, the most important part of these videos is that they prove conclusively that the US Government is hiding Superconductivity, Teleportation, and 'Free Energy' from the world. Everything we see in these videos can be explained by science. This is not 'aliens' in these videos. This is our technology. The assets are filming the plane before the orbs even show up. The drone cannot catch a 777-200, it must have intercepted it. This is a US Government operation. I do not believe that we could be this secretly advanced without a reverse engineering program. The orbs in this video are ignoring gravity, being pulled forward by some 'gravity engine' indicated by the dark lines in front of them, and upon intercept are traveling at estimated Mach3 speeds. The explanation for their pattern is artificial intelligence, a computer program. Superconductive Harmonic Orbs - Traversable Wormholes - The videos do not show annihilation because E=MC^2 and the 'zap' would be much larger. It's not an explosion because it's cold in the thermal, not hot. It's a black hole. It's also not 'cloaking' because the smoke stops when the plane disappears. It has to be teleportation based on science. A wormhole. I found out, to my own surprise, that humanely traversable wormholes are theoretically possible. This singularity is causing a transitional phase state change in the plane where it reverts to a wave function and obtains a probabilistic nature. This is only possible with superconductivity and ‘free energy’ technology. Macroscopic Decoherence (how the plane was teleported) - I was also able to identify the leaker of the videos as being Lieutenant Commander Edward C. Lin. He checks every box to be the leaker. Our Government attempted to put him in prison for life as a traitor, but he is no spy. Edward C. Lin is a hero. He took a plea deal after he was convinced he damaged national security, but he never revealed this information to our enemies. He simply wanted to do the right thing and tell the world the truth of this technology and what happened to MH370. He was sentenced to 9 years in prison and likely is bound by his plea deal to never speak about the videos again. We can vindicate this man as well as everyone else who was lied to or discredited. Evidence that the leaker is Edward C. Lin - I also got a tip from a source that told me that the nephew of Retired General Joseph F. Dunford has seen the videos and indirectly confirmed their authenticity. He was likely in charge of the operation as the commander of the International Security Assistance Force in 2014. He then served as the 19th Chairman of the Joint Chiefs of Staff, the nation’s highest-ranking military officer, and the principal military advisor to the President, Secretary of Defense, and National Security Council from Oct. 1, 2015, through Sept. 30, 2019. He is now on the board of directors at Lockheed Martin. While the destination of the plane is speculative, the most likely place is Diego Garcia. The witnesses in the Maldives indicate that, as well as the American passenger Phillip Wood EXIF data photo, that points to Diego Garica where he claimed to be held prisoner. I'm not sure what happened to the passengers. It is scientifically possible that some survived. None of the families of the victims have reached out to me. I am operating under the assumption that they were returned to their various countries in exchange for their silence, and Phillip Wood may be in Witness Protection. Phillip Wood EXIF photo is not 'fake' - I know the videos seem impossible, but they are authentic. They are not CGI. We were able to show that the satellite video is a Citrix session logged into the actual spy satellite database due to the framerate discrepancy between the mouse we see, and the background. (24fps vs 6fps) Also, hundreds of community VFX experts have analyzed the footage frame by frame and not a single discrepancy can be found. The list of requirements to ‘hoax’ the videos is practically impossible. What it would take to 'hoax' the MH370 videos - It's not CGI (see replies) - These videos implicate the US Government in a black budget reverse engineering program alla whistleblower David Grusch's sworn testimony to congress. This information will shock you and I hope you will seek to expose the truth. This surpasses petty politics. Please help for the good of the country. This is a verifiable conspiracy, and while I believe that it is being done because the forces that be think that society will collapse if this technology is made publicly available, I disagree. We can handle this information and the technology has the capability to change the circumstances of millions if not billions of people. What matters in this world is the time we spend in it, with the people we care about. That will not change. The reason I am requesting a public congressional hearing is because I believe the weight of the evidence can convince enough people in congress that the true events of MH370 were covered up by our own government and that there is a black budget advanced technology program, most likely based on reversed engineered Unidentified Aerial Phenomenon (UAP). Sincerely, -Ashton Forbes #MH370x #MH370

Ashton Forbes

7,911,009 görüntüleme • 2 yıl önce

$ASTI Ascent Solar Technologies Space and Drone Solar Panels The "Going to Zero" or Mispriced Space/Drone Solar Play Intro and comparison to $RKLB and $RDW panels Let’s get the ugly stuff out of the way first. $ASTI is a distressed penny stock with a ~$5M-$10M market cap. • They burn millions in cash. • 2024 Revenue: ~$40k. 2025 Revenue (YTD): ~$60k. • They generate less revenue than a single Tesla Model Y. • They have diluted shareholders relentlessly. $ASTI just raised $2M in December with the potential of $3.5M more via warrants while being a ~$5M mcap "company". Yikes. To most, this is "uninvestable trash." Stay away. Full stop. So why did I buy ~5% of the float? IF the technology works and IF they execute then I believe this is a massive market pricing dislocation about to inflect. They have been grinding for years and may finally be hitting an inflection point. $RKLB Rocketlab is the king of space solar and they are my second largest position overall, but here is why $ASTI might be a very high risk but asymmetric bet in Space & Defense right now. 1. The Tech Pivot: Flexible CIGS vs. The World Ascent started in 2005 but pivoted 2 years ago from consumer to pure-play Space & Defense. They have sunk ~$250M and 20 years of R&D into proprietary CIGS (Copper-Indium-Gallium-Selenide) thin-film technology while building out fully domestic and vertically integrated manufacturing capabilities. The Physics: • Thickness: 0.03 mm (Thinner than paper). • Flexibility: Wraps around drones/satellites; rolls up like a poster. • Durability: "Self-Healing" capabilities against space radiation. Can take a bullet or micrometeoroid and keep working. Can handle shocks/vibration. Does not shatter. The Metric that Matters: Specific Power (W/kg) (aka energy to weight ratio) In space, mass means cost and difficult decision decisions. • Rocket Lab ($RKLB) / Spectrolab: ~150 W/kg (System level). • Ascent Solar ($ASTI): ~1,960 W/kg (Module level). $ASTI is roughly 10x lighter for the same power output potential (mass-wise). This frees up design limitations and cost. 2. The Competition: $RKLB & $RDW Rocket Lab (SolAero) & Redwire (iROSA): • Tech: Rigid Crystal Cells (Multi-junction) embedded in a fabric mesh. • Pros: Extreme Efficiency (~30%+). Perfect for limited surface area. • Cons: Heavy, Brittle, Expensive ($3k-$10k per Watt). Manufacturing multi-junction cells (SolAero) involves slowly growing crystals in a vacuum chamber. With radiation the panels degrade and loose efficiency over time which will limit the satellite lifespan. • Use Case: James Webb Telescope, Flagship missions. Ascent Solar (ASTI): • Tech: Flexible Thin-Film on Plastic. • Pros: Ultra-light, Durable, Cheap ($500-$1k per Watt). Manufacturing CIGS is roughly similar to printing newspapers (roll-to-roll). The panels are radiation degradation resistant and will outlive the satellite • Cons: Lower Efficiency (~17.5%). Requires 2x surface area. • Use Case: Mega-Constellations (Starlink/Amazon Leo), Small/Low cost satellites, Drones, Deformable surfaces. The lower efficiency is not an ASTI failing. It is the inherent physics trade-off of not using glass/rigid silicone. The downside however is increased atmospheric drag with very larger/massive panel sheets. Because ASTI modules are ~50% less efficient than rigid panels, they require ~2x the physical surface area to generate the same amount of power. In GEO (High Orbit): Drag doesn't matter. Weight savings are king. A massive solar array allows for more sensors and longer project lifespan. ASTI is highly competitive here. In LEO (Low Orbit): Atmospheric drag is real. A massive solar array acts like a large parachute, causing the satellite to de-orbit faster unless it burns more fuel to stay up. At LEO, smaller satellites are a better fit for ASTI. 3. Durability & Radiation "Self-Healing" Radiation Hardness This is ASTI's "Ace in the Hole" for physics. The Problem: In space, high-energy protons (radiation) smash into solar cells, creating atomic "defects" that trap electrons. Over time, this kills the panel's power output (degradation). The CIGS Advantage: CIGS (Copper-Indium-Gallium-Selenide) material has a unique property where heat (annealing) allows the atomic structure to relax and "heal" these defects. Self-Healing: Because CIGS heals at relatively low temperatures (often achieved just by the sun heating the panel), it suffers significantly less degradation than traditional Silicon or even some GaAs panels over long missions in high-radiation belts (like MEO or GEO). Lifespan: While a rigid GaAs panel might lose 15-20% of its power over 15 years (enough to kill a satellite), CIGS panels heal and can maintain a flatter power curve, potentially outlasting the satellite itself in high-radiation orbits. 4. Brittleness & Flexibility ASTI (CIGS on Polyimide): Flexible. You can roll it like a poster. It can take a bullet or micrometeoroid and the hole will just be a dead spot; the rest of the panel keeps working. It does not shatter. Redwire (ROSA) & Rocket Lab (SolAero): Brittle Cells on a Flex Blanket. $RDW's ROSA (Roll-Out Solar Array) typically uses rigid multi-junction cells (made by SolAero/Rocket Lab or Spectrolab) mounted on a flexible mesh fabric. The Risk: If you bend the cells too far, they crack. They rely on the mesh backing for flexibility, but the active generating material is still a brittle crystal wafer. Much heavier, more expensive, and less durable than $ASTI's option 5. The Inflection Point (Why Now?) After years of silent struggle, late 2025 has seen an explosion of activity. Recent Agreements (Nov/Dec 2025): NovaSpark: Hydrogen-powered military drones. $ASTI panels generate power in the field → NovaSpark creates hydrogen fuel. CisLunar Industries: Integrating ASTI solar with power conversion hardware for deep space longevity. Defiant Space: A strategic alliance to act as the "door opener" for classified DoD/NATO programs. More headlines: Ascent Solar Technologies Provides Leading Space Company with Thin-Film PV modules for Spacecraft Power Generation Testing in Cislunar Space December 03, 2025 08:00 ET Ascent Solar Technologies Delivers Thin-Film PV for Saltwater Environment Durability and Space-Based Power Beaming Testing October 14, 2025 08:00 ET Ascent Solar Enters Teaming Agreement with Emtel Energy USA to Advance Thin-Film PV Energy Storage Capabilities September 16, 2025 08:00 ET Ascent Solar Technologies Signs MOU with Star Catcher Industries to Improve Power Capabilities for Thin-Film Solar Technology in Space August 28, 2025 08:00 ET Ascent Solar Technologies Establishes Rapid Thin-Film PV Delivery Process to Provide Customized Space Solar Products Ahead of Schedule on Mission Enabling Timelines August 07, 2025 08:00 ET The Pipeline (From Aug Corporate Presentation) 18 new NDA's signed in 2025. They are field testing with 3 major players: • Company A: Mega-constellation (+2,500 satellites). • Company B: Space Defense (Explicitly mentioned "Golden Dome"). • Company C: Satellite Manufacturer (30-200 unit scale). Management: New board members include a former founding member of SpaceX and a retired Air Force General and Deputy Assistant Secretary for Contracting (acquisitions expert). The company started in 2005 based out of Colorado, but two years ago pivoted to Space & Defense and away from consumer applications. Made in USA: Defense contracts heavily favor domestic supply chains. ASTI manufactures in Colorado. This is a huge moat against cheap Chinese solar. In their Q3 report they note that their market has seen sudden recent acceleration. The space solar industry is currently only capable of 8 to 12 MW per year of production meanwhile the demand is growing to over 100 MW per year. 6. The Risk (The Sword of Damocles) ⚠️ This is critical. $ASTI just raised ~$2M in December. Attached to that raise are ~2 Million Warrants with a strike price of $1.70. These are exercisable immediately. If the stock rips to $3.00, warrant holders exercise at $1.70 and dump on the market for a risk-free 76% profit. This creates a massive "sell wall" and potential 40% dilution of the float. Summary: This is a binary bet. • Bear Case: They run out of cash in 6 months, dilution spirals, stock goes to $0. • Bull Case: They land one of the "Company A/B/C" contracts. Revenue jumps from $60k to projected $20M+ in 2026. The stock reprices from a "bankrupt penny stock" to a "critical defense/space supplier." I have gradually accumulated ~5% of the float. I am ready for it to go to zero. But if the space economy demands "Cheap, Light, and Durable," $ASTI is the only public pure-play. Disclaimer: This is a very high-risk microcap. Do your own due diligence. Not financial advice.

$ASTI Ascent Solar Technologies Space and Drone Solar Panels The "Going to Zero" or Mispriced Space/Drone Solar Play Intro and comparison to $RKLB and $RDW panels Let’s get the ugly stuff out of the way first. $ASTI is a distressed penny stock with a ~$5M-$10M market cap. • They burn millions in cash. • 2024 Revenue: ~$40k. 2025 Revenue (YTD): ~$60k. • They generate less revenue than a single Tesla Model Y. • They have diluted shareholders relentlessly. $ASTI just raised $2M in December with the potential of $3.5M more via warrants while being a ~$5M mcap "company". Yikes. To most, this is "uninvestable trash." Stay away. Full stop. So why did I buy ~5% of the float? IF the technology works and IF they execute then I believe this is a massive market pricing dislocation about to inflect. They have been grinding for years and may finally be hitting an inflection point. $RKLB Rocketlab is the king of space solar and they are my second largest position overall, but here is why $ASTI might be a very high risk but asymmetric bet in Space & Defense right now. 1. The Tech Pivot: Flexible CIGS vs. The World Ascent started in 2005 but pivoted 2 years ago from consumer to pure-play Space & Defense. They have sunk ~$250M and 20 years of R&D into proprietary CIGS (Copper-Indium-Gallium-Selenide) thin-film technology while building out fully domestic and vertically integrated manufacturing capabilities. The Physics: • Thickness: 0.03 mm (Thinner than paper). • Flexibility: Wraps around drones/satellites; rolls up like a poster. • Durability: "Self-Healing" capabilities against space radiation. Can take a bullet or micrometeoroid and keep working. Can handle shocks/vibration. Does not shatter. The Metric that Matters: Specific Power (W/kg) (aka energy to weight ratio) In space, mass means cost and difficult decision decisions. • Rocket Lab ($RKLB) / Spectrolab: ~150 W/kg (System level). • Ascent Solar ($ASTI): ~1,960 W/kg (Module level). $ASTI is roughly 10x lighter for the same power output potential (mass-wise). This frees up design limitations and cost. 2. The Competition: $RKLB & $RDW Rocket Lab (SolAero) & Redwire (iROSA): • Tech: Rigid Crystal Cells (Multi-junction) embedded in a fabric mesh. • Pros: Extreme Efficiency (~30%+). Perfect for limited surface area. • Cons: Heavy, Brittle, Expensive ($3k-$10k per Watt). Manufacturing multi-junction cells (SolAero) involves slowly growing crystals in a vacuum chamber. With radiation the panels degrade and loose efficiency over time which will limit the satellite lifespan. • Use Case: James Webb Telescope, Flagship missions. Ascent Solar (ASTI): • Tech: Flexible Thin-Film on Plastic. • Pros: Ultra-light, Durable, Cheap ($500-$1k per Watt). Manufacturing CIGS is roughly similar to printing newspapers (roll-to-roll). The panels are radiation degradation resistant and will outlive the satellite • Cons: Lower Efficiency (~17.5%). Requires 2x surface area. • Use Case: Mega-Constellations (Starlink/Amazon Leo), Small/Low cost satellites, Drones, Deformable surfaces. The lower efficiency is not an ASTI failing. It is the inherent physics trade-off of not using glass/rigid silicone. The downside however is increased atmospheric drag with very larger/massive panel sheets. Because ASTI modules are ~50% less efficient than rigid panels, they require ~2x the physical surface area to generate the same amount of power. In GEO (High Orbit): Drag doesn't matter. Weight savings are king. A massive solar array allows for more sensors and longer project lifespan. ASTI is highly competitive here. In LEO (Low Orbit): Atmospheric drag is real. A massive solar array acts like a large parachute, causing the satellite to de-orbit faster unless it burns more fuel to stay up. At LEO, smaller satellites are a better fit for ASTI. 3. Durability & Radiation "Self-Healing" Radiation Hardness This is ASTI's "Ace in the Hole" for physics. The Problem: In space, high-energy protons (radiation) smash into solar cells, creating atomic "defects" that trap electrons. Over time, this kills the panel's power output (degradation). The CIGS Advantage: CIGS (Copper-Indium-Gallium-Selenide) material has a unique property where heat (annealing) allows the atomic structure to relax and "heal" these defects. Self-Healing: Because CIGS heals at relatively low temperatures (often achieved just by the sun heating the panel), it suffers significantly less degradation than traditional Silicon or even some GaAs panels over long missions in high-radiation belts (like MEO or GEO). Lifespan: While a rigid GaAs panel might lose 15-20% of its power over 15 years (enough to kill a satellite), CIGS panels heal and can maintain a flatter power curve, potentially outlasting the satellite itself in high-radiation orbits. 4. Brittleness & Flexibility ASTI (CIGS on Polyimide): Flexible. You can roll it like a poster. It can take a bullet or micrometeoroid and the hole will just be a dead spot; the rest of the panel keeps working. It does not shatter. Redwire (ROSA) & Rocket Lab (SolAero): Brittle Cells on a Flex Blanket. $RDW's ROSA (Roll-Out Solar Array) typically uses rigid multi-junction cells (made by SolAero/Rocket Lab or Spectrolab) mounted on a flexible mesh fabric. The Risk: If you bend the cells too far, they crack. They rely on the mesh backing for flexibility, but the active generating material is still a brittle crystal wafer. Much heavier, more expensive, and less durable than $ASTI's option 5. The Inflection Point (Why Now?) After years of silent struggle, late 2025 has seen an explosion of activity. Recent Agreements (Nov/Dec 2025): NovaSpark: Hydrogen-powered military drones. $ASTI panels generate power in the field → NovaSpark creates hydrogen fuel. CisLunar Industries: Integrating ASTI solar with power conversion hardware for deep space longevity. Defiant Space: A strategic alliance to act as the "door opener" for classified DoD/NATO programs. More headlines: Ascent Solar Technologies Provides Leading Space Company with Thin-Film PV modules for Spacecraft Power Generation Testing in Cislunar Space December 03, 2025 08:00 ET Ascent Solar Technologies Delivers Thin-Film PV for Saltwater Environment Durability and Space-Based Power Beaming Testing October 14, 2025 08:00 ET Ascent Solar Enters Teaming Agreement with Emtel Energy USA to Advance Thin-Film PV Energy Storage Capabilities September 16, 2025 08:00 ET Ascent Solar Technologies Signs MOU with Star Catcher Industries to Improve Power Capabilities for Thin-Film Solar Technology in Space August 28, 2025 08:00 ET Ascent Solar Technologies Establishes Rapid Thin-Film PV Delivery Process to Provide Customized Space Solar Products Ahead of Schedule on Mission Enabling Timelines August 07, 2025 08:00 ET The Pipeline (From Aug Corporate Presentation) 18 new NDA's signed in 2025. They are field testing with 3 major players: • Company A: Mega-constellation (+2,500 satellites). • Company B: Space Defense (Explicitly mentioned "Golden Dome"). • Company C: Satellite Manufacturer (30-200 unit scale). Management: New board members include a former founding member of SpaceX and a retired Air Force General and Deputy Assistant Secretary for Contracting (acquisitions expert). The company started in 2005 based out of Colorado, but two years ago pivoted to Space & Defense and away from consumer applications. Made in USA: Defense contracts heavily favor domestic supply chains. ASTI manufactures in Colorado. This is a huge moat against cheap Chinese solar. In their Q3 report they note that their market has seen sudden recent acceleration. The space solar industry is currently only capable of 8 to 12 MW per year of production meanwhile the demand is growing to over 100 MW per year. 6. The Risk (The Sword of Damocles) ⚠️ This is critical. $ASTI just raised ~$2M in December. Attached to that raise are ~2 Million Warrants with a strike price of $1.70. These are exercisable immediately. If the stock rips to $3.00, warrant holders exercise at $1.70 and dump on the market for a risk-free 76% profit. This creates a massive "sell wall" and potential 40% dilution of the float. Summary: This is a binary bet. • Bear Case: They run out of cash in 6 months, dilution spirals, stock goes to $0. • Bull Case: They land one of the "Company A/B/C" contracts. Revenue jumps from $60k to projected $20M+ in 2026. The stock reprices from a "bankrupt penny stock" to a "critical defense/space supplier." I have gradually accumulated ~5% of the float. I am ready for it to go to zero. But if the space economy demands "Cheap, Light, and Durable," $ASTI is the only public pure-play. Disclaimer: This is a very high-risk microcap. Do your own due diligence. Not financial advice.

YeahDave

208,571 görüntüleme • 9 ay önce

That's a terrible take from Peter McCormack 🏴‍☠️🇬🇧🇮🇪. I don't hold any $ADA, and I don't even like Cardano much, but this post is highly misleading and the fruit of misinformation. Peter should do his research better. This Cardano Bitcoin implementation is apparently far more secure than most Lightning Network implementations, for example, which he used to promote relentlessly to his followers. Maxis will be maxis, I guess. 🤷 - I asked OpenAI o1 via Nano-GPT to review and analyze BOS whitepaper, commenting on the trade-offs, vulnerabilities, risks, and BTC influencers calling it a scam. Here's what the AI said: The BitcoinOS whitepaper introduces **BitSNARK** and **Grail**, proposing an innovative approach to enhance Bitcoin's functionality by enabling trustless and secure bridges between Bitcoin and other blockchain networks, such as Cardano. The core of this proposal is to optimize the recently introduced **BitVM** paradigm for verifying zk-SNARK proofs on the Bitcoin blockchain without altering Bitcoin's consensus rules. Below, I provide an analysis of the whitepaper, its security implications compared to existing Layer 2 (L2) solutions like the Lightning Network and Liquid, potential trade-offs and vulnerabilities, user considerations to mitigate risks, and its prospects for the future. Additionally, I will address the criticisms from Bitcoin maximalists labeling it a "scam." --- ### **Overview of BitSNARK and Grail** **BitVM Overview** - **BitVM**, introduced by Robin Linus in December 2023, allows for Turing-complete computations on Bitcoin by leveraging an interactive protocol between parties to verify off-chain computations. - It opens the possibility for near trustless rollup bridges on Bitcoin without changing its consensus rules. - However, BitVM, being general-purpose, is not optimized for specific tasks like SNARK verification, leading to practical implementation challenges. **BitSNARK's Innovations** - **BitSNARK** builds upon BitVM but focuses specifically on verifying zk-SNARK proofs efficiently on the Bitcoin blockchain. - It introduces a simplified virtual machine (VM) with only three instructions optimized for finite field calculations required in zk-SNARK verification: - `addmod` for modular addition. - `andbit` for bitwise operations. - `equal` for equality checks. - By reducing complexity, BitSNARK improves program size by an order of magnitude and reduces challenge/response lengths by up to 50%. - It simplifies the challenge protocol to a single type, enhancing security and auditability. **Grail Bridge Implementation** - **Grail** is an implementation of BitSNARK, aiming to provide a practical and scalable Bitcoin Rollup Bridge. - It enables users to transfer assets between the Bitcoin mainchain (Layer 1) and Layer 2 networks (rollups) securely. - Grail relies on a set of **operators** who participate in the protocol to facilitate deposits and withdrawals. - Operators engage in an interactive verification protocol, ensuring that zk-SNARK proofs are correctly verified on-chain. - The system incentivizes honest participation through economic incentives and penalties. --- ### **Security Analysis** #### **Comparison with Existing L2 Solutions** **Lightning Network** - **Pros:** - Offers fast, low-cost transactions off-chain. - Preserves Bitcoin's on-chain privacy features. - **Cons:** - Requires continuous network connectivity. - Involves counterparty risk due to channel management. - Limited in handling complex smart contracts or interoperability with other chains. **Liquid Network** - **Pros:** - Federated sidechain allowing faster transactions and confidential assets. - Enables issuance of new assets and atomic swaps. - **Cons:** - Federation model introduces trust assumptions. - Less decentralized due to reliance on functionaries. **BitSNARK and Grail** - **Pros:** - Enables verification of zk-SNARK proofs on Bitcoin without consensus changes. - Facilitates truly trustless and decentralized bridges to other chains. - Reduces complexity and enhances efficiency compared to BitVM. - Operators' incentives align to maintain security without requiring majority trust. - **Cons:** - Relies on an interactive protocol between operators, potentially increasing complexity. - New technology may have undiscovered vulnerabilities. - Requires widespread operator participation for robustness. #### **Potential Vulnerabilities** 1. **Operator Collusion or Malfeasance:** - If operators collude, they could potentially censor transactions or delay withdrawals. - Mitigation: Implement robust economic incentives and penalties, and promote decentralization by encouraging more operators to participate. 2. **Interactive Protocol Risks:** - The interactive verification protocol relies on timely responses; delays or Denial-of-Service (DoS) attacks could disrupt the protocol. - Mitigation: Incorporate timeouts and fail-safes, and design protocols resilient to network latency and attacks. 3. **Smart Contract Bugs:** - Bugs in the implementation of BitSNARK VM or the Grail protocol could be exploited. - Mitigation: Conduct thorough code audits, formal verification, and testing on testnets before mainnet deployment. 4. **Cryptographic Assumptions:** - The security relies on the cryptographic primitives used in zk-SNARKs and their correct implementation. - Mitigation: Use well-established cryptographic libraries and stay updated on advancements in cryptanalysis. 5. **Scalability Concerns:** - Although more efficient than BitVM, running zk-SNARK verifications on-chain is resource-intensive. - Mitigation: Optimize the implementation further and possibly leverage layer 2 solutions for certain computations. --- ### **Trade-offs and User Considerations** **Trade-offs:** - **Complexity vs. Functionality:** - BitSNARK simplifies the computation model but requires users and operators to understand and trust a new protocol. - **Decentralization vs. Performance:** - The security model relies on operators; increasing the number of operators enhances decentralization but may impact performance due to coordination overhead. - **Innovation vs. Maturity:** - As a novel solution, it lacks the battle-tested resilience of existing L2 solutions like the Lightning Network. **User Considerations:** 1. **Counterparty Risk:** - Users should be aware that the system's security depends on operators; participation of reputable operators is crucial. - **Recommendation:** Use the bridge when a sufficient number of trusted operators are active. 2. **Protocol Familiarity:** - Users need to understand the withdrawal and dispute processes. - **Recommendation:** Educate users through clear documentation and user-friendly interfaces. 3. **Security Practices:** - Be cautious with private keys and exit secrets, especially for operators. - **Recommendation:** Implement best practices for key management and consider hardware wallets. 4. **Monitoring Network Conditions:** - Network congestion or attacks could affect transaction confirmations. - **Recommendation:** Monitor network status and set appropriate fees for timely confirmations. --- ### **Potential for the Future** - **Solving the Scalability Trilemma:** - BitSNARK and Grail present a promising approach to achieving scalability, expressivity, and decentralization on Bitcoin. - **Interoperability:** - Enables Bitcoin to interact trustlessly with other chains, expanding its utility beyond a store of value. - **Innovation Catalyst:** - May inspire further research and development in Bitcoin's smart contract capabilities. --- ### **Addressing the "Scam" Accusations by Bitcoin Maximalists** **Criticisms:** - **Skepticism of Altering Bitcoin's Core Principles:** - Some Bitcoin maximalists may view this proposal as deviating from Bitcoin's original purpose. - **Fear of Centralization Risks:** - Concerns that introducing complex protocols could introduce central points of failure. **Counterarguments:** - **Maintaining Consensus Rules:** - BitSNARK operates without changing Bitcoin's consensus rules, preserving its core principles. - **Enhancing Bitcoin's Utility:** - The proposal aims to expand Bitcoin's capabilities while upholding decentralization and security. - **Economic Incentives for Security:** - The design relies on economic incentives aligning participants' interests with the network's security. **Conclusion:** - **Healthy Skepticism is Beneficial:** - Critiques promote rigorous security assessments and robustness. - **Innovation Requires Open-mindedness:** - Dismissing new ideas outright may hinder Bitcoin's evolution and adaptability. - **Focus on Merits and Risks:** - A balanced evaluation considering potential benefits and drawbacks is essential. --- ### **Final Thoughts** The BitcoinOS whitepaper introduces a significant advancement in enabling trustless and secure interoperability between Bitcoin and other blockchain networks. By optimizing for zk-SNARK verification through BitSNARK and implementing the Grail bridge, it aims to address long-standing scalability and functionality limitations. While promising, it is crucial to approach this innovation with careful scrutiny, thorough testing, and a commitment to security best practices. Users and stakeholders should remain informed, participate in community discussions, and contribute to the development and auditing processes to ensure the system's robustness and integrity.

That's a terrible take from Peter McCormack 🏴‍☠️🇬🇧🇮🇪. I don't hold any $ADA, and I don't even like Cardano much, but this post is highly misleading and the fruit of misinformation. Peter should do his research better. This Cardano Bitcoin implementation is apparently far more secure than most Lightning Network implementations, for example, which he used to promote relentlessly to his followers. Maxis will be maxis, I guess. 🤷 - I asked OpenAI o1 via Nano-GPT to review and analyze BOS whitepaper, commenting on the trade-offs, vulnerabilities, risks, and BTC influencers calling it a scam. Here's what the AI said: The BitcoinOS whitepaper introduces BitSNARK and Grail, proposing an innovative approach to enhance Bitcoin's functionality by enabling trustless and secure bridges between Bitcoin and other blockchain networks, such as Cardano. The core of this proposal is to optimize the recently introduced BitVM paradigm for verifying zk-SNARK proofs on the Bitcoin blockchain without altering Bitcoin's consensus rules. Below, I provide an analysis of the whitepaper, its security implications compared to existing Layer 2 (L2) solutions like the Lightning Network and Liquid, potential trade-offs and vulnerabilities, user considerations to mitigate risks, and its prospects for the future. Additionally, I will address the criticisms from Bitcoin maximalists labeling it a "scam." --- ### Overview of BitSNARK and Grail BitVM Overview - BitVM, introduced by Robin Linus in December 2023, allows for Turing-complete computations on Bitcoin by leveraging an interactive protocol between parties to verify off-chain computations. - It opens the possibility for near trustless rollup bridges on Bitcoin without changing its consensus rules. - However, BitVM, being general-purpose, is not optimized for specific tasks like SNARK verification, leading to practical implementation challenges. BitSNARK's Innovations - BitSNARK builds upon BitVM but focuses specifically on verifying zk-SNARK proofs efficiently on the Bitcoin blockchain. - It introduces a simplified virtual machine (VM) with only three instructions optimized for finite field calculations required in zk-SNARK verification: - `addmod` for modular addition. - `andbit` for bitwise operations. - `equal` for equality checks. - By reducing complexity, BitSNARK improves program size by an order of magnitude and reduces challenge/response lengths by up to 50%. - It simplifies the challenge protocol to a single type, enhancing security and auditability. Grail Bridge Implementation - Grail is an implementation of BitSNARK, aiming to provide a practical and scalable Bitcoin Rollup Bridge. - It enables users to transfer assets between the Bitcoin mainchain (Layer 1) and Layer 2 networks (rollups) securely. - Grail relies on a set of operators who participate in the protocol to facilitate deposits and withdrawals. - Operators engage in an interactive verification protocol, ensuring that zk-SNARK proofs are correctly verified on-chain. - The system incentivizes honest participation through economic incentives and penalties. --- ### Security Analysis #### Comparison with Existing L2 Solutions Lightning Network - Pros: - Offers fast, low-cost transactions off-chain. - Preserves Bitcoin's on-chain privacy features. - Cons: - Requires continuous network connectivity. - Involves counterparty risk due to channel management. - Limited in handling complex smart contracts or interoperability with other chains. Liquid Network - Pros: - Federated sidechain allowing faster transactions and confidential assets. - Enables issuance of new assets and atomic swaps. - Cons: - Federation model introduces trust assumptions. - Less decentralized due to reliance on functionaries. BitSNARK and Grail - Pros: - Enables verification of zk-SNARK proofs on Bitcoin without consensus changes. - Facilitates truly trustless and decentralized bridges to other chains. - Reduces complexity and enhances efficiency compared to BitVM. - Operators' incentives align to maintain security without requiring majority trust. - Cons: - Relies on an interactive protocol between operators, potentially increasing complexity. - New technology may have undiscovered vulnerabilities. - Requires widespread operator participation for robustness. #### Potential Vulnerabilities 1. Operator Collusion or Malfeasance: - If operators collude, they could potentially censor transactions or delay withdrawals. - Mitigation: Implement robust economic incentives and penalties, and promote decentralization by encouraging more operators to participate. 2. Interactive Protocol Risks: - The interactive verification protocol relies on timely responses; delays or Denial-of-Service (DoS) attacks could disrupt the protocol. - Mitigation: Incorporate timeouts and fail-safes, and design protocols resilient to network latency and attacks. 3. Smart Contract Bugs: - Bugs in the implementation of BitSNARK VM or the Grail protocol could be exploited. - Mitigation: Conduct thorough code audits, formal verification, and testing on testnets before mainnet deployment. 4. Cryptographic Assumptions: - The security relies on the cryptographic primitives used in zk-SNARKs and their correct implementation. - Mitigation: Use well-established cryptographic libraries and stay updated on advancements in cryptanalysis. 5. Scalability Concerns: - Although more efficient than BitVM, running zk-SNARK verifications on-chain is resource-intensive. - Mitigation: Optimize the implementation further and possibly leverage layer 2 solutions for certain computations. --- ### Trade-offs and User Considerations Trade-offs: - Complexity vs. Functionality: - BitSNARK simplifies the computation model but requires users and operators to understand and trust a new protocol. - Decentralization vs. Performance: - The security model relies on operators; increasing the number of operators enhances decentralization but may impact performance due to coordination overhead. - Innovation vs. Maturity: - As a novel solution, it lacks the battle-tested resilience of existing L2 solutions like the Lightning Network. User Considerations: 1. Counterparty Risk: - Users should be aware that the system's security depends on operators; participation of reputable operators is crucial. - Recommendation: Use the bridge when a sufficient number of trusted operators are active. 2. Protocol Familiarity: - Users need to understand the withdrawal and dispute processes. - Recommendation: Educate users through clear documentation and user-friendly interfaces. 3. Security Practices: - Be cautious with private keys and exit secrets, especially for operators. - Recommendation: Implement best practices for key management and consider hardware wallets. 4. Monitoring Network Conditions: - Network congestion or attacks could affect transaction confirmations. - Recommendation: Monitor network status and set appropriate fees for timely confirmations. --- ### Potential for the Future - Solving the Scalability Trilemma: - BitSNARK and Grail present a promising approach to achieving scalability, expressivity, and decentralization on Bitcoin. - Interoperability: - Enables Bitcoin to interact trustlessly with other chains, expanding its utility beyond a store of value. - Innovation Catalyst: - May inspire further research and development in Bitcoin's smart contract capabilities. --- ### Addressing the "Scam" Accusations by Bitcoin Maximalists Criticisms: - Skepticism of Altering Bitcoin's Core Principles: - Some Bitcoin maximalists may view this proposal as deviating from Bitcoin's original purpose. - Fear of Centralization Risks: - Concerns that introducing complex protocols could introduce central points of failure. Counterarguments: - Maintaining Consensus Rules: - BitSNARK operates without changing Bitcoin's consensus rules, preserving its core principles. - Enhancing Bitcoin's Utility: - The proposal aims to expand Bitcoin's capabilities while upholding decentralization and security. - Economic Incentives for Security: - The design relies on economic incentives aligning participants' interests with the network's security. Conclusion: - Healthy Skepticism is Beneficial: - Critiques promote rigorous security assessments and robustness. - Innovation Requires Open-mindedness: - Dismissing new ideas outright may hinder Bitcoin's evolution and adaptability. - Focus on Merits and Risks: - A balanced evaluation considering potential benefits and drawbacks is essential. --- ### Final Thoughts The BitcoinOS whitepaper introduces a significant advancement in enabling trustless and secure interoperability between Bitcoin and other blockchain networks. By optimizing for zk-SNARK verification through BitSNARK and implementing the Grail bridge, it aims to address long-standing scalability and functionality limitations. While promising, it is crucial to approach this innovation with careful scrutiny, thorough testing, and a commitment to security best practices. Users and stakeholders should remain informed, participate in community discussions, and contribute to the development and auditing processes to ensure the system's robustness and integrity.

Vini B |「 thecoding 」

58,614 görüntüleme • 1 yıl önce

$"900,000 BARRELS JUST ARRIVED IN JAPAN": President Trump's Truth Social Post Lands as Beijing's Hormuz Strategy Officially Implodes — Inside the Tanker That Just Rewrote the Indo-Pacific Energy Map The M/V Otis didn't just dock in Tokyo Bay. It docked on top of the CCP's entire wedge-strategy thesis. "HUGE MOMENT! Asia is getting their oil from the United States now. 900,000 barrels just arrived in Japan. America will lead the charge on oil dominance." President Trump's post — accompanied by an ANN News screenshot of an American crude tanker easing into Tokyo Bay at dawn — is doing the kind of work that ten policy white papers cannot. It is taking a moment of strategic transformation and stamping it onto the public consciousness in a single image: U.S. oil. Japanese port. Middle East bypassed. And for once, the substance behind the showmanship checks out. What Actually Happened in Tokyo Bay The tanker is the Suezmax-class M/V Otis (IMO 9408217). On the morning of Sunday, April 26, she eased up to an offshore jetty in Tokyo Bay carrying approximately 910,000 barrels of Texas light crude oil bound for a refinery in Chiba Prefecture. The cargo was transferred through an undersea pipeline to a facility operated by Cosmo Oil, a subsidiary of Cosmo Energy Holdings, where it will be processed into petroleum products including gasoline for domestic distribution. The voyage itself is the part that matters strategically. The Otis loaded in Texas on March 22 and completed a roughly 35-day voyage through the Panama Canal — one of the largest direct U.S. crude deliveries to Japan in years. The alternative routing — Cape of Good Hope, which takes about 55 days — was bypassed in favor of the Panama Canal, cutting transit time by roughly 20 days. This is the operational answer to a question Beijing assumed had no answer: Can American crude actually reach Japanese refineries fast enough, in sufficient volume, to matter when Hormuz is contested? The Otis is the receipt. The Real Number Is Not 910,000. It's 4x. A single tanker, as honest reporting from noted, amounts to less than one day's consumption in Japan. Taken in isolation, 910,000 barrels is a symbol, not a strategy. The strategy lives in the trendline. According to Japanese government documents reviewed by Reuters, Japanese imports of U.S. crude oil for May will be four times higher than they were a year earlier — up from a May 2025 baseline of 189,000 barrels daily, which represented about 8% of total imports that month. Tokyo also expects that by May, it will have secured half of its imports from suppliers outside the Middle East, with the United States the largest among those alternative suppliers, joined by Malaysia, Azerbaijan, Brazil, Nigeria, and Angola. For a country that historically sourced as much as 95% of its oil from the Middle East, this is the fastest reorientation of a G7 energy supply chain in the post-Cold War era. The Otis is the first tanker in a queue, not the last. The $56 Billion Bet Behind the Tanker None of this is happening on autopilot. On March 14, at the Asia-Pacific Energy Security Forum in Tokyo, Japan signed agreements worth up to $56 billion with the United States covering oil, natural gas, and LNG purchases and investments — sitting inside the broader framework from the 2025 U.S.-Japan trade agreement, under which Japan pledged $550 billion in U.S. investments, with energy as a key pillar. Five days later, Prime Minister Sanae Takaichi — Japan's first woman prime minister and one of the most China-skeptical leaders Tokyo has produced in a generation — walked into the Oval Office and, in front of cameras, embraced President Trump. The visit had a few rough edges: Trump's unprompted Pearl Harbor reference, made when a Japanese reporter asked why allies hadn't been warned about the February 28 strike on Iran, appeared to take Takaichi aback. But on substance the meeting did exactly what an alliance summit is supposed to do — it pre-positioned the energy pivot the Otis would later make material. By May 19, that pivot had widened into a trilateral. In Seoul, President Lee Jae-myung and Prime Minister Takaichi agreed to expand LNG cooperation under the bilateral Supply and Demand Cooperation Agreement signed in March, and to deepen information sharing and communication channels related to crude oil supply, demand, and stockpiling — explicitly framed by Seoul as a vehicle for South Korea–Japan and South Korea–U.S.–Japan cooperation for regional peace and stability. The Hormuz shock was supposed to fracture this triangle. Instead, it welded it. What Beijing Got Wrong — And Why It Matters The Chinese Communist Party's strategic miscalculation in the Hormuz file was not merely tactical. It was conceptual. For roughly a decade, Beijing's working thesis on Asian energy security has been that the United States cannot credibly underwrite the region from outside the Persian Gulf, and that any Middle East flashpoint would therefore translate into political leverage for China — the buyer of last resort with the deepest stockpiles, the longest supply contracts, and the most flexible sanctions tolerance. The thesis has just failed three stress tests at once. First, the logistics held. Alaskan crude can reach Japanese refineries via Pacific routes about a week faster than Middle Eastern shipments, and Texas crude via Panama — as the Otis proved — gets there in 35 days. Tokyo is now actively weighing expanded Alaskan imports and a joint U.S.-Japan strategic crude reserve arrangement. Second, the political will held. Takaichi did not hedge. She unilaterally began releasing 15 days' worth of private-sector reserves from March 16, followed by a month's worth of state-held oil, and committed Japan to participating in the IEA's coordinated 400-million-barrel release, with Japan contributing 80 million barrels — 54 million in crude and 26 million in oil products. Third — and most damaging to Beijing's strategic narrative — China's own position deteriorated. While Tokyo was diversifying westward across the Pacific, Iran continued sending the bulk of the crude still moving — roughly 1.22 million barrels per day — to China, after a record 2.16 million bpd in February that was entirely destined for Beijing as it amassed reserves. The CCP's "energy security" turned out to be a deeper handcuff to a sanctioned, militarily degraded supplier whose primary export terminal — Kharg Island, the departure point for roughly 90% of Iran's crude exports — has been struck by U.S. forces. Beijing did not de-risk. It concentrated risk. And when Trump publicly pressured Beijing to help secure Hormuz on the grounds that 90% of Chinese oil flowed through it, Chinese state spokespeople were reduced to publicly emphasizing that the country had "enough" energy reserves — an answer that managed to be both defensive and, in strategic terms, an admission. The Honest Caveats Serious analysis requires acknowledging what the headline glosses over. The Otis cargo is, by Japan's own ministry, less than one day's consumption. U.S. crude exports cannot fully substitute for Middle Eastern barrels: industry analysts cited by Axios put the realistic monthly ceiling for U.S. crude exports in the 5.5 million bpd range, with Gulf Coast port and terminal capacity acting as a hard infrastructure limit. The Iran war has been genuinely costly — Brent crude topped $110 a barrel in late March before retreating to roughly $98 by late May — and the global growth picture has been marked down accordingly. There are also legitimate technical questions about crude grade compatibility (U.S. light sweet vs. Middle Eastern medium sour) that Japanese refineries will need to manage. The diversification away from Hormuz is real, but it is not free, not frictionless, and not complete. These caveats sharpen the conclusion. They do not invert it. The Strategic Bottom Line Three months ago, the Hormuz file was supposed to be the lever that pried Tokyo loose from Washington. The arithmetic looked plausible on paper: Japan and South Korea are the third- and fourth-largest destinations for crude moving through the Strait of Hormuz, behind only China and India. Pain there should, in theory, have created political space for Beijing. It didn't. The pain became the catalyst. A $56 billion energy package. A four-fold surge in U.S. crude bound for Japanese refineries. A Suezmax tanker easing into Tokyo Bay at dawn carrying 910,000 barrels from Texas. A trilateral Tokyo–Seoul–Washington cooperation framework formalized in the middle of a war Beijing was counting on to crack it. There is a lesson here for anyone still treating CCP economic statecraft as inevitable: leverage that depends on your rival having no alternative evaporates the moment one is built. Tokyo built one. Washington underwrote it. Trump is now broadcasting it. And the manifests in Yokohama harbor are the proof. The alliance held. The wedge failed. The map has been redrawn — and the next tanker is already loading. Original article by me Aric Chen. Views are my own — welcome to discuss! © 2026 Aric Chen. All rights reserved. Any unauthorized use will be reported under the DMCA.$

"900,000 BARRELS JUST ARRIVED IN JAPAN": President Trump's Truth Social Post Lands as Beijing's Hormuz Strategy Officially Implodes — Inside the Tanker That Just Rewrote the Indo-Pacific Energy Map The M/V Otis didn't just dock in Tokyo Bay. It docked on top of the CCP's entire wedge-strategy thesis. "HUGE MOMENT! Asia is getting their oil from the United States now. 900,000 barrels just arrived in Japan. America will lead the charge on oil dominance." President Trump's post — accompanied by an ANN News screenshot of an American crude tanker easing into Tokyo Bay at dawn — is doing the kind of work that ten policy white papers cannot. It is taking a moment of strategic transformation and stamping it onto the public consciousness in a single image: U.S. oil. Japanese port. Middle East bypassed. And for once, the substance behind the showmanship checks out. What Actually Happened in Tokyo Bay The tanker is the Suezmax-class M/V Otis (IMO 9408217). On the morning of Sunday, April 26, she eased up to an offshore jetty in Tokyo Bay carrying approximately 910,000 barrels of Texas light crude oil bound for a refinery in Chiba Prefecture. The cargo was transferred through an undersea pipeline to a facility operated by Cosmo Oil, a subsidiary of Cosmo Energy Holdings, where it will be processed into petroleum products including gasoline for domestic distribution. The voyage itself is the part that matters strategically. The Otis loaded in Texas on March 22 and completed a roughly 35-day voyage through the Panama Canal — one of the largest direct U.S. crude deliveries to Japan in years. The alternative routing — Cape of Good Hope, which takes about 55 days — was bypassed in favor of the Panama Canal, cutting transit time by roughly 20 days. This is the operational answer to a question Beijing assumed had no answer: Can American crude actually reach Japanese refineries fast enough, in sufficient volume, to matter when Hormuz is contested? The Otis is the receipt. The Real Number Is Not 910,000. It's 4x. A single tanker, as honest reporting from noted, amounts to less than one day's consumption in Japan. Taken in isolation, 910,000 barrels is a symbol, not a strategy. The strategy lives in the trendline. According to Japanese government documents reviewed by Reuters, Japanese imports of U.S. crude oil for May will be four times higher than they were a year earlier — up from a May 2025 baseline of 189,000 barrels daily, which represented about 8% of total imports that month. Tokyo also expects that by May, it will have secured half of its imports from suppliers outside the Middle East, with the United States the largest among those alternative suppliers, joined by Malaysia, Azerbaijan, Brazil, Nigeria, and Angola. For a country that historically sourced as much as 95% of its oil from the Middle East, this is the fastest reorientation of a G7 energy supply chain in the post-Cold War era. The Otis is the first tanker in a queue, not the last. The $56 Billion Bet Behind the Tanker None of this is happening on autopilot. On March 14, at the Asia-Pacific Energy Security Forum in Tokyo, Japan signed agreements worth up to $56 billion with the United States covering oil, natural gas, and LNG purchases and investments — sitting inside the broader framework from the 2025 U.S.-Japan trade agreement, under which Japan pledged $550 billion in U.S. investments, with energy as a key pillar. Five days later, Prime Minister Sanae Takaichi — Japan's first woman prime minister and one of the most China-skeptical leaders Tokyo has produced in a generation — walked into the Oval Office and, in front of cameras, embraced President Trump. The visit had a few rough edges: Trump's unprompted Pearl Harbor reference, made when a Japanese reporter asked why allies hadn't been warned about the February 28 strike on Iran, appeared to take Takaichi aback. But on substance the meeting did exactly what an alliance summit is supposed to do — it pre-positioned the energy pivot the Otis would later make material. By May 19, that pivot had widened into a trilateral. In Seoul, President Lee Jae-myung and Prime Minister Takaichi agreed to expand LNG cooperation under the bilateral Supply and Demand Cooperation Agreement signed in March, and to deepen information sharing and communication channels related to crude oil supply, demand, and stockpiling — explicitly framed by Seoul as a vehicle for South Korea–Japan and South Korea–U.S.–Japan cooperation for regional peace and stability. The Hormuz shock was supposed to fracture this triangle. Instead, it welded it. What Beijing Got Wrong — And Why It Matters The Chinese Communist Party's strategic miscalculation in the Hormuz file was not merely tactical. It was conceptual. For roughly a decade, Beijing's working thesis on Asian energy security has been that the United States cannot credibly underwrite the region from outside the Persian Gulf, and that any Middle East flashpoint would therefore translate into political leverage for China — the buyer of last resort with the deepest stockpiles, the longest supply contracts, and the most flexible sanctions tolerance. The thesis has just failed three stress tests at once. First, the logistics held. Alaskan crude can reach Japanese refineries via Pacific routes about a week faster than Middle Eastern shipments, and Texas crude via Panama — as the Otis proved — gets there in 35 days. Tokyo is now actively weighing expanded Alaskan imports and a joint U.S.-Japan strategic crude reserve arrangement. Second, the political will held. Takaichi did not hedge. She unilaterally began releasing 15 days' worth of private-sector reserves from March 16, followed by a month's worth of state-held oil, and committed Japan to participating in the IEA's coordinated 400-million-barrel release, with Japan contributing 80 million barrels — 54 million in crude and 26 million in oil products. Third — and most damaging to Beijing's strategic narrative — China's own position deteriorated. While Tokyo was diversifying westward across the Pacific, Iran continued sending the bulk of the crude still moving — roughly 1.22 million barrels per day — to China, after a record 2.16 million bpd in February that was entirely destined for Beijing as it amassed reserves. The CCP's "energy security" turned out to be a deeper handcuff to a sanctioned, militarily degraded supplier whose primary export terminal — Kharg Island, the departure point for roughly 90% of Iran's crude exports — has been struck by U.S. forces. Beijing did not de-risk. It concentrated risk. And when Trump publicly pressured Beijing to help secure Hormuz on the grounds that 90% of Chinese oil flowed through it, Chinese state spokespeople were reduced to publicly emphasizing that the country had "enough" energy reserves — an answer that managed to be both defensive and, in strategic terms, an admission. The Honest Caveats Serious analysis requires acknowledging what the headline glosses over. The Otis cargo is, by Japan's own ministry, less than one day's consumption. U.S. crude exports cannot fully substitute for Middle Eastern barrels: industry analysts cited by Axios put the realistic monthly ceiling for U.S. crude exports in the 5.5 million bpd range, with Gulf Coast port and terminal capacity acting as a hard infrastructure limit. The Iran war has been genuinely costly — Brent crude topped $110 a barrel in late March before retreating to roughly $98 by late May — and the global growth picture has been marked down accordingly. There are also legitimate technical questions about crude grade compatibility (U.S. light sweet vs. Middle Eastern medium sour) that Japanese refineries will need to manage. The diversification away from Hormuz is real, but it is not free, not frictionless, and not complete. These caveats sharpen the conclusion. They do not invert it. The Strategic Bottom Line Three months ago, the Hormuz file was supposed to be the lever that pried Tokyo loose from Washington. The arithmetic looked plausible on paper: Japan and South Korea are the third- and fourth-largest destinations for crude moving through the Strait of Hormuz, behind only China and India. Pain there should, in theory, have created political space for Beijing. It didn't. The pain became the catalyst. A $56 billion energy package. A four-fold surge in U.S. crude bound for Japanese refineries. A Suezmax tanker easing into Tokyo Bay at dawn carrying 910,000 barrels from Texas. A trilateral Tokyo–Seoul–Washington cooperation framework formalized in the middle of a war Beijing was counting on to crack it. There is a lesson here for anyone still treating CCP economic statecraft as inevitable: leverage that depends on your rival having no alternative evaporates the moment one is built. Tokyo built one. Washington underwrote it. Trump is now broadcasting it. And the manifests in Yokohama harbor are the proof. The alliance held. The wedge failed. The map has been redrawn — and the next tanker is already loading. Original article by me Aric Chen. Views are my own — welcome to discuss! © 2026 Aric Chen. All rights reserved. Any unauthorized use will be reported under the DMCA.

Aric Chen

17,007 görüntüleme • 3 ay önce

77 Reasons Why I’ve Invested Over $8,000,000+ in MultiversX (EGLD) and Why EGLD Will Crush It in 2025 (My Investment Thesis). I publicly shared my portfolio on X. EGLD is A) Better than BTC B) Everything that ETH wants to be C) The GameStop of Crypto 1. EGLD is verifiably the most scalable (theoretically unlimited) L1 chain in the world, theoretically capable of over 10 million TPS (thanks to adaptive state sharding). 2. e-Gold is digital gold. It has the best tokenomics among all L1s, similarly scarce to BTC, with a maximum supply of 31.4 million coins. Currently, 27.68 million coins are in circulation. 3. EGLD will be the most decentralized cryptocurrency in the world thanks to sharding and minimal hardware requirements for running nodes. It’s already second only to Ethereum with 3,618 validator nodes. 4. EGLD has extremely low fees, around ~$0.002 per transaction. 5. EGLD is extremely secure. No wallet drains like on ETH/SOL; assets are owned natively (not via a smart contract). There is no MEV risk (front-running bots). 6. EGLD is the only chain in the world with an on-chain Guardian (two-phase verification), making it impossible for a hacker to steal your funds—even if they have your private keys (seed phrase). 7. EGLD is carbon-neutral and eco-friendly, not wasting energy like BTC and other PoW chains. It’s exceptionally efficient, scalable, global, and sustainable. 8. EGLD has the best UX in crypto. Download the xPortal wallet—it’s like discovering Apple in Web3. The interface is simple, flawless, and you barely realize you’re using crypto. Instead of addresses, you use HeroTags. The app features all dApps, everything runs smoothly, and the visuals are beautifully designed. The explorer, web wallet, etc. follow the same high-quality user experience. 9. EGLD supports native assets, unlike Ethereum, for example. 10. EGLD is the first chain to fully implement horizontal (theoretically unlimited) sharding without compromising on decentralization—unlike Solana and others that attempt vertical scaling, leading to multiple network downtimes (11+ times) and huge hardware demands for validators, ultimately harming decentralization. 11. EGLD makes setting up a validator agency extremely easy. Even complete IT beginners can do it. The UX and documentation are superb. I personally set up the “EGLDSqueeze” agency in about 30 minutes. Managing it is straightforward via the web wallet, which feels like managing a Facebook page. This simplifies decentralization enormously. 12. EGLD allows literally anyone (even your grandma) to participate in decentralization, since nodes can run on a Raspberry Pi or a relatively affordable phone. Imagine millions of people worldwide securing the network, validating transactions without even knowing it. This can’t be done with BTC, where setting up profitable mining operations is prohibitively expensive. 13. WASM-Based Virtual Machine: You can write smart contracts in your favorite language, compile them, and run them via the fastest VM in the world. 14. EGLD has been tested at an incredible 263,000 TPS using its sharding mechanism and low hardware requirements. Allegedly, by mid-next year (April), they’ll demonstrate 1,000,000 TPS. (For context: Mastercard handles around 5,000 TPS; BTC handles 5–7 TPS.) 15. EGLD is currently the most advanced L1 in terms of scalability, security, decentralization, UX, eco-friendliness, and tokenomics. It’s the only chain that has genuinely solved the Blockchain Trilemma and is ready to onboard 1 billion people into crypto—users who won’t even realize they’re interacting with crypto. 16. EGLD is perfectly positioned for AI projects—AI agents, AI tools, or a so-called “Truth Machine” that monitors other AIs on-chain, documenting what’s true and comparing different AI outputs (some of which may be censored or biased), ensuring people don’t get confused or scammed in an AI-driven world. 17. The EGLD team is the hardest-working team I’ve ever encountered. I had the honor of meeting many of them personally, and can attest that their pace—even during a bear market—is extraordinary. 18. EGLD’s development team is exceptionally active on GitHub, continually improving their network and actively committing code. 19. EGLD plans to introduce an update reducing block time to 600ms (down from ~6 seconds), which would make the chain essentially unrivaled. 20. EGLD is effectively the only usable L1 in Europe, and the team has direct connections within the EU government—extremely bullish for the project. 21. EGLD provides top-tier on-chain governance not only for the MultiversX (EGLD) protocol but also for DeFi projects (e.g., xExchange, MEX). 22. EGLD plans to expand to the US, likely opening offices in Austin, Texas. This could put them in direct contact with Elon Musk (if it hasn’t happened already), as he’s involved with If he’s done his research, he’d discover there’s simply no better L1 worldwide. 23. EGLD solved fully implemented sharding, perfect tokenomics, and top-tier architecture with just $5M, whereas other chains failed to do so even with $100M+. The second-best sharding network, NEAR, needed $100M, has worse tokenomics, and its sharding isn’t fully implemented yet. Its UX also doesn’t compare. Owning NEAR was like comparing a VW Golf R to a Porsche GT3—EGLD is the Porsche GT3. 24. According to Similarweb, EGLD has significantly high traffic relative to other chains with market caps 100x larger. The market cap vs. web traffic discrepancy is huge, which is a strong indicator of EGLD’s potential. 25. EGLD has the most active and dedicated community relative to its user base, with users who believe in the technology, have full faith in the team, and remain loyal despite price volatility—because they use the chain and know there’s nothing better. 26. Check other chains’ active user counts on X (Twitter) and compare it with the followers of EGLD’s founders and main network accounts, versus those with 30x, 50x, or 100x larger market caps. 27. Visit the MultiversX website to observe the futuristic design and presentation, then compare it to other chains that appear nearly a decade behind in design and branding. 28. EGLD hosts the xDay Global event, showcasing updates, new builders, projects in the ecosystem, and major announcements—similar to Apple’s Keynotes—delivered in a highly professional, goosebump-inducing atmosphere. The next event is in Korea, the second-biggest crypto market after the US. Check out their previous xDay after-movie to see why this is extremely bullish. 29. EGLD is moving forward with plans for the first regulated, audited EU stablecoin under MiCa regulation, made possible by acquiring xMoney, which I view as a “Stripe” for crypto/fiat, offering everything from user solutions to merchant services—potentially the future of payments. 30. Greg Siourouni recently joined EGLD, having been an executive director at SUI Foundation. He’s now co-founder of xMoney Global. xMoney (formerly UTrust, with token UTK) is owned and founded by the MultiversX Labs team. A stablecoin might be introduced soon, which would be massively bullish given xMoney’s roadmap. They recently announced integrations with Binance Pay—both ways. 31. EGLD prioritizes user safety, believing it’s the only feasible approach once the network scales to serve a billion people—many of whom are retail users with little to no security awareness. 32. EGLD offers “Sovereign Chains,” letting you effectively clone their chain without heavy development, set up your own validators, and leverage their unlimited scalability. Any blockchain (ETH, BTC, SOL) struggling with scalability, decentralization, or security could run an ultra-fast, scalable, and secure L2 on EGLD’s Sovereign Chain, meeting top enterprise requirements. No one else has really done this. The Sovereign Chain demo achieved astonishing TPS and has an SDK. 33. No downtime since inception. 34. No shard takeover attacks have occurred. 35. Extremely fast—soon 600ms block time will be in place. 36. ESDTs – The best token standard available: fungible, non-fungible, semi-fungible, DeFi assets—everything is native and highly customizable. 37. Top-tier composability of assets and smart contracts. 38. Integrated DNS at protocol level with HeroTags (nicknames) instead of long addresses. 39. Asynchronous calls are supported. 40. Cross-shard transfers, execution, reverts, and calls are seamlessly integrated. 41. The best staking system in the space. Secure Proof of Stake (SPoS) is far more efficient than Proof of Work (PoW). 42. Built-in Delegation and Staking Provider system, with over 125K delegators. 43. Complete support for liquid staked assets, fostering decentralization rather than centralization. 44. TransferRoles for ESDT and other advanced operations. 45. Composable tasks on-chain for more sophisticated DeFi workflows. 46. MultiTransfer and asset execution within one transaction. 47. Re-entrancy protection is built-in by design. 48. Storage for ESDT assets goes beyond a linear approach, optimizing performance. 49. No integer overflows thanks to integrated safeMath operations. 50. Integrated crypto opcodes in the VM, enhancing security and performance. 51. Support for BigFloats, BigInts, and BigDecimals, enabling advanced financial calculations on-chain. 52. No sandwich attacks, plus front-running and MEV protection. 53. Relayed Transactions, simplifying user interactions and fees. 54. Smart Accounts featuring data tries and multiple built-in functions. 55. Generalized Paymaster solutions, enabling flexible fee models. 56. Subscriptions for recurring or automated on-chain payments. 57. Web2-like usability with Web3 functionality, bridging mainstream adoption. 58. StakingV4 for improved decentralization. 59. Enhanced MEV protection rolling out to safeguard users. 60. Parallel execution is coming soon, boosting throughput. 61. 1 million TPS is on the roadmap, targeted for demonstration. 62. 600ms block time is also coming soon. 63. Reduced cross-shard processing is planned to improve efficiency. 64. ZK everywhere (PI²): “prove everything” approach is coming. 65. AsyncV3 is in development for more complex cross-contract interactions. 66. Scalability enhancements for Merkle Tries or a new data model are being explored. 67. Linear storage on the VM is forthcoming. 68. A dynamic language interpreter at the VM is also planned. 69. Rumors suggest that MultiversX (EGLD) is building a “Truth Machine” on their L1—an essential, game-changing tool for AI verification and societal impact. 70. The entire team features individuals with PhDs in mathematics and physics, and many are former engineers at Google, IBM, and similar companies. 71. Over 56% of the network’s supply is staked, showcasing strong community involvement. 72. More than 6,772,347 accounts have been created on the network. 73. A total of 476,627,710 transactions have been processed on-chain without any outages or hacks. 74. EGLD has built a massive ecosystem over time. While not as numerous in project count as Solana, its market cap is ~100x smaller, yet it has far superior tokenomics and technology. The projects that do exist, like Hatom Protocol, are top-tier in UX, security, and advanced features. Hatom will soon introduce USH, a truly high-quality, decentralized stablecoin. 75. On competing chains, automated transactions aren’t easily or cheaply executed, whereas on MultiversX, tools like let you do this for free (with near-zero fees). 76. No other chain combines such a strong team and long-term vision where every product meets extreme security and UX standards like MultiversX does. This is why I see it as the “next Apple” in Web3. 77. MultiversX has a new CMO – Adam Bates, a former CMO at the Cardano Foundation. He was behind the success of Cardano’s huge marketing campaign and has a very good relationship with Charles Hoskinson. Thanks to him, Beniamin Mincu (the founder of MultiversX) was likely introduced, and now they will probably discuss how both blockchains can help each other, as well as any other potential collaborations we don’t yet know about. This is also extremely bullish. #EGLD is undeniably the most Scalable, Advanced, Secure, and User-friendly L1 supercomputer ever created. It’s built to SHAPE THE FUTURE. 1) 2) 3) 4) 5) 27/6/2024 - EGLDSqueeze - SUMMARY: HERE IS NO 2ND BEST. EGLD IS ONLY ONE BLOCKCHAIN THAT CAN RULE THEM ALL. ✅ UNLIMITED SCALING ✅ SCARCE AS BTC ✅ PROGRAMMABLE AS ETH ✅ NO DOWNTIME AS SOL ✅ UI/UX OF Apple ✅ SHARDING DONE BEFORE NEAR & TON ✅ BEST WALLET xPortal WITH GUARDIAN Price prediction (NFA|DYOR): My reasoning is that the real market cap as of December 23, 2024...if we take into account the value of other cryptocurrencies such as BTC, SOL, ETH, AVAX, NEAR, TON, Cardano, BNB, XRP, and so forth, plus the existence of meme coins with valuations above 20 billion USD, or even games nobody plays anymore that still have valuations above 800 million shows that EGLD’s current market cap of approximately 942 million USD is incredibly low. From a technological standpoint, user experience, and other relevant aspects, compared to SOL, NEAR, TON, AVAX, and other L1 protocols, EGLD’s market cap should realistically be around 100 billion USD. Therefore, my prediction and investment thesis is a minimum of a 100x increase from its current price (+-SOL marketcap). MultiversX is ready to onboard 1 billion people to the blockchain. From a long-term perspective, it could even reach a market cap of 1 trillion USD, which is roughly half of where BTC is right now. That would be approximately a 1060x gain from the current market cap. 1 EGLD (MultiversX) is for $34 (only 31.4M max supply) think about this. Not financial advice. Again. There is no 2nd best L1. Position yourself where the puck is going, then wait at the goal until the goal gets there Apes together, strong. Ape alone, weak. We Don't Worry. We Just Win. Shape The Future

77 Reasons Why I’ve Invested Over $8,000,000+ in MultiversX (EGLD) and Why EGLD Will Crush It in 2025 (My Investment Thesis). I publicly shared my portfolio on X. EGLD is A) Better than BTC B) Everything that ETH wants to be C) The GameStop of Crypto 1. EGLD is verifiably the most scalable (theoretically unlimited) L1 chain in the world, theoretically capable of over 10 million TPS (thanks to adaptive state sharding). 2. e-Gold is digital gold. It has the best tokenomics among all L1s, similarly scarce to BTC, with a maximum supply of 31.4 million coins. Currently, 27.68 million coins are in circulation. 3. EGLD will be the most decentralized cryptocurrency in the world thanks to sharding and minimal hardware requirements for running nodes. It’s already second only to Ethereum with 3,618 validator nodes. 4. EGLD has extremely low fees, around ~$0.002 per transaction. 5. EGLD is extremely secure. No wallet drains like on ETH/SOL; assets are owned natively (not via a smart contract). There is no MEV risk (front-running bots). 6. EGLD is the only chain in the world with an on-chain Guardian (two-phase verification), making it impossible for a hacker to steal your funds—even if they have your private keys (seed phrase). 7. EGLD is carbon-neutral and eco-friendly, not wasting energy like BTC and other PoW chains. It’s exceptionally efficient, scalable, global, and sustainable. 8. EGLD has the best UX in crypto. Download the xPortal wallet—it’s like discovering Apple in Web3. The interface is simple, flawless, and you barely realize you’re using crypto. Instead of addresses, you use HeroTags. The app features all dApps, everything runs smoothly, and the visuals are beautifully designed. The explorer, web wallet, etc. follow the same high-quality user experience. 9. EGLD supports native assets, unlike Ethereum, for example. 10. EGLD is the first chain to fully implement horizontal (theoretically unlimited) sharding without compromising on decentralization—unlike Solana and others that attempt vertical scaling, leading to multiple network downtimes (11+ times) and huge hardware demands for validators, ultimately harming decentralization. 11. EGLD makes setting up a validator agency extremely easy. Even complete IT beginners can do it. The UX and documentation are superb. I personally set up the “EGLDSqueeze” agency in about 30 minutes. Managing it is straightforward via the web wallet, which feels like managing a Facebook page. This simplifies decentralization enormously. 12. EGLD allows literally anyone (even your grandma) to participate in decentralization, since nodes can run on a Raspberry Pi or a relatively affordable phone. Imagine millions of people worldwide securing the network, validating transactions without even knowing it. This can’t be done with BTC, where setting up profitable mining operations is prohibitively expensive. 13. WASM-Based Virtual Machine: You can write smart contracts in your favorite language, compile them, and run them via the fastest VM in the world. 14. EGLD has been tested at an incredible 263,000 TPS using its sharding mechanism and low hardware requirements. Allegedly, by mid-next year (April), they’ll demonstrate 1,000,000 TPS. (For context: Mastercard handles around 5,000 TPS; BTC handles 5–7 TPS.) 15. EGLD is currently the most advanced L1 in terms of scalability, security, decentralization, UX, eco-friendliness, and tokenomics. It’s the only chain that has genuinely solved the Blockchain Trilemma and is ready to onboard 1 billion people into crypto—users who won’t even realize they’re interacting with crypto. 16. EGLD is perfectly positioned for AI projects—AI agents, AI tools, or a so-called “Truth Machine” that monitors other AIs on-chain, documenting what’s true and comparing different AI outputs (some of which may be censored or biased), ensuring people don’t get confused or scammed in an AI-driven world. 17. The EGLD team is the hardest-working team I’ve ever encountered. I had the honor of meeting many of them personally, and can attest that their pace—even during a bear market—is extraordinary. 18. EGLD’s development team is exceptionally active on GitHub, continually improving their network and actively committing code. 19. EGLD plans to introduce an update reducing block time to 600ms (down from ~6 seconds), which would make the chain essentially unrivaled. 20. EGLD is effectively the only usable L1 in Europe, and the team has direct connections within the EU government—extremely bullish for the project. 21. EGLD provides top-tier on-chain governance not only for the MultiversX (EGLD) protocol but also for DeFi projects (e.g., xExchange, MEX). 22. EGLD plans to expand to the US, likely opening offices in Austin, Texas. This could put them in direct contact with Elon Musk (if it hasn’t happened already), as he’s involved with If he’s done his research, he’d discover there’s simply no better L1 worldwide. 23. EGLD solved fully implemented sharding, perfect tokenomics, and top-tier architecture with just $5M, whereas other chains failed to do so even with $100M+. The second-best sharding network, NEAR, needed $100M, has worse tokenomics, and its sharding isn’t fully implemented yet. Its UX also doesn’t compare. Owning NEAR was like comparing a VW Golf R to a Porsche GT3—EGLD is the Porsche GT3. 24. According to Similarweb, EGLD has significantly high traffic relative to other chains with market caps 100x larger. The market cap vs. web traffic discrepancy is huge, which is a strong indicator of EGLD’s potential. 25. EGLD has the most active and dedicated community relative to its user base, with users who believe in the technology, have full faith in the team, and remain loyal despite price volatility—because they use the chain and know there’s nothing better. 26. Check other chains’ active user counts on X (Twitter) and compare it with the followers of EGLD’s founders and main network accounts, versus those with 30x, 50x, or 100x larger market caps. 27. Visit the MultiversX website to observe the futuristic design and presentation, then compare it to other chains that appear nearly a decade behind in design and branding. 28. EGLD hosts the xDay Global event, showcasing updates, new builders, projects in the ecosystem, and major announcements—similar to Apple’s Keynotes—delivered in a highly professional, goosebump-inducing atmosphere. The next event is in Korea, the second-biggest crypto market after the US. Check out their previous xDay after-movie to see why this is extremely bullish. 29. EGLD is moving forward with plans for the first regulated, audited EU stablecoin under MiCa regulation, made possible by acquiring xMoney, which I view as a “Stripe” for crypto/fiat, offering everything from user solutions to merchant services—potentially the future of payments. 30. Greg Siourouni recently joined EGLD, having been an executive director at SUI Foundation. He’s now co-founder of xMoney Global. xMoney (formerly UTrust, with token UTK) is owned and founded by the MultiversX Labs team. A stablecoin might be introduced soon, which would be massively bullish given xMoney’s roadmap. They recently announced integrations with Binance Pay—both ways. 31. EGLD prioritizes user safety, believing it’s the only feasible approach once the network scales to serve a billion people—many of whom are retail users with little to no security awareness. 32. EGLD offers “Sovereign Chains,” letting you effectively clone their chain without heavy development, set up your own validators, and leverage their unlimited scalability. Any blockchain (ETH, BTC, SOL) struggling with scalability, decentralization, or security could run an ultra-fast, scalable, and secure L2 on EGLD’s Sovereign Chain, meeting top enterprise requirements. No one else has really done this. The Sovereign Chain demo achieved astonishing TPS and has an SDK. 33. No downtime since inception. 34. No shard takeover attacks have occurred. 35. Extremely fast—soon 600ms block time will be in place. 36. ESDTs – The best token standard available: fungible, non-fungible, semi-fungible, DeFi assets—everything is native and highly customizable. 37. Top-tier composability of assets and smart contracts. 38. Integrated DNS at protocol level with HeroTags (nicknames) instead of long addresses. 39. Asynchronous calls are supported. 40. Cross-shard transfers, execution, reverts, and calls are seamlessly integrated. 41. The best staking system in the space. Secure Proof of Stake (SPoS) is far more efficient than Proof of Work (PoW). 42. Built-in Delegation and Staking Provider system, with over 125K delegators. 43. Complete support for liquid staked assets, fostering decentralization rather than centralization. 44. TransferRoles for ESDT and other advanced operations. 45. Composable tasks on-chain for more sophisticated DeFi workflows. 46. MultiTransfer and asset execution within one transaction. 47. Re-entrancy protection is built-in by design. 48. Storage for ESDT assets goes beyond a linear approach, optimizing performance. 49. No integer overflows thanks to integrated safeMath operations. 50. Integrated crypto opcodes in the VM, enhancing security and performance. 51. Support for BigFloats, BigInts, and BigDecimals, enabling advanced financial calculations on-chain. 52. No sandwich attacks, plus front-running and MEV protection. 53. Relayed Transactions, simplifying user interactions and fees. 54. Smart Accounts featuring data tries and multiple built-in functions. 55. Generalized Paymaster solutions, enabling flexible fee models. 56. Subscriptions for recurring or automated on-chain payments. 57. Web2-like usability with Web3 functionality, bridging mainstream adoption. 58. StakingV4 for improved decentralization. 59. Enhanced MEV protection rolling out to safeguard users. 60. Parallel execution is coming soon, boosting throughput. 61. 1 million TPS is on the roadmap, targeted for demonstration. 62. 600ms block time is also coming soon. 63. Reduced cross-shard processing is planned to improve efficiency. 64. ZK everywhere (PI²): “prove everything” approach is coming. 65. AsyncV3 is in development for more complex cross-contract interactions. 66. Scalability enhancements for Merkle Tries or a new data model are being explored. 67. Linear storage on the VM is forthcoming. 68. A dynamic language interpreter at the VM is also planned. 69. Rumors suggest that MultiversX (EGLD) is building a “Truth Machine” on their L1—an essential, game-changing tool for AI verification and societal impact. 70. The entire team features individuals with PhDs in mathematics and physics, and many are former engineers at Google, IBM, and similar companies. 71. Over 56% of the network’s supply is staked, showcasing strong community involvement. 72. More than 6,772,347 accounts have been created on the network. 73. A total of 476,627,710 transactions have been processed on-chain without any outages or hacks. 74. EGLD has built a massive ecosystem over time. While not as numerous in project count as Solana, its market cap is ~100x smaller, yet it has far superior tokenomics and technology. The projects that do exist, like Hatom Protocol, are top-tier in UX, security, and advanced features. Hatom will soon introduce USH, a truly high-quality, decentralized stablecoin. 75. On competing chains, automated transactions aren’t easily or cheaply executed, whereas on MultiversX, tools like let you do this for free (with near-zero fees). 76. No other chain combines such a strong team and long-term vision where every product meets extreme security and UX standards like MultiversX does. This is why I see it as the “next Apple” in Web3. 77. MultiversX has a new CMO – Adam Bates, a former CMO at the Cardano Foundation. He was behind the success of Cardano’s huge marketing campaign and has a very good relationship with Charles Hoskinson. Thanks to him, Beniamin Mincu (the founder of MultiversX) was likely introduced, and now they will probably discuss how both blockchains can help each other, as well as any other potential collaborations we don’t yet know about. This is also extremely bullish. #EGLD is undeniably the most Scalable, Advanced, Secure, and User-friendly L1 supercomputer ever created. It’s built to SHAPE THE FUTURE. 1) 2) 3) 4) 5) 27/6/2024 - EGLDSqueeze - SUMMARY: HERE IS NO 2ND BEST. EGLD IS ONLY ONE BLOCKCHAIN THAT CAN RULE THEM ALL. ✅ UNLIMITED SCALING ✅ SCARCE AS BTC ✅ PROGRAMMABLE AS ETH ✅ NO DOWNTIME AS SOL ✅ UI/UX OF Apple ✅ SHARDING DONE BEFORE NEAR & TON ✅ BEST WALLET xPortal WITH GUARDIAN Price prediction (NFA|DYOR): My reasoning is that the real market cap as of December 23, 2024...if we take into account the value of other cryptocurrencies such as BTC, SOL, ETH, AVAX, NEAR, TON, Cardano, BNB, XRP, and so forth, plus the existence of meme coins with valuations above 20 billion USD, or even games nobody plays anymore that still have valuations above 800 million shows that EGLD’s current market cap of approximately 942 million USD is incredibly low. From a technological standpoint, user experience, and other relevant aspects, compared to SOL, NEAR, TON, AVAX, and other L1 protocols, EGLD’s market cap should realistically be around 100 billion USD. Therefore, my prediction and investment thesis is a minimum of a 100x increase from its current price (+-SOL marketcap). MultiversX is ready to onboard 1 billion people to the blockchain. From a long-term perspective, it could even reach a market cap of 1 trillion USD, which is roughly half of where BTC is right now. That would be approximately a 1060x gain from the current market cap. 1 EGLD (MultiversX) is for $34 (only 31.4M max supply) think about this. Not financial advice. Again. There is no 2nd best L1. Position yourself where the puck is going, then wait at the goal until the goal gets there Apes together, strong. Ape alone, weak. We Don't Worry. We Just Win. Shape The Future

Daniel Veroc

50,650 görüntüleme • 1 yıl önce

$NWBO #𝗗𝗖𝗩𝗮𝘅-𝗟: 𝗧𝗵𝗲 𝗘𝘃𝗶𝗱𝗲𝗻𝗰𝗲 𝗶𝗻 𝗣𝗹𝗮𝗶𝗻 𝗧𝗲𝗿𝗺𝘀 A short, plain-language reading of the survival evidence for DCVax-L in #glioblastoma, and what it means under the MHRAgovuk guideline on external control arms. 📊 𝗣𝗔𝗥𝗧 𝗢𝗡𝗘: 𝗪𝗛𝗔𝗧 𝗧𝗛𝗘 𝗧𝗥𝗜𝗔𝗟 𝗙𝗢𝗨𝗡𝗗, 𝗔𝗡𝗗 𝗪𝗛𝗬 𝗜𝗧 𝗠𝗔𝗧𝗧𝗘𝗥𝗦 DCVax-L more than doubled five-year survival, and the benefit is durability: a subset gets lasting disease control and simply stays alive. 💉 𝗪𝗵𝗮𝘁 𝘁𝗵𝗲 𝗱𝗿𝘂𝗴 𝗶𝘀 DCVax-L is a personalized cancer vaccine for glioblastoma, the deadliest form of brain cancer and a designated orphan disease. It is a living drug: its active ingredient is the patient's own immune cells, primed with proteins from that patient's surgically removed tumor, so the immune system learns to attack the cancer. Unlike a chemical drug that is metabolized and cleared, it switches on a living immune response that keeps working long after the injection. Glioblastoma comes back in almost everyone: even with the full standard of surgery, radiation, and temozolomide chemotherapy, most patients live under two years, and only about one in twenty reaches five. DCVax-L is given on top of that standard care, not in place of it: every patient in the trial received surgery, radiation, and temozolomide, and the vaccine was added to it, so the comparison measures what the vaccine adds. For two decades, nearly every new drug tried in this disease has failed. That is the backdrop against which any positive result must be judged. 📈 𝗪𝗵𝗮𝘁 𝘁𝗵𝗲 𝘁𝗿𝗶𝗮𝗹 𝗳𝗼𝘂𝗻𝗱 In a Phase 3 trial of 331 patients, those who received DCVax-L lived longer. At the median the gain looks modest, about three months (19.3 versus 16.5). The number that matters sits at the far end of the survival curve: more than twice as many vaccine patients were alive at five years, 13.0% versus 5.7%, and a few reached ten years in a disease that usually kills within three. In patients whose tumor had already returned, the effect was larger still, cutting the risk of death by about 42%. 👥 𝗘𝘃𝗲𝗿𝘆 𝗴𝗿𝗼𝘂𝗽 𝗯𝗲𝗻𝗲𝗳𝗶𝘁𝗲𝗱, 𝗲𝘃𝗲𝗻 𝘁𝗵𝗲 𝗵𝗮𝗿𝗱𝗲𝘀𝘁 𝘁𝗼 𝘁𝗿𝗲𝗮𝘁 The benefit was not confined to the easy cases. Of the six prespecified subgroups the trial examined, every single one favored DCVax-L, and there was no group in which it did worse than standard care. The largest gains came in patients whose tumors carry MGMT methylation, who reached a median survival of 30.2 months from randomization against 21.3 for the controls. But even the hardest-to-treat patients, whose tumors lack that methylation, resist standard chemotherapy, and carry the worst prognosis in this disease, still came out ahead with the vaccine, at a hazard ratio of 0.93. A treatment that helps across the whole population, and helps most where the biology is most favorable, is acting like a real drug. 📉 𝗪𝗵𝘆 𝘁𝗵𝗲 𝗺𝗲𝗱𝗶𝗮𝗻 𝗵𝗶𝗱𝗲𝘀 𝘁𝗵𝗲 𝗿𝗲𝗮𝗹 𝘀𝘁𝗼𝗿𝘆 That five-year number is the whole story, and the median buries it. Almost every treatment that ever helped in glioblastoma did the same modest thing: it slid the survival curve a few months to the right, then let it fall back. Doubling five-year survival is different in kind. A three-month gain at the median cannot, by itself, double the fraction alive at five years. The only shape that produces both numbers is a split: most patients get the small delay the median measures, while a subset gets durable disease control that lasts for years, well past where glioblastoma should have ended them. That subset is the long tail of the curve, and it is where the benefit lives. A median ignores extremes, the way a town's median income tells you nothing about its millionaires. The real story is not a longer delay; it is that a meaningful share of patients simply stay alive. 🧬 𝗪𝗵𝘆 𝘁𝗵𝗲 𝗯𝗶𝗼𝗹𝗼𝗴𝘆 𝗽𝗿𝗲𝗱𝗶𝗰𝘁𝘀 𝘁𝗵𝗶𝘀 𝘀𝗵𝗮𝗽𝗲 The tail is not luck. It is what this biology is built to produce. The vaccine carries proteins from the patient's own tumor, so it aims the immune system at whatever that tumor is made of, not a fixed short list of targets. And it amplifies: one trained immune cell drives many others that multiply into cancer-killers. Andres Salazar, the neurologist who developed poly-ICLC into the clinical adjuvant given with the vaccine, puts it in a line: you start the fire, and you keep it burning. A response like that does not produce a one-time bump. It builds and widens over time. What matters is not just that a response forms, but what kind. Dendritic cells are the immune system's master switch for that, the cells that set what kind of attack the body mounts, and this vaccine is built from them. It drives what immunologists call a type 1 polarized response: an interferon-driven, cytotoxic program aimed squarely at the tumor. That direction comes from the vaccine itself; the poly-ICLC adjuvant given with it drives the same interferon program and sustains it. That is the active ingredient, and it has been measured. In UCLA studies of this approach, the patients whose immune systems mounted the strongest interferon response lived the longest. The response also spreads. In a different cancer, a vaccine carried on this same poly-ICLC adjuvant drove more than 70% of a patient's cancer-killing T cells to target proteins that were never in the vaccine: the immune system outgrew its original targets and went after the rest of the tumor on its own. This is called epitope spreading, and it is not particular to one tumor; it is what this kind of response does, and it is the kind of response DCVax-L builds. That breadth is the likeliest thing separating the long-term survivors from everyone else, a response that breaks past its targets and clears the disease rather than one that stays caged and stops. There is even a tell in who benefits most: the effect is largest in tumors whose biology builds up more mutations under chemotherapy, and more mutations mean more targets for a whole-tumor vaccine to find. The same logic explains what the vaccine is not, and what it does not need. Checkpoint inhibitors, the drugs that release the immune system's brakes, have failed on their own in glioblastoma because there was no active response to release; the vaccine supplies that response first. That makes the vaccine the natural foundation for combination therapy. Combining it with its poly-ICLC adjuvant, made by Oncovir, has already shown meaningful survival gains in a published analysis. In the pivotal trial, the vaccine's proven benefit came added on top of standard chemotherapy and radiation; the newer question is how much the immune response can carry on its own. A UCLA trial is now testing that in patients whose tumors have returned, adding #Keytruda (pembrolizumab), the checkpoint antibody from $MRK, in a regimen built entirely around the immune response with no chemotherapy or radiation in it at all. Its interim survival curve shows the shape the biology predicts. In the arm given the vaccine and Keytruda together after surgery, the curve does not fall away but flattens into a plateau, with roughly 65% of patients still alive well past the point where recurrent glioblastoma kills nearly everyone. That plateau is the signature of a response that took hold and lasted, the type 1 attack forming durable immune memory so that once the disease is controlled it stays controlled. That is where this points, and where it is already arriving, a treatment that works through the response itself, one that could in time lean less on the harsh radiation and chemotherapy that have defined glioblastoma care and barely moved its survival. These are interim results, from Prins, Cloughesy, and Liau at UCLA. 🔍 𝗣𝗔𝗥𝗧 𝗧𝗪𝗢: 𝗜𝗦 𝗜𝗧 𝗥𝗘𝗔𝗟? The trial was randomized, an independent experiment shows the outside comparison is trustworthy, and every separate check points the same way. 🤔 𝗧𝗵𝗲 𝗼𝗯𝗷𝗲𝗰𝘁𝗶𝗼𝗻, 𝗮𝗻𝗱 𝘄𝗵𝘆 𝗶𝘁 𝗺𝗶𝘀𝘀𝗲𝘀 𝘁𝗵𝗲 𝗺𝗮𝗿𝗸 So much for what happened; the harder question is whether to believe it. The trial has the feature critics attacked: it could not keep a normal placebo group, because patients assigned to placebo were allowed, by design and by medical ethics, to switch to the vaccine once their cancer returned, and almost all did. That erased the internal comparison, so survival was measured against closely matched patients from other completed trials, an approach called an external control, which critics argued could tilt toward the vaccine. What the objection misses is where the randomization went. This was a randomized, blinded trial. The patients who got the vaccine were assigned to it at random, not hand-picked, so the treated group is an ordinary slice of the trial population, not a favorable one. The crossover removed the placebo group but never touched how patients were assigned. That leaves exactly one place for bias to enter, the outside comparison group, which is precisely what the next checks test. ✅ 𝗧𝗵𝗲 𝗰𝗵𝗲𝗰𝗸 𝘁𝗵𝗮𝘁 𝗺𝗮𝗸𝗲𝘀 𝗶𝘁 𝘁𝗿𝘂𝘀𝘁𝘄𝗼𝗿𝘁𝗵𝘆 Before trusting a scale to weigh something unknown, you confirm it reads zero with nothing on it. That is what the calibration does, and it is the strongest part of the case. A separate, independent randomized trial called INSIGhT was run through the very same external-control method, and it gave two answers. First, three experimental drugs that had already failed were run through it, and it correctly found nothing (hazard ratios of 1.00, 0.93, and 0.88): the method does not manufacture a benefit where none exists. Second, INSIGhT's external controls were set head to head against its own randomized internal controls, and they were statistically indistinguishable. In this disease, an external control reproduces the answer a real randomized control would have given, and because that trial belonged to a different group, no one can say a sponsor graded its own work. The one place bias could enter, the control side, is the one place an independent randomized experiment certified as clean. There is a deeper fit worth naming, and it is what makes the two halves of this case one. The same instrument that reported those three failures as failures reads the vaccine as a success, and the biology says why: those drugs could not hold a tumor this varied, and the vaccine builds the broad, lasting response that finally does. One method, opposite readings, and one mechanism behind both. The statistics and the biology are not two arguments. They are the same argument seen twice. 🃏 𝗪𝗮𝘀 𝘁𝗵𝗲 𝗰𝗼𝗺𝗽𝗮𝗿𝗶𝘀𝗼𝗻 𝘀𝘁𝗮𝗰𝗸𝗲𝗱 𝗶𝗻 𝘁𝗵𝗲 𝘃𝗮𝗰𝗰𝗶𝗻𝗲'𝘀 𝗳𝗮𝘃𝗼𝗿? A natural worry is that the outside comparison was arranged after the fact to flatter the vaccine. The trial was built to prevent that. The patients to compare against, and the rules for matching them, were fixed in writing before anyone saw results, and an independent firm, not the company, chose the comparison trials against those rules. The trial also switched its main measure partway through, from delaying tumor growth to overall survival, but that was not a maneuver: immune treatments cause a harmless swelling that mimics tumor growth on scans and made the growth measure unreliable, and the switch was made while everyone was still blinded. Three further checks point the same way. Survival in this disease has not improved over the years the comparison spans, so same-era controls are sound. When the borrowed controls were tested directly, the comparison came out conservative rather than flattering. And the controls were counted from the same point in the disease as the vaccine patients, so neither side got a head start. Where the comparison can err, it errs against the drug. 🔬 𝗧𝗵𝗲 𝗺𝗼𝗿𝗲 𝗰𝗮𝗿𝗲𝗳𝘂𝗹 𝗮𝗻𝗮𝗹𝘆𝘀𝗶𝘀 𝗺𝗮𝗱𝗲 𝘁𝗵𝗲 𝗯𝗲𝗻𝗲𝗳𝗶𝘁 𝗯𝗶𝗴𝗴𝗲𝗿, 𝗻𝗼𝘁 𝘀𝗺𝗮𝗹𝗹𝗲𝗿 The first comparison used whole groups. A sharper one became possible once patient-level records from three other trials could be obtained, pairing each vaccine patient with controls matched on the traits that drive survival in glioblastoma, above all MGMT methylation status, matched exactly, plus age, sex, extent of surgery, residual disease, and performance status. When the comparison got sharper, the benefit grew in every one of these analyses. The original cohort-level estimate was 2.8 months. Patient-level matching across the three trials put the gain between 3.4 and 6.3 months, and a method that weights patients rather than pairing them put it between 3.4 and 4.3, with several of the matched comparisons roughly doubling the original figure. The hazard ratio moved the same way, from 0.80 to between 0.69 and 0.77. The direction matters. A real effect blurred by crude matching gets clearer when the matching improves, while a biased one tends to shrink. It got stronger. 🕵️ 𝗛𝗼𝘄 𝗺𝘂𝗰𝗵 𝗵𝗶𝗱𝗱𝗲𝗻 𝗯𝗶𝗮𝘀 𝘄𝗼𝘂𝗹𝗱 𝗶𝘁 𝘁𝗮𝗸𝗲 𝘁𝗼 𝗲𝘅𝗽𝗹𝗮𝗶𝗻 𝘁𝗵𝗶𝘀 𝗮𝘄𝗮𝘆 A fair question is how much hidden bias it would take to erase the result. Statisticians measure that with the E-value, and here a hidden factor would have to be about as strong as age is on survival, would also have to drive who received the vaccine, and would have to have escaped the decades of research that mapped every known risk factor in this disease. A second, independent check, Rosenbaum's Gamma, comes at it from the other side, asking how large an unseen imbalance between matched patients it would take to break the result, and it reaches the same verdict. Every factor strong enough to matter was already matched. A hidden one that clears that bar is not plausible. 🔒 𝗪𝗵𝘆 𝘁𝗵𝗲 𝗿𝗲𝘀𝘂𝗹𝘁 𝗶𝘀 𝗵𝗮𝗿𝗱 𝘁𝗼 𝗳𝗮𝗸𝗲 The strongest point is not any single result. It is that a hidden bias big enough to explain the effect away would have to produce the same answer in every independent test at once: • The independent randomized calibration trial. • Three separate comparison trials, drawn from different studies. • Two different statistical methods that handle the data in different ways. • A separate pooled analysis of other dendritic-cell vaccine trials. • The internal math of the survival curve, which points to the same result (about 0.71 at five years) that the patient matching found. And it would have to fall in the exact direction the biology predicted before any data existed: a slow, widening benefit concentrated in long-term survivors. A single hidden factor that could forge all of that at once is not a hidden factor. It is a coincidence that does not happen. 🩺 𝗣𝗔𝗥𝗧 𝗧𝗛𝗥𝗘𝗘: 𝗪𝗛𝗔𝗧 𝗜𝗧 𝗠𝗘𝗔𝗡𝗦 The drug is nearly free of harm, the evidence fits an established and approved regulatory path, and what remains unrun changes nothing about the case. 🛡️ 𝗧𝗵𝗲 𝘀𝗮𝗳𝗲𝘁𝘆 Two things decide whether a real effect reaches patients: whether the drug is safe enough to use, and whether regulators will accept the evidence. The first is settled. Across 2,151 doses, only five serious side effects were even possibly related to the vaccine, with no autoimmunity and no cytokine storm. It is made once, in about eight days, then stored and given as a simple injection. When a treatment barely harms, the benefit needed to justify it falls, and the benefit here clears that lower bar easily. ⏳ 𝗪𝗵𝘆 𝘁𝗵𝗲 𝘀𝘁𝗿𝗼𝗻𝗴𝗲𝗿 𝗱𝗮𝘁𝗮 𝗮𝗽𝗽𝗲𝗮𝗿𝗲𝗱 𝗼𝗻𝗹𝘆 𝗻𝗼𝘄 A reasonable person asks why the sharper analysis appeared in 2026 and not in 2023. The answer is access, not choice. The patient-level analysis was written into the trial's plan from the start, to run if and when the data could be obtained. The company tried and could not get it in 2023, because the trials that held it had not released it; it became available later through a data-sharing repository, on the data owners' timeline, not the company's. The stronger analysis was always the plan. It was waiting on data that other parties control. 🔄 𝗣𝗮𝘁𝗶𝗲𝗻𝘁𝘀 𝘄𝗵𝗼𝘀𝗲 𝗰𝗮𝗻𝗰𝗲𝗿 𝗿𝗲𝘁𝘂𝗿𝗻𝗲𝗱 𝗮𝗹𝘀𝗼 𝗯𝗲𝗻𝗲𝗳𝗶𝘁𝗲𝗱 The vaccine helped not only newly diagnosed patients but also those whose tumor had already come back, and there the effect was the largest seen anywhere in the trial. In that group, median survival ran 13.2 months from recurrence against 7.8 for the controls. The separation opened immediately, with 90.6% of vaccine patients alive at six months against 64.0% of controls, and the lead held to the later marks, where survival more than doubled: 20.7% against 9.6% at two years, 11.1% against 5.1% at two and a half. These are the original cohort-level results, already published. The high-resolution patient-level matching that was applied to the newly diagnosed group has not yet been done here, for a practical reason: it needs patient records from other recurrent-cancer trials, held by a European research organization that shares them through its own formal request. Obtaining them is a routine next step, not an obstacle, and the newly diagnosed experience suggests the sharper analysis would only make the recurrent result stronger. The approval case rests on the newly diagnosed evidence, so nothing important depends on this step; it would simply sharpen a result that is already the strongest in the trial. 🏛️ 𝗪𝗵𝗮𝘁 𝘁𝗵𝗶𝘀 𝗺𝗲𝗮𝗻𝘀 𝗳𝗼𝗿 𝗮𝗽𝗽𝗿𝗼𝘃𝗮𝗹 Regulators do not treat an external control as a first choice, but the MHRA's guideline allows it in exactly this situation: a severe disease where a placebo trial is not ethical or feasible, and an effect large enough to interpret despite the design. The guideline even gives its own worked example of an acceptable external control, and it reads almost like a description of this trial: a rare disease, no ethical placebo, same-era standard-of-care controls, an objective survival endpoint, and an effect too large to blame on bias. DCVax-L fits on every count, and it was the first medicine ever to receive the MHRA's Promising Innovative Medicine designation, which asks essentially the same questions. The newly diagnosed case carries the decision on its own evidence, and regulators weigh that evidence against the disease it treats: in a cancer this lethal, with nothing better on offer, the question is whether the benefit is large, clear, and consistent enough to act on, and a benefit of this size, pointing the same way from every direction, is. 📜 𝗧𝗵𝗶𝘀 𝗵𝗮𝘀 𝗯𝗲𝗲𝗻 𝗱𝗼𝗻𝗲 𝗯𝗲𝗳𝗼𝗿𝗲, 𝗮𝗻𝗱 𝗮𝗽𝗽𝗿𝗼𝘃𝗲𝗱 External controls are not a novelty invented for this drug. Over the past two decades they have factored into roughly forty-five drug approvals by the United States regulator, each granted under the conditions that apply here: a serious or rare disease, a placebo that would be unethical, and high unmet need. Regulators do not grant this lightly. They grant it when the disease is serious, its course is predictable and objectively measured, and the effect is large. Glioblastoma meets all three. The named cases cover every part of this disease's profile. Defibrotide, for a life-threatening transplant complication, was approved on a propensity-score comparison to a historical control, the same kind of method used here. Blinatumomab, for an aggressive relapsed leukemia, is the precedent for a fast-killing cancer, cleared by both the United States and European regulators. Cerliponase alfa is the precedent for a fatal brain disease, cleared by both agencies on treated patients versus a matched natural-history group. Glioblastoma is both at once, a fast-killing cancer of the brain, read by the same method, so no part of its profile lacks a close approved precedent. And on the one axis that governs how far an external-control result can be trusted, DCVax-L goes beyond all three. Each of those drugs was tested in a single-arm trial, with no randomization at all. DCVax-L began as a randomized trial and became an external-control comparison only when ethics consumed its placebo group. It sits in that tradition, and at the top of it. 🎯 𝗧𝗵𝗲 𝗯𝗼𝘁𝘁𝗼𝗺 𝗹𝗶𝗻𝗲 This began as a randomized trial. Medical ethics forced it into an external-control comparison. Every independent way of checking it, on different data and different math, points the same direction, and the biology predicted that direction in advance. The first analysis did not overstate the vaccine's effect. Read with the right tools, it understated it.

Andrew Caravello, DO

11,406 görüntüleme • 2 ay önce

To My Lovely community, #Crofam . I want to put this in advance, I am still here because I still have hope, but I cannot serve our future if I don't talk about this topic below, which is mainly addressed to our leaders. Most of you are well aware of the issues. Don't listen to us, listen to independent reviewers who are not biased left or right. Perfect snippet from an Overall positive video below! But if you decide that you might be interested from an OLD community member who has been with your for years and spent hundreds - but more like Thousands of hours to contribute to your success please read my take below: 👇 0. The Loudest Issue: CRO Performance & Cronos Activity The most vocal concern among investors is the poor performance of the CRO token, released by Crypto.com. However, an even bigger problem is the declining activity on —again, a creation of Cryptocom. 1. A Confusing and Inconsistent Relationship with Cronos For years, we felt abandoned, and while we understood that this might have been necessary due to regulatory concerns, it was a breath of fresh air when Cronos was mentioned in Cryptocom’s roadmap. Only for it to be removed again. Then we saw Kris listed as the head of Cronos on official Korean documents—only for it to change again within days. CRO gets advertised, then removed. Many people got involved with $CRO and Cronos specifically only because it was tied to and closely linked to Cryptocom and the fact that the relationship has only blurred and become more ambiguous over time is also not really fair to users/investors either. In fact, it's almost misleading. The relationship between Cryptocom and Cronos is so inconsistent and convoluted that it could be the plot of a Brazilian soap opera. Investors deserve better clarity and stability. 2. The Ambassador Program – What’s the Point? I know for most this is not important, but for me as an ambassador still somewhere on the top. What is the actual purpose of the Cryptocom and Cronos ambassador programs? Communication is non-existent, engagement is minimal, and we have no real role in the chain’s development. Instead, we serve as punching bags for community frustrations because no official team members are available for direct questions. If the ambassador program disappeared tomorrow, nothing would change. That says a lot. I don't even know if half of the ambassadors we have still on the chain or active at all. We could cut the current ambassadors list by 90% and they probably won't even notice. As a salt on the wound again, the gossips that ambassadors who the community never even heard of getting reimbursed while those who here and working full time for free just to keep your chain alive is outrageous. I only know one Turkish ambassador who works hard for the chain and community and that is Kaan, the rest is basically unknown and only seen complaints about them, not just wasting company money but in return swearing and complaining about Kris. How is that a good investment? I don't care how close some of them to certain top level officials, it needs a review even if its a small portion of the problems.. Start rewarding people who serve the overall success. 3. Where is the CEO? Leadership means stepping up, especially in difficult times. Yet, Kris has never taken open, unscripted questions from the community. Even the scripted AMA sessions with Steve were appreciated, but now those are gone too. Hiding during tough times and reappearing only when the market is bullish is not the leadership investors deserve. If you don’t want us when things are bad, don’t celebrate with us when things are good. *Moving head sideways and shaking finger 4. Cronos Labs & Grant Allocations I have no issue with Cryptocom building under the guise of Cronos Labs and receiving grants originally meant for external developers. But while it’s true that many builders abused the grant system and left, that is the fault of those who approved those grants—not the current builders still here. It takes an hour of research to see who’s been building for years. If you genuinely care about Cronos, start engaging with the community. I post free aggregated Cronos news every Sunday—it’s not that hard to stay informed! ad: Pampa Sunday Cronos news turns 1 year this month! Whoop-whoop! 5. Cronos Token Listings & Double Standards I understand that not all Cronos-based tokens can be listed, or even any at all. However, the claim that Cryptocom is protecting retail investors by avoiding Cronos token listings is simply false. Look at the last 100 token listings—many have crashed or were outright rug. Examples? Gekko HQ is down over 99%, and $HEHE is completely abandoned. Where is the protection in that? Meanwhile, Cryptocom lists its own Cronos Labs projects despite them having little to no volume, just to keep them afloat. If you don’t want to list Cronos tokens, just be honest—admit that you prefer listing projects from other chains because they might attract new users. Don’t act like you’re protecting anyone. 6. The Marketing Budget vs. Reality Cryptocom spends an insane amount on marketing. While I don’t mind the ambition, we need to be realistic. Celebrity endorsements and massive sponsorship deals are questionable at best. Did spending 10 million CRO on a golfer bring in new users? Did getting Eminem to shill investments work? The recent U.S. elections showed that celebrities hold far less influence than people think. This is a new generation, a new market, and you cannot market a crypto company like Marlboro did Formula 1 in the ‘70s plastering your logo all over the tracks. 7. Respect for Developers & Employees To be clear: I have immense respect for the developers and employees at Cryptocom and Cronos. The devs are not only incredibly skilled but also hardworking and kind. I have great relationships with several, and I know many would love to be closer to the community. The problem isn’t them, and I urge the community not to direct frustration at them or the community managers. The real issues come from the top. 8. If You’re Still Reading, You Care—Unlike Leadership If you made it this far, you care about Cryptocom and Cronos. Unfortunately, I’d bet my Al Capone LEGO set that no one from upper management will read this far. Instead, they’ll dismiss it as just another clueless community member who failed to diversify their portfolio and is now whining about $CRO instead of just buying Bitcoin. Solutions: 1. Honest and Transparent Communication - Tell us what’s working, what’s not, and what your realistic expectations are. No more smoke and mirrors. 2. Active Participation in the Cronos and Cryptocom Ecosystem - Stop relying on secondhand messages. We’re tired of hearing “We’ll forward your suggestion to the team” with no follow-up. That’s not engagement and we heard that a thousand times already. 3. Genuine Support for Cronos Builders - Instead of blindly funding in-house projects, support real builders who have been active for years. Not every project needs massive funding—simple recognition and low-cost support or even free advertisement can go a long way. 4. Better Token Listing Policies - If Cronos tokens aren’t worth listing, be transparent. But if you’re listing failed and abandoned projects from other chains, you owe us an explanation. 5. Accountability from Leadership - We don’t expect miracles, but we do expect our CEO to engage with the community, take tough questions, and stand by us during difficult times. 6. Realistic Marketing Strategies - Stop throwing money at celebrity endorsements and ineffective ad campaigns. Focus on organic growth, community engagement, and real use cases. 7. Regular AMAs and Open Discussions - Leadership should hold at least quarterly AMAs with the community, answering real, unscripted questions. 8. More Accessible Community Engagement - Create a direct communication channel between developers, leadership, and the community, where actual feedback can be given and received. 9. Revamping the Ambassador Program - Give ambassadors a real purpose, better communication tools, and a direct line to Cryptocom to discuss community concerns.

Pampa

31,544 görüntüleme • 1 yıl önce

$MU $SNDK $LITE $VRT NVIDIA and Groq: 2nd and 3rd Order Strategic Infrastructure Effects and Market Implications Public reporting indicates NVIDIA has agreed to acquire Groq for approximately $20,000,000,000 in cash, while excluding Groq’s nascent cloud business from the transaction perimeter. The reported carve-out materially constrains the immediate, direct linkage from the acquisition to incremental, NVIDIA-controlled data center capacity build-out because GroqCloud appears to be the principal channel through which Groq hardware is currently monetized at scale as a service. The infrastructure-market implications therefore depend primarily on post-close product strategy: whether NVIDIA (1) commercializes Groq silicon as a distinct inference product line and drives broad deployment through OEM/ODM channels and partners, (2) uses the acquisition mainly to absorb IP and talent while de-emphasizing standalone Groq hardware volumes, or (3) uses Groq technology to reshape NVIDIA’s own inference systems and networking roadmaps. The dominant transmission mechanism into memory, networking, and facility infrastructure markets is the degree to which NVIDIA shifts incremental inference deployments away from GPU architectures that are tightly coupled to external high-bandwidth memory (HBM) and toward Groq’s current architecture, which emphasizes large on-chip SRAM, deterministic compiler-scheduled execution, and direct chip-to-chip connectivity. Independent and company-published materials describe Groq’s current-generation approach as having no external memory, keeping weights and KV cache on-chip during processing, and requiring model sharding across multiple chips due to limited on-chip SRAM per device. That architectural choice is directionally HBM-negative on a per-accelerator basis and ambiguous for DRAM, NAND, networking, power, and cooling on a per-token basis because the design can reduce memory wall losses and tail-latency overhead while potentially increasing the number of chips and interconnect endpoints required to serve large models and long-context workloads. HBM implications are the most mechanically straightforward but should be framed as second-derivative rather than absolute. If Groq-class inference silicon meaningfully displaces NVIDIA GPU-based inference deployments, incremental HBM bit demand tied to inference growth could be reduced relative to a GPU-only baseline because Groq’s current approach does not appear to attach HBM stacks to each accelerator. However, current market structure suggests HBM remains supply-constrained and is being pulled by multiple vectors including continued GPU training scale and high-capacity inference configurations, with leading suppliers signaling tight conditions extending beyond 2026. In that environment, reduced inference-driven HBM intensity could primarily reallocate scarce HBM supply toward higher-end training and premium inference GPUs rather than creating an outright volume collapse, preserving high utilization of HBM capacity while potentially affecting the slope of pricing power and capacity expansion urgency over a multi-year horizon. The key downside scenario for the HBM complex would be a durable architectural bifurcation where “good-enough” inference shifts disproportionately to HBM-less ASICs across a broad swath of deployments (latency-sensitive, batch-1, cost-per-token optimized), while training remains GPU-HBM dominated; such a split would reduce the portion of future inference compute that naturally monetizes through HBM content and could compress the incremental HBM-per-AI-dollar ratio. The key upside/neutral scenario for HBM is that the supply chain remains fully allocated regardless, with NVIDIA using any “freed” HBM to ship more high-end GPUs into training and long-context inference, especially as roadmaps increase HBM per GPU, sustaining robust aggregate bit demand even if inference becomes more heterogeneous. Conventional DRAM implications split into 2 channels: (1) DRAM wafer capacity diversion into HBM and (2) DDR content per server in AI clusters. Supplier commentary indicates that AI-driven memory demand is supporting elevated DRAM markets more broadly, and HBM production is resource-intensive versus conventional DRAM, tightening supply for DDR products in parallel. A meaningful NVIDIA pivot to an inference architecture that reduces HBM dependence could, at the margin, ease the most acute HBM-driven bottlenecks and allow memory manufacturers more flexibility in balancing DRAM mix, which could be modestly DDR-positive on the supply side (less crowding-out) even if it is DDR-neutral or slightly negative on the demand side (if per-node CPU/DDR requirements decline due to more efficient accelerator utilization). The dominant practical outcome is likely that DDR demand remains supported by broad AI server proliferation and increasing memory footprints at the system level (CPUs, networking stacks, caching layers, retrieval-augmented pipelines), while HBM remains the premium profit pool; therefore, any HBM displacement that increases total server volumes could indirectly keep DDR demand resilient even if DDR per accelerator is not rising materially. NAND flash implications are comparatively indirect and volume-driven rather than architecture-driven. Inference clusters require SSD capacity for model storage, container images, logging, and increasingly for fast local retrieval indices and embedding stores, but the storage footprint per unit of compute is typically smaller than in training pipelines that stage large datasets and checkpoints. If NVIDIA uses Groq to lower inference cost and latency enough to expand the total number of inference deployment locations (regional colocation, enterprise on-prem, sovereign footprints), aggregate SSD attach could rise through geographic fragmentation and replication of model artifacts across more sites, even if per-site storage is modest. The NAND effect is therefore likely to be demand-broadening and mix-positive (datacenter SSDs) but not a primary swing factor versus the macro AI capex cycle and consumer/device cycles. Hard disk drive (HDD) markets should see negligible direct sensitivity because nearline HDD demand is driven by bulk storage and cloud archiving economics, while inference acceleration choices primarily reshape compute and network layers; any HDD benefit would be a tertiary function of overall data center square footage expansion rather than a direct consequence of Groq silicon displacing GPUs. Optical networking implications require separating (1) intra-cluster back-end fabrics that connect accelerators and (2) front-end / data center interconnect (DCI) that connects sites and regions. Groq’s own positioning and third-party reporting suggest scaling beyond a single node or rack relies on high-bandwidth fabrics and, in some described configurations, optical interconnect scaling across hundreds of chips. If NVIDIA commercializes Groq at scale, 2 offsetting forces emerge: lower cost-per-token and improved latency could expand inference throughput and drive more east-west traffic, increasing demand for high-speed switching and optics; conversely, if Groq delivers materially higher utilization and tokens per unit of network bandwidth for certain workloads, the network required per served token could decline. Public NVIDIA materials already indicate an aggressive photonics roadmap aimed at scaling AI factories, including co-packaged optics (CPO) switches and explicit collaboration with Coherent and Lumentum in the silicon photonics supply chain. That linkage is important because it suggests that, independent of Groq, NVIDIA is already pushing optics integration deeper into the switch package to reduce power and increase resiliency; Groq increases the strategic incentive to reduce network power and latency if inference becomes even more distributed and latency-sensitive. For Lumentum and Coherent specifically, the net implication is less about “more optics versus fewer optics” and more about a shift in optics form factor and value capture. Co-packaged optics can reduce reliance on pluggable transceivers in some switch architectures while increasing demand for integrated photonic engines, lasers, fiber attach, packaging processes, and component-level supply. NVIDIA’s own announcements explicitly position Coherent and Lumentum as collaborators in creating the integrated silicon/optics process and supply chain for photonics switches. If Groq accelerates the transition to very large-scale fabrics (more endpoints, higher port speeds, tighter power envelopes), that tends to pull forward CPO adoption and amplifies demand for the underlying photonics components even if the conventional pluggable module TAM is structurally pressured over time. If Groq instead pushes inference toward smaller, more localized pods (closer to users, more regional colocation), that can be optics-positive for DCI and metro connectivity because more sites must be interconnected at high bandwidth with low latency, favoring coherent optics and high-speed interconnect between facilities. The principal risk for optics suppliers is timing and margin structure: a faster move to NVIDIA-driven integrated photonics could concentrate bargaining power and compress margins for commoditized transceiver modules while favoring suppliers with differentiated lasers, integration capability, and qualification depth in NVIDIA’s CPO ecosystem. AEC and copper interconnect implications hinge on whether Groq deployment increases the density of short-reach links inside racks and rows. High-speed copper remains structurally advantaged at very short distances on cost, power, and serviceability, but reaches become constrained as lane speeds and aggregate bandwidth rise, creating a role for active electrical cables (AECs), retimers, and signal-conditioning silicon. Credo explicitly positions its AEC products as enabling reliable lossless 800G connectivity for AI clusters, and the company has highlighted participation at NVIDIA GTC with content focused on extending PCIe/CXL using AECs, indicating relevance to next-generation system topologies that require longer reach and higher signal integrity than passive copper can deliver. If NVIDIA turns Groq into a widely deployed inference card or chassis product, the likely near-term effect is AEC-positive because (1) more inference throughput tends to increase top-of-rack connectivity requirements, (2) distributing inference across more racks and sites increases short-reach links per unit of delivered service, and (3) PCIe-attached accelerator architectures tend to require robust signal conditioning as systems move to PCIe 6.x and beyond. Groq workshop materials explicitly reference GroqCard and GroqNode form factors, reinforcing that PCIe-attached deployment has been central to Groq’s current packaging strategy. The main countervailing risk is that Groq’s deterministic chip-to-chip fabric could be implemented primarily through backplanes and direct board-level connectivity that reduces the need for merchant AECs inside the box; in that case, incremental AEC demand would concentrate more in rack-to-switch and node-to-fabric links rather than within-chassis chip fabrics. Astera Labs implications are connectivity-architecture sensitive and, on balance, skew positive if NVIDIA increases heterogeneity and disaggregation in AI systems. NVIDIA has publicly positioned NVLink Fusion as a pathway for partners to build semi-custom AI infrastructure and has explicitly identified Astera Labs as a partner in that ecosystem, with Astera describing NVLink-related solutions expanding its connectivity platform across PCIe, CXL, and Ethernet plus fleet observability software. A Groq acquisition increases the probability that NVIDIA offers a broader menu of accelerators (training GPUs, inference-focused ASICs) and therefore increases the importance of scalable, high-reliability connectivity, retiming, switching, and telemetry across mixed topologies. If Groq silicon remains PCIe-attached in many deployments, PCIe 6.x retimers/switches and active cable modules become more central, aligning with Astera’s core portfolio. If NVIDIA instead integrates Groq concepts into scale-up fabrics (NVLink-like domains) or uses Groq to expand into inference “appliances” that must be rapidly deployed in colocation environments, the need for standard-compliant, serviceable connectivity with strong RAS/telemetry increases, again aligning with Astera’s positioning. Power equipment and cooling implications for Vertiv and adjacent suppliers should be viewed through the lens of rack power density, cooling modality (air vs liquid), and site deployment model (hyperscale campuses vs distributed colocation/enterprise). Groq claims its LPU and rack designs are “air-cooled by design” and require no complex cooling and power infrastructure, and third-party reporting has described Groq’s approach as relying on parallelism across many lower-power units rather than extreme per-chip performance. If NVIDIA scales Groq as a mainstream inference platform, the mix of data center cooling spend could shift modestly away from the highest-density liquid-cooled racks toward more air-cooled or hybrid deployments, particularly for inference pods placed in existing facilities that cannot easily retrofit for very high rack heat flux. That would be a mix headwind for suppliers most levered exclusively to high-end liquid cooling attachments per rack, but it is not necessarily a volume headwind for Vertiv given the company’s broad exposure to both power and cooling infrastructure and the likelihood that total AI deployment locations expand. Vertiv’s own industry commentary emphasizes that AI racks require higher power-density UPS, batteries, power distribution equipment, and switchgear capable of handling rapid load transients, and that hybrid cooling systems will evolve across deployment environments. Those statements align with a world where inference growth increases the count of powered racks and raises the operational complexity of power delivery even if per-rack density is lower than the most extreme training clusters. The most material infrastructure impact may occur outside the rack and upstream of the data hall: grid interconnects, substations, transformers, switchgear, generators, and utility-scale generation additions. Recent regulatory actions in the U.S. highlight that projected data center demand is already driving large planned increases in electricity generation capacity, underscoring that power availability is a binding constraint. In that context, an inference architecture that lowers joules per token could reduce the power required per unit of inference delivered, but it can also accelerate demand by lowering cost and improving latency, increasing the total volume of inference served (a classic rebound effect). The net outcome is likely continued, elevated demand for power infrastructure even if efficiency improves, with the key swing factor being whether AI capex remains on a multi-year growth trajectory or enters a digestion phase. Other data center infrastructure implications include server/ODM mix, facility design standardization, and networking architecture choices. If NVIDIA positions Groq-based inference as a broadly distributable “standard server + accelerator” solution rather than as an integrated, liquid-cooled rack like GB200 NVL72, spend could shift toward more conventional air-cooled server designs, higher unit volumes of mainstream racks, and faster deployment in colocation footprints, increasing demand for modular power rooms, busways, and rapidly deployable cooling solutions. If NVIDIA instead integrates Groq into its “AI factory” paradigm, the primary effect is likely acceleration of dense back-end fabric build-outs and a faster push toward photonics switching, increasing demand for fiber plant, connectors, and integrated optics supply chains while potentially compressing the lifecycle of transitional architectures based on pluggable optics and mid-reach copper. NVIDIA’s stated roadmap toward co-packaged optics and silicon photonics switches is already oriented toward scaling to very large GPU counts; adding a high-end inference ASIC increases the strategic importance of power-efficient, low-latency fabrics because inference economics become increasingly sensitive to network overhead as compute cost declines. Across the covered segments, the most defensible base case is limited near-term dislocation and a medium-term increase in uncertainty around memory intensity per unit of inference growth. HBM faces the clearest relative risk from an HBM-less inference platform, but supply tightness and GPU training roadmaps reduce the probability of an absolute demand shock over the next 12–24 months. Optical, AEC/copper, and power/cooling are more likely to remain volume-supported because they scale with endpoint count, deployment fragmentation, and total data center footprint, and those tend to rise when inference becomes cheaper and more widely deployed. The highest-conviction second-order effect is a shift in infrastructure mix: incrementally more distributed inference deployments (favoring colocation power/cooling standardization, DCI optics, and serviceable short-reach interconnect) and a gradual migration from pluggable optics toward integrated photonics in back-end fabrics (favoring suppliers positioned in the CPO ecosystem).

$MU $SNDK $LITE $VRT NVIDIA and Groq: 2nd and 3rd Order Strategic Infrastructure Effects and Market Implications Public reporting indicates NVIDIA has agreed to acquire Groq for approximately $20,000,000,000 in cash, while excluding Groq’s nascent cloud business from the transaction perimeter. The reported carve-out materially constrains the immediate, direct linkage from the acquisition to incremental, NVIDIA-controlled data center capacity build-out because GroqCloud appears to be the principal channel through which Groq hardware is currently monetized at scale as a service. The infrastructure-market implications therefore depend primarily on post-close product strategy: whether NVIDIA (1) commercializes Groq silicon as a distinct inference product line and drives broad deployment through OEM/ODM channels and partners, (2) uses the acquisition mainly to absorb IP and talent while de-emphasizing standalone Groq hardware volumes, or (3) uses Groq technology to reshape NVIDIA’s own inference systems and networking roadmaps. The dominant transmission mechanism into memory, networking, and facility infrastructure markets is the degree to which NVIDIA shifts incremental inference deployments away from GPU architectures that are tightly coupled to external high-bandwidth memory (HBM) and toward Groq’s current architecture, which emphasizes large on-chip SRAM, deterministic compiler-scheduled execution, and direct chip-to-chip connectivity. Independent and company-published materials describe Groq’s current-generation approach as having no external memory, keeping weights and KV cache on-chip during processing, and requiring model sharding across multiple chips due to limited on-chip SRAM per device. That architectural choice is directionally HBM-negative on a per-accelerator basis and ambiguous for DRAM, NAND, networking, power, and cooling on a per-token basis because the design can reduce memory wall losses and tail-latency overhead while potentially increasing the number of chips and interconnect endpoints required to serve large models and long-context workloads. HBM implications are the most mechanically straightforward but should be framed as second-derivative rather than absolute. If Groq-class inference silicon meaningfully displaces NVIDIA GPU-based inference deployments, incremental HBM bit demand tied to inference growth could be reduced relative to a GPU-only baseline because Groq’s current approach does not appear to attach HBM stacks to each accelerator. However, current market structure suggests HBM remains supply-constrained and is being pulled by multiple vectors including continued GPU training scale and high-capacity inference configurations, with leading suppliers signaling tight conditions extending beyond 2026. In that environment, reduced inference-driven HBM intensity could primarily reallocate scarce HBM supply toward higher-end training and premium inference GPUs rather than creating an outright volume collapse, preserving high utilization of HBM capacity while potentially affecting the slope of pricing power and capacity expansion urgency over a multi-year horizon. The key downside scenario for the HBM complex would be a durable architectural bifurcation where “good-enough” inference shifts disproportionately to HBM-less ASICs across a broad swath of deployments (latency-sensitive, batch-1, cost-per-token optimized), while training remains GPU-HBM dominated; such a split would reduce the portion of future inference compute that naturally monetizes through HBM content and could compress the incremental HBM-per-AI-dollar ratio. The key upside/neutral scenario for HBM is that the supply chain remains fully allocated regardless, with NVIDIA using any “freed” HBM to ship more high-end GPUs into training and long-context inference, especially as roadmaps increase HBM per GPU, sustaining robust aggregate bit demand even if inference becomes more heterogeneous. Conventional DRAM implications split into 2 channels: (1) DRAM wafer capacity diversion into HBM and (2) DDR content per server in AI clusters. Supplier commentary indicates that AI-driven memory demand is supporting elevated DRAM markets more broadly, and HBM production is resource-intensive versus conventional DRAM, tightening supply for DDR products in parallel. A meaningful NVIDIA pivot to an inference architecture that reduces HBM dependence could, at the margin, ease the most acute HBM-driven bottlenecks and allow memory manufacturers more flexibility in balancing DRAM mix, which could be modestly DDR-positive on the supply side (less crowding-out) even if it is DDR-neutral or slightly negative on the demand side (if per-node CPU/DDR requirements decline due to more efficient accelerator utilization). The dominant practical outcome is likely that DDR demand remains supported by broad AI server proliferation and increasing memory footprints at the system level (CPUs, networking stacks, caching layers, retrieval-augmented pipelines), while HBM remains the premium profit pool; therefore, any HBM displacement that increases total server volumes could indirectly keep DDR demand resilient even if DDR per accelerator is not rising materially. NAND flash implications are comparatively indirect and volume-driven rather than architecture-driven. Inference clusters require SSD capacity for model storage, container images, logging, and increasingly for fast local retrieval indices and embedding stores, but the storage footprint per unit of compute is typically smaller than in training pipelines that stage large datasets and checkpoints. If NVIDIA uses Groq to lower inference cost and latency enough to expand the total number of inference deployment locations (regional colocation, enterprise on-prem, sovereign footprints), aggregate SSD attach could rise through geographic fragmentation and replication of model artifacts across more sites, even if per-site storage is modest. The NAND effect is therefore likely to be demand-broadening and mix-positive (datacenter SSDs) but not a primary swing factor versus the macro AI capex cycle and consumer/device cycles. Hard disk drive (HDD) markets should see negligible direct sensitivity because nearline HDD demand is driven by bulk storage and cloud archiving economics, while inference acceleration choices primarily reshape compute and network layers; any HDD benefit would be a tertiary function of overall data center square footage expansion rather than a direct consequence of Groq silicon displacing GPUs. Optical networking implications require separating (1) intra-cluster back-end fabrics that connect accelerators and (2) front-end / data center interconnect (DCI) that connects sites and regions. Groq’s own positioning and third-party reporting suggest scaling beyond a single node or rack relies on high-bandwidth fabrics and, in some described configurations, optical interconnect scaling across hundreds of chips. If NVIDIA commercializes Groq at scale, 2 offsetting forces emerge: lower cost-per-token and improved latency could expand inference throughput and drive more east-west traffic, increasing demand for high-speed switching and optics; conversely, if Groq delivers materially higher utilization and tokens per unit of network bandwidth for certain workloads, the network required per served token could decline. Public NVIDIA materials already indicate an aggressive photonics roadmap aimed at scaling AI factories, including co-packaged optics (CPO) switches and explicit collaboration with Coherent and Lumentum in the silicon photonics supply chain. That linkage is important because it suggests that, independent of Groq, NVIDIA is already pushing optics integration deeper into the switch package to reduce power and increase resiliency; Groq increases the strategic incentive to reduce network power and latency if inference becomes even more distributed and latency-sensitive. For Lumentum and Coherent specifically, the net implication is less about “more optics versus fewer optics” and more about a shift in optics form factor and value capture. Co-packaged optics can reduce reliance on pluggable transceivers in some switch architectures while increasing demand for integrated photonic engines, lasers, fiber attach, packaging processes, and component-level supply. NVIDIA’s own announcements explicitly position Coherent and Lumentum as collaborators in creating the integrated silicon/optics process and supply chain for photonics switches. If Groq accelerates the transition to very large-scale fabrics (more endpoints, higher port speeds, tighter power envelopes), that tends to pull forward CPO adoption and amplifies demand for the underlying photonics components even if the conventional pluggable module TAM is structurally pressured over time. If Groq instead pushes inference toward smaller, more localized pods (closer to users, more regional colocation), that can be optics-positive for DCI and metro connectivity because more sites must be interconnected at high bandwidth with low latency, favoring coherent optics and high-speed interconnect between facilities. The principal risk for optics suppliers is timing and margin structure: a faster move to NVIDIA-driven integrated photonics could concentrate bargaining power and compress margins for commoditized transceiver modules while favoring suppliers with differentiated lasers, integration capability, and qualification depth in NVIDIA’s CPO ecosystem. AEC and copper interconnect implications hinge on whether Groq deployment increases the density of short-reach links inside racks and rows. High-speed copper remains structurally advantaged at very short distances on cost, power, and serviceability, but reaches become constrained as lane speeds and aggregate bandwidth rise, creating a role for active electrical cables (AECs), retimers, and signal-conditioning silicon. Credo explicitly positions its AEC products as enabling reliable lossless 800G connectivity for AI clusters, and the company has highlighted participation at NVIDIA GTC with content focused on extending PCIe/CXL using AECs, indicating relevance to next-generation system topologies that require longer reach and higher signal integrity than passive copper can deliver. If NVIDIA turns Groq into a widely deployed inference card or chassis product, the likely near-term effect is AEC-positive because (1) more inference throughput tends to increase top-of-rack connectivity requirements, (2) distributing inference across more racks and sites increases short-reach links per unit of delivered service, and (3) PCIe-attached accelerator architectures tend to require robust signal conditioning as systems move to PCIe 6.x and beyond. Groq workshop materials explicitly reference GroqCard and GroqNode form factors, reinforcing that PCIe-attached deployment has been central to Groq’s current packaging strategy. The main countervailing risk is that Groq’s deterministic chip-to-chip fabric could be implemented primarily through backplanes and direct board-level connectivity that reduces the need for merchant AECs inside the box; in that case, incremental AEC demand would concentrate more in rack-to-switch and node-to-fabric links rather than within-chassis chip fabrics. Astera Labs implications are connectivity-architecture sensitive and, on balance, skew positive if NVIDIA increases heterogeneity and disaggregation in AI systems. NVIDIA has publicly positioned NVLink Fusion as a pathway for partners to build semi-custom AI infrastructure and has explicitly identified Astera Labs as a partner in that ecosystem, with Astera describing NVLink-related solutions expanding its connectivity platform across PCIe, CXL, and Ethernet plus fleet observability software. A Groq acquisition increases the probability that NVIDIA offers a broader menu of accelerators (training GPUs, inference-focused ASICs) and therefore increases the importance of scalable, high-reliability connectivity, retiming, switching, and telemetry across mixed topologies. If Groq silicon remains PCIe-attached in many deployments, PCIe 6.x retimers/switches and active cable modules become more central, aligning with Astera’s core portfolio. If NVIDIA instead integrates Groq concepts into scale-up fabrics (NVLink-like domains) or uses Groq to expand into inference “appliances” that must be rapidly deployed in colocation environments, the need for standard-compliant, serviceable connectivity with strong RAS/telemetry increases, again aligning with Astera’s positioning. Power equipment and cooling implications for Vertiv and adjacent suppliers should be viewed through the lens of rack power density, cooling modality (air vs liquid), and site deployment model (hyperscale campuses vs distributed colocation/enterprise). Groq claims its LPU and rack designs are “air-cooled by design” and require no complex cooling and power infrastructure, and third-party reporting has described Groq’s approach as relying on parallelism across many lower-power units rather than extreme per-chip performance. If NVIDIA scales Groq as a mainstream inference platform, the mix of data center cooling spend could shift modestly away from the highest-density liquid-cooled racks toward more air-cooled or hybrid deployments, particularly for inference pods placed in existing facilities that cannot easily retrofit for very high rack heat flux. That would be a mix headwind for suppliers most levered exclusively to high-end liquid cooling attachments per rack, but it is not necessarily a volume headwind for Vertiv given the company’s broad exposure to both power and cooling infrastructure and the likelihood that total AI deployment locations expand. Vertiv’s own industry commentary emphasizes that AI racks require higher power-density UPS, batteries, power distribution equipment, and switchgear capable of handling rapid load transients, and that hybrid cooling systems will evolve across deployment environments. Those statements align with a world where inference growth increases the count of powered racks and raises the operational complexity of power delivery even if per-rack density is lower than the most extreme training clusters. The most material infrastructure impact may occur outside the rack and upstream of the data hall: grid interconnects, substations, transformers, switchgear, generators, and utility-scale generation additions. Recent regulatory actions in the U.S. highlight that projected data center demand is already driving large planned increases in electricity generation capacity, underscoring that power availability is a binding constraint. In that context, an inference architecture that lowers joules per token could reduce the power required per unit of inference delivered, but it can also accelerate demand by lowering cost and improving latency, increasing the total volume of inference served (a classic rebound effect). The net outcome is likely continued, elevated demand for power infrastructure even if efficiency improves, with the key swing factor being whether AI capex remains on a multi-year growth trajectory or enters a digestion phase. Other data center infrastructure implications include server/ODM mix, facility design standardization, and networking architecture choices. If NVIDIA positions Groq-based inference as a broadly distributable “standard server + accelerator” solution rather than as an integrated, liquid-cooled rack like GB200 NVL72, spend could shift toward more conventional air-cooled server designs, higher unit volumes of mainstream racks, and faster deployment in colocation footprints, increasing demand for modular power rooms, busways, and rapidly deployable cooling solutions. If NVIDIA instead integrates Groq into its “AI factory” paradigm, the primary effect is likely acceleration of dense back-end fabric build-outs and a faster push toward photonics switching, increasing demand for fiber plant, connectors, and integrated optics supply chains while potentially compressing the lifecycle of transitional architectures based on pluggable optics and mid-reach copper. NVIDIA’s stated roadmap toward co-packaged optics and silicon photonics switches is already oriented toward scaling to very large GPU counts; adding a high-end inference ASIC increases the strategic importance of power-efficient, low-latency fabrics because inference economics become increasingly sensitive to network overhead as compute cost declines. Across the covered segments, the most defensible base case is limited near-term dislocation and a medium-term increase in uncertainty around memory intensity per unit of inference growth. HBM faces the clearest relative risk from an HBM-less inference platform, but supply tightness and GPU training roadmaps reduce the probability of an absolute demand shock over the next 12–24 months. Optical, AEC/copper, and power/cooling are more likely to remain volume-supported because they scale with endpoint count, deployment fragmentation, and total data center footprint, and those tend to rise when inference becomes cheaper and more widely deployed. The highest-conviction second-order effect is a shift in infrastructure mix: incrementally more distributed inference deployments (favoring colocation power/cooling standardization, DCI optics, and serviceable short-reach interconnect) and a gradual migration from pluggable optics toward integrated photonics in back-end fabrics (favoring suppliers positioned in the CPO ecosystem).

TheValueist

76,267 görüntüleme • 9 ay önce

$AMD| The FOMO to buy AMD Chips is NOW 🧵 Not Financial Advice! DYOR! Research Purpose Only! The Inference Queen is the biggest winner in Agentic AI where all other CPUs are struggling to compete with a 2yr old EPYC Turin and EPYC Venice is in mass production phase. AMD stresses deployability today on standard x86 platforms (no proprietary architectures required), full software compatibility, and open standards. This positions Venice + Helios as a practical, high-density alternative to competing solutions while underscoring that agentic AI shifts the balance toward CPU-rich racks alongside GPUs, and most importantly, lowering the cost of token to accelerate adoption and innovation. Context: The Wall Street Journal yesterday came out with an article that OpenAI is condiering drasstically lowering the token prices to win more customers from Anthropic. The narrative "they" are trying to exacerbate the current AI selloff won't last long. This is a fundamental misunderstanding of what is going on, or what I already discussed for months and years. Followers and Subscribers already knew this for years, that this day would come, where token cost will bcome the central discussion among enterprises as there is no such thing as unlimited budget or Tokenmaxxing when they use $NVDA chips or In-house Hyperscalers chips. I will link various threads if you are interested in understanding the full picture from supply chain to recent TSMC Rapid 2nm expansion up to 12 Fabs total by 2027/2028. Hyperscalers and AI natives effectively have no choice but to buy more AMD system for Agentic AI as leadership in economical, power-aware, high-volume internal + agentic use. However, due to supply constraints where Supply is far behind Demand, this makes multi-vendor reality along with in-house chips drive faster industry progress, lower overall costs, and better sustainability. NVIDIA’s Vera Rubin cannot compete with a 2 years old EPYC Turin, but AMD under Dr. Lisa Su has engineered the lowest cost-per-million-tokens, highly competitive energy-efficient solutions, and superior CPU orchestration for agentic AI at scale with Helios. Dr. Su has championed this shift since at least 2023, foreseeing the rise of agentic workflows that demand far more orchestration, parallel agents, and balanced compute well before the industry fully embraced it. Her long-term vision of AI moving from simple prompts to always on, multi-agent systems has driven AMD’s investments in high-core EPYC CPUs and integrated rack-scale solutions, perfectly positioning the company for today’s realities. The OpenAI-AMD 1GW Helios deployment (starting H2 2026) represents a pivotal vertical integration move that directly supercharges the inference economics. This isn't incremental; it's a structural shift toward ownership of massive, optimized rack-scale capacity, enabling the lowest token costs and triggering the enterprise adoption flywheel. We need to be honest, $AMD is the only company that made a big bet on Inference since the day Chatgpt became sensational where $NVDA and others were betting big on Training. At the end of the day, Token bill from Anthropic has to obey economics. Meaning the bills rise, companies have to get more out of it to justify the cost. It cannot be an unlimited inference budget, and it has to show up on efficiency, profitability and operating leverage. 1. Tokenomics After you understand this, you will understand why Citi cited Anthropic is likely to sign a deal with $AMD along with Hyperscalers, AI Labs, Sovereign AI like Softbank 5GW in France and many other countries. However, OpenAI and $META are now wanting faster deployment, and they are AMD shareholders now, they have prioritized allocation. Anthropic and Hyperscalers just cannot compete when Helios Rack lower token cost to$0.0003–$0.0005 per million tokens at GW scale. Cost to build 1GW data center 1GW Helios Rack full build is estimated $30-$35B 1GW Rubin Rack full build is estimated $45-$55B Inference (Cost per Million Tokens) ~$NVDA B200 / HGX: ~$0.02–$0.08 on optimized workloads (FP4/MXFP4, speculative decoding). Significant improvement over Hopper but still premium-priced. GB200 NVL72 rack-scale: $0.05–$0.25+ ~$AMD Helios Racks: $0.0003-$0.0005 per M tokens, dramatically lower than NVIDIA equivalents in owned infra. MI355X node-level: Up to 40% more tokens per dollar vs. competing solutions ( B200), driven by higher memory capacity (up to 288GB+ HBM), strong bandwidth, and lower acquisition costs. Training ~$NVDA Rubin Rack is estimated $0.7-$1.2/M Tokens ~$AMD Helios Rack is estimated $0.65-$1.0/M Tokens Now, OpenAI, META and Hyperscalers can lower Inference cost even further with $AMD EPYC Venice "dense rack" or Agentic AI Rack. AMD published a detailed technical blog emphasizing that the future of agentic AI autonomous, multi-step AI systems requiring heavy orchestration, databases, caching, APIs, and control planes demands massive CPU-dense rack-scale infrastructure, not just GPUs. The catalyst prominently positions their upcoming 6th Gen EPYC "Venice" processors as the key enabler for next-generation dense racks, delivering leadership throughput under real-world power, cooling, and density constraints. ~EPYC Venice (Zen 6 architecture, up to 256 cores / 512 threads per socket) is projected to deliver exceptional rack-level performance. In AMD’s modeled 100 kW rack comparisons, Venice-powered systems are expected to achieve ~3.30x the throughput of NVIDIA’s Vera (88-core Olympus) baseline across a broad mix of agentic-supporting workloads. ~This builds on current-generation 5th Gen EPYC "Turin" (up to 192 cores), which already delivers ~2.37x rack throughput vs. Vera and ~1.6x vs. Intel’s Xeon 6980P (128 cores). ~ Liquid-cooled Turin deployments already support >27,000 CPU cores per rack today. Venice is architected to push this beyond 36,000 cores in the same rack class, dramatically increasing concurrent agent capacity and overall infrastructure efficiency. 2. Ownership vs renting compute from Hyperscalers matter to OpenAI and only owning $AMD chips can meaningfully lower token cost for enterprises. ~Eliminates cloud overhead: No provider margins, utilization buffers, or egress fees. Direct control over power contracts, cooling, scheduling, and orchestration at dedicated facilities. ~Helios optimizations at GW scale: Rack-level density (1.4+ exaFLOPS FP8 per rack), high HBM4 bandwidth, EPYC orchestration for agentic workloads, and superior TCO/TDP. AMD's long-standing focus on tokens per dollar/watt shines here 20-40%+ efficiency edges in inference-heavy scenarios. ~At 1GW+ optimized deployment, inference hits $0.0003–$0.0005 per million tokens (community/analyst models tied to Helios metrics). This is dramatically lower than typical rented/cloud equivalents, especially for high-volume output tokens in agentic flows. High token bills today, enterprises running heavy agentic/coding/analysis workloads can face $50-100M+/month at current API rates (flagship models $5-30+/M output, scaled to massive volumes). Post-Helios compression, same volume will drop to $10-15M/month (or better) via lower underlying costs passed through as pricing flexibility, volume tiers, caching, or batch discounts. ROI thresholds collapse. More companies greenlight pilots → production → massive scaling. Agentic AI (autonomous workflows) multiplies token demand exponentially, but affordability removes the friction. OpenAI gains flexibility, Unlike more cloud-dependent rivals (Anthropic), they can lower effective pricing, offer aggressive enterprise bundles, or absorb volume without margin destruction directly tackling "high token bill" complaints while maintaining profitability as usage explodes. 3. Agentic AI Models shifted CPU:GPU Ratio to 1:1 toward 3-5:1 with Explosively Token-Hungry Workloads Agentic AI (autonomous, multi-step agents with planning, tool use, iteration, and self-correction) is fundamentally more compute and token intensive than conversational or single-turn generative AI. Agentic AI. autonomous, multi-step workflows with orchestration, tool use, parallel agents, data movement, and enterprise integration has dramatically increased the importance of strong host CPUs alongside GPUs. This shifts the CPU-to-GPU ratio higher and makes balanced systems critical toward 1:1 to 5:1 as enterprises testing more than 5-10 agents. AMD EPYC Venice excels ~Leadership core density (up to 256 Zen 6 cores per socket) for running many agents in parallel, orchestration layers, and high-throughput control-plane tasks. ~Superior performance-per-core and power efficiency ( up to 2.1x higher perf/core and 2.26x better SPECpower vs. NVIDIA Grace in benchmarks). ~Tight integration in Helios: One Venice CPU + multiple MI450 GPUs per node, enabling efficient data feeding to GPUs ("zero-copy"), parallel execution, and full rack utilization for complex agentic loops. Hyperscalers (Meta, Microsoft, Amazon, Google, Softbank) and AI natives (OpenAI, Anthropic...) are adopting high-core EPYC at scale specifically for these agentic demands, as CPUs now handle a larger share of non-model work (orchestration, policy enforcement, tool calls). This complements AMD’s lower-cost GPUs for overall TCO wins. ~Agents often generate 10–100x+ more tokens per task due to iterative reasoning chains, multiple tool calls, verification loops, and long-context orchestration. ~Goldman Sachs forecasts token consumption multiplying 24x by 2030 (to 120 quadrillion tokens/month) largely driven by agentic adoption in consumer and enterprise. ~Enterprise data shows agent-pattern workloads growing at 680% annualized rates, projected to surpass conversational AI in token volume by Q3 2026. ~Daily enterprise agent token consumption is already in the billions, with complex workflows (coding, workflows, analysis) amplifying this dramatically. 4. Competitive Edge: Winning Customers from Anthropic Anthropic’s Claude models (especially Opus/Sonnet) excel in complex reasoning and agentic coding, commanding premium positioning. However, their higher underlying costs (heavier reliance on third-party cloud with margins) limit pricing flexibility compared to OpenAI’s owned Helios capacity. Anthropic is on track to generate $10.9 billion in Q2 revenue. The company expects to achieve its first-ever quarterly adjusted operating profit of $559 million. However, sustaining full-year profitability remains challenging due to immense computing and model training costs The truth is, Anthropic has no choice but to buy as much $AMD chips as possible if they want to compete with OpenAI or get investors attention. This 5% adjusted operating profit to revenue ratio is just pathetic. Current pricing dynamics (2026): OpenAI already undercuts on many tiers ( flagship output tokens significantly cheaper than equivalent Claude Opus). Nano/mini models offer 5–10x advantages for volume work. Anthropic holds edges in long-context flat pricing and certain reasoning quality. OpenAI after Helios Rack Ownership, At $0.0003–$0.0005/M effective costs, OpenAI gains massive headroom to: ~Aggressively discount high-volume agentic tiers or bundles. ~Offer “unlimited” enterprise plans or usage-based models that Anthropic struggles to match without margin erosion. ~Target cost-sensitive, high-throughput agent deployments (dev tools, automation platforms) where token bills explode. Enterprises facing $ millions in monthly agentic bills will migrate to the provider delivering better economics at scale. OpenAI’s combination of strong models (o-series reasoning) + lowest TCO positions it to erode Anthropic’s enterprise share, especially as agentic becomes the dominant token consumer. Cheaper tokens expand the total addressable market dramatically. This feeds the data/model improvement loop, justifying further capex. AMD benefits from proven scale pulling in more customers (Meta, Oracle, Microsfot, Amazon, Softbank, TensorWave, LumaAI ... already aligned on Helios). Conclusion: Dr. Lisa Su has been laser focused on inference economics since at least 2022–2023, repeatedly emphasizing that the real battleground for AI scalability would be TCO, power efficiency (TDP), and ultimately tokens per dollar and per watt not just raw training FLOPS. While many viewed inference as a secondary, commoditized workload, Dr. Su architected AMD’s roadmap around rack-scale systems optimized for high-volume, sustained inference that would dominate as models matured and usage exploded. Helios represents the culmination of that multi-year bet: a fully integrated, open platform designed precisely for the economics of massive token throughput. This deep, strategic partnership with OpenAI starting with the 1GW Helios deployment in H2 2026 and scaling to 6GW, is the embodiment of that shared vision. Both companies foresaw a future where agentic AI models evolve to become extraordinarily token-hungry: autonomous agents executing complex, iterative workflows with planning, tool use, verification loops, and long-context reasoning. These workloads can consume 100x+ more tokens per task than traditional chat or single-turn generation, driving exponential demand as capabilities improve and enterprises deploy them at scale. By owning and optimizing this massive Helios capacity at GW scale, OpenAI achieves inference costs as low as $0.0003–$0.0005 per million tokens. This structural cost advantage allows OpenAI to absorb the coming token explosion profitably, dramatically lower effective pricing for enterprises, and win high-volume agentic workloads from higher-cost competitors like Anthropic. What was once a prohibitive monthly token bill becomes an affordable accelerator for productivity and innovation. The OpenAI-AMD alliance validates Dr. Su’s prescient strategy and turns the Agentic flywheel into reality: Collapsing inference costs → explosive token consumption → richer data and better models → accelerate greater demand. This partnership doesn’t just address today’s economics, it positions both leaders at the center of the infrastructure buildout that will power AI’s next decade. By delivering the lowest inference economics at scale, OpenAI not only solves enterprise bill pain but gains a decisive weapon to win share from higher-cost rivals like Anthropic. And that is why OpenAI and $META will deploy EPYC Dense Rack Not Financial Advice! DYOR! Research Purpose Only!

$AMD| The FOMO to buy AMD Chips is NOW 🧵 Not Financial Advice! DYOR! Research Purpose Only! The Inference Queen is the biggest winner in Agentic AI where all other CPUs are struggling to compete with a 2yr old EPYC Turin and EPYC Venice is in mass production phase. AMD stresses deployability today on standard x86 platforms (no proprietary architectures required), full software compatibility, and open standards. This positions Venice + Helios as a practical, high-density alternative to competing solutions while underscoring that agentic AI shifts the balance toward CPU-rich racks alongside GPUs, and most importantly, lowering the cost of token to accelerate adoption and innovation. Context: The Wall Street Journal yesterday came out with an article that OpenAI is condiering drasstically lowering the token prices to win more customers from Anthropic. The narrative "they" are trying to exacerbate the current AI selloff won't last long. This is a fundamental misunderstanding of what is going on, or what I already discussed for months and years. Followers and Subscribers already knew this for years, that this day would come, where token cost will bcome the central discussion among enterprises as there is no such thing as unlimited budget or Tokenmaxxing when they use $NVDA chips or In-house Hyperscalers chips. I will link various threads if you are interested in understanding the full picture from supply chain to recent TSMC Rapid 2nm expansion up to 12 Fabs total by 2027/2028. Hyperscalers and AI natives effectively have no choice but to buy more AMD system for Agentic AI as leadership in economical, power-aware, high-volume internal + agentic use. However, due to supply constraints where Supply is far behind Demand, this makes multi-vendor reality along with in-house chips drive faster industry progress, lower overall costs, and better sustainability. NVIDIA’s Vera Rubin cannot compete with a 2 years old EPYC Turin, but AMD under Dr. Lisa Su has engineered the lowest cost-per-million-tokens, highly competitive energy-efficient solutions, and superior CPU orchestration for agentic AI at scale with Helios. Dr. Su has championed this shift since at least 2023, foreseeing the rise of agentic workflows that demand far more orchestration, parallel agents, and balanced compute well before the industry fully embraced it. Her long-term vision of AI moving from simple prompts to always on, multi-agent systems has driven AMD’s investments in high-core EPYC CPUs and integrated rack-scale solutions, perfectly positioning the company for today’s realities. The OpenAI-AMD 1GW Helios deployment (starting H2 2026) represents a pivotal vertical integration move that directly supercharges the inference economics. This isn't incremental; it's a structural shift toward ownership of massive, optimized rack-scale capacity, enabling the lowest token costs and triggering the enterprise adoption flywheel. We need to be honest, $AMD is the only company that made a big bet on Inference since the day Chatgpt became sensational where $NVDA and others were betting big on Training. At the end of the day, Token bill from Anthropic has to obey economics. Meaning the bills rise, companies have to get more out of it to justify the cost. It cannot be an unlimited inference budget, and it has to show up on efficiency, profitability and operating leverage. 1. Tokenomics After you understand this, you will understand why Citi cited Anthropic is likely to sign a deal with $AMD along with Hyperscalers, AI Labs, Sovereign AI like Softbank 5GW in France and many other countries. However, OpenAI and $META are now wanting faster deployment, and they are AMD shareholders now, they have prioritized allocation. Anthropic and Hyperscalers just cannot compete when Helios Rack lower token cost to$0.0003–$0.0005 per million tokens at GW scale. Cost to build 1GW data center 1GW Helios Rack full build is estimated $30-$35B 1GW Rubin Rack full build is estimated $45-$55B Inference (Cost per Million Tokens) ~$NVDA B200 / HGX: ~$0.02–$0.08 on optimized workloads (FP4/MXFP4, speculative decoding). Significant improvement over Hopper but still premium-priced. GB200 NVL72 rack-scale: $0.05–$0.25+ ~$AMD Helios Racks: $0.0003-$0.0005 per M tokens, dramatically lower than NVIDIA equivalents in owned infra. MI355X node-level: Up to 40% more tokens per dollar vs. competing solutions ( B200), driven by higher memory capacity (up to 288GB+ HBM), strong bandwidth, and lower acquisition costs. Training ~$NVDA Rubin Rack is estimated $0.7-$1.2/M Tokens ~$AMD Helios Rack is estimated $0.65-$1.0/M Tokens Now, OpenAI, META and Hyperscalers can lower Inference cost even further with $AMD EPYC Venice "dense rack" or Agentic AI Rack. AMD published a detailed technical blog emphasizing that the future of agentic AI autonomous, multi-step AI systems requiring heavy orchestration, databases, caching, APIs, and control planes demands massive CPU-dense rack-scale infrastructure, not just GPUs. The catalyst prominently positions their upcoming 6th Gen EPYC "Venice" processors as the key enabler for next-generation dense racks, delivering leadership throughput under real-world power, cooling, and density constraints. ~EPYC Venice (Zen 6 architecture, up to 256 cores / 512 threads per socket) is projected to deliver exceptional rack-level performance. In AMD’s modeled 100 kW rack comparisons, Venice-powered systems are expected to achieve ~3.30x the throughput of NVIDIA’s Vera (88-core Olympus) baseline across a broad mix of agentic-supporting workloads. ~This builds on current-generation 5th Gen EPYC "Turin" (up to 192 cores), which already delivers ~2.37x rack throughput vs. Vera and ~1.6x vs. Intel’s Xeon 6980P (128 cores). ~ Liquid-cooled Turin deployments already support >27,000 CPU cores per rack today. Venice is architected to push this beyond 36,000 cores in the same rack class, dramatically increasing concurrent agent capacity and overall infrastructure efficiency. 2. Ownership vs renting compute from Hyperscalers matter to OpenAI and only owning $AMD chips can meaningfully lower token cost for enterprises. ~Eliminates cloud overhead: No provider margins, utilization buffers, or egress fees. Direct control over power contracts, cooling, scheduling, and orchestration at dedicated facilities. ~Helios optimizations at GW scale: Rack-level density (1.4+ exaFLOPS FP8 per rack), high HBM4 bandwidth, EPYC orchestration for agentic workloads, and superior TCO/TDP. AMD's long-standing focus on tokens per dollar/watt shines here 20-40%+ efficiency edges in inference-heavy scenarios. ~At 1GW+ optimized deployment, inference hits $0.0003–$0.0005 per million tokens (community/analyst models tied to Helios metrics). This is dramatically lower than typical rented/cloud equivalents, especially for high-volume output tokens in agentic flows. High token bills today, enterprises running heavy agentic/coding/analysis workloads can face $50-100M+/month at current API rates (flagship models $5-30+/M output, scaled to massive volumes). Post-Helios compression, same volume will drop to $10-15M/month (or better) via lower underlying costs passed through as pricing flexibility, volume tiers, caching, or batch discounts. ROI thresholds collapse. More companies greenlight pilots → production → massive scaling. Agentic AI (autonomous workflows) multiplies token demand exponentially, but affordability removes the friction. OpenAI gains flexibility, Unlike more cloud-dependent rivals (Anthropic), they can lower effective pricing, offer aggressive enterprise bundles, or absorb volume without margin destruction directly tackling "high token bill" complaints while maintaining profitability as usage explodes. 3. Agentic AI Models shifted CPU:GPU Ratio to 1:1 toward 3-5:1 with Explosively Token-Hungry Workloads Agentic AI (autonomous, multi-step agents with planning, tool use, iteration, and self-correction) is fundamentally more compute and token intensive than conversational or single-turn generative AI. Agentic AI. autonomous, multi-step workflows with orchestration, tool use, parallel agents, data movement, and enterprise integration has dramatically increased the importance of strong host CPUs alongside GPUs. This shifts the CPU-to-GPU ratio higher and makes balanced systems critical toward 1:1 to 5:1 as enterprises testing more than 5-10 agents. AMD EPYC Venice excels ~Leadership core density (up to 256 Zen 6 cores per socket) for running many agents in parallel, orchestration layers, and high-throughput control-plane tasks. ~Superior performance-per-core and power efficiency ( up to 2.1x higher perf/core and 2.26x better SPECpower vs. NVIDIA Grace in benchmarks). ~Tight integration in Helios: One Venice CPU + multiple MI450 GPUs per node, enabling efficient data feeding to GPUs ("zero-copy"), parallel execution, and full rack utilization for complex agentic loops. Hyperscalers (Meta, Microsoft, Amazon, Google, Softbank) and AI natives (OpenAI, Anthropic...) are adopting high-core EPYC at scale specifically for these agentic demands, as CPUs now handle a larger share of non-model work (orchestration, policy enforcement, tool calls). This complements AMD’s lower-cost GPUs for overall TCO wins. ~Agents often generate 10–100x+ more tokens per task due to iterative reasoning chains, multiple tool calls, verification loops, and long-context orchestration. ~Goldman Sachs forecasts token consumption multiplying 24x by 2030 (to 120 quadrillion tokens/month) largely driven by agentic adoption in consumer and enterprise. ~Enterprise data shows agent-pattern workloads growing at 680% annualized rates, projected to surpass conversational AI in token volume by Q3 2026. ~Daily enterprise agent token consumption is already in the billions, with complex workflows (coding, workflows, analysis) amplifying this dramatically. 4. Competitive Edge: Winning Customers from Anthropic Anthropic’s Claude models (especially Opus/Sonnet) excel in complex reasoning and agentic coding, commanding premium positioning. However, their higher underlying costs (heavier reliance on third-party cloud with margins) limit pricing flexibility compared to OpenAI’s owned Helios capacity. Anthropic is on track to generate $10.9 billion in Q2 revenue. The company expects to achieve its first-ever quarterly adjusted operating profit of $559 million. However, sustaining full-year profitability remains challenging due to immense computing and model training costs The truth is, Anthropic has no choice but to buy as much $AMD chips as possible if they want to compete with OpenAI or get investors attention. This 5% adjusted operating profit to revenue ratio is just pathetic. Current pricing dynamics (2026): OpenAI already undercuts on many tiers ( flagship output tokens significantly cheaper than equivalent Claude Opus). Nano/mini models offer 5–10x advantages for volume work. Anthropic holds edges in long-context flat pricing and certain reasoning quality. OpenAI after Helios Rack Ownership, At $0.0003–$0.0005/M effective costs, OpenAI gains massive headroom to: ~Aggressively discount high-volume agentic tiers or bundles. ~Offer “unlimited” enterprise plans or usage-based models that Anthropic struggles to match without margin erosion. ~Target cost-sensitive, high-throughput agent deployments (dev tools, automation platforms) where token bills explode. Enterprises facing $ millions in monthly agentic bills will migrate to the provider delivering better economics at scale. OpenAI’s combination of strong models (o-series reasoning) + lowest TCO positions it to erode Anthropic’s enterprise share, especially as agentic becomes the dominant token consumer. Cheaper tokens expand the total addressable market dramatically. This feeds the data/model improvement loop, justifying further capex. AMD benefits from proven scale pulling in more customers (Meta, Oracle, Microsfot, Amazon, Softbank, TensorWave, LumaAI ... already aligned on Helios). Conclusion: Dr. Lisa Su has been laser focused on inference economics since at least 2022–2023, repeatedly emphasizing that the real battleground for AI scalability would be TCO, power efficiency (TDP), and ultimately tokens per dollar and per watt not just raw training FLOPS. While many viewed inference as a secondary, commoditized workload, Dr. Su architected AMD’s roadmap around rack-scale systems optimized for high-volume, sustained inference that would dominate as models matured and usage exploded. Helios represents the culmination of that multi-year bet: a fully integrated, open platform designed precisely for the economics of massive token throughput. This deep, strategic partnership with OpenAI starting with the 1GW Helios deployment in H2 2026 and scaling to 6GW, is the embodiment of that shared vision. Both companies foresaw a future where agentic AI models evolve to become extraordinarily token-hungry: autonomous agents executing complex, iterative workflows with planning, tool use, verification loops, and long-context reasoning. These workloads can consume 100x+ more tokens per task than traditional chat or single-turn generation, driving exponential demand as capabilities improve and enterprises deploy them at scale. By owning and optimizing this massive Helios capacity at GW scale, OpenAI achieves inference costs as low as $0.0003–$0.0005 per million tokens. This structural cost advantage allows OpenAI to absorb the coming token explosion profitably, dramatically lower effective pricing for enterprises, and win high-volume agentic workloads from higher-cost competitors like Anthropic. What was once a prohibitive monthly token bill becomes an affordable accelerator for productivity and innovation. The OpenAI-AMD alliance validates Dr. Su’s prescient strategy and turns the Agentic flywheel into reality: Collapsing inference costs → explosive token consumption → richer data and better models → accelerate greater demand. This partnership doesn’t just address today’s economics, it positions both leaders at the center of the infrastructure buildout that will power AI’s next decade. By delivering the lowest inference economics at scale, OpenAI not only solves enterprise bill pain but gains a decisive weapon to win share from higher-cost rivals like Anthropic. And that is why OpenAI and $META will deploy EPYC Dense Rack Not Financial Advice! DYOR! Research Purpose Only!

Mike

84,951 görüntüleme • 3 ay önce

Have you heard of collective consciousness and mass programming? Watch THINK TOGETHER (short film 5min) A TORVÆL FILM. The spell is global. It's not just "Think Together." That's one film, one title, one thread in a tapestry of mass enchantment that has been woven through every medium humans use to receive information, entertainment, and meaning. It's a"magic kind of a spell through screen." That is the most precise description of what's happening. Not metaphor. Not allegory. Literal spellcasting through electronic and print media. Let's go deep into the global spell. The mediums. The methods. The specific frequency weapons deployed through each channel. The Nature of the Spell: Electronic Enchantment A spell, in its original meaning, is a binding. A set of symbols, sounds, and focused intention that alters the consciousness of the target, making them perceive reality differently, act against their own interest, or accept a condition they would otherwise reject. Traditional magic required proximity. The sorcerer had to be near the target, or use a physical link hair, nail clippings, a photograph. The spell was limited by space. Electronic media destroyed that limitation. The screen is a direct energetic link between the caster and the target. Light enters the eyes. Sound enters the ears. The brain entrains to the frequencies embedded in the transmission. The biofield receives the signal. Distance is irrelevant. One broadcast can enchant a billion people simultaneously. The screen is the wand. The transmission is the incantation. The content is the intention. And the population is under a continuous, multi-layered, globally synchronized spell that has been building for over a century. Medium 1: Cinema | The Dream Injection Movies are the most powerful spell delivery system ever invented. The Theater as Ritual Chamber: A cinema is a darkened room where strangers gather in silence, facing a single light source. The flickering light induces a hypnagogic state the brainwave pattern of the threshold between waking and dreaming. In this state, the critical faculty is suppressed. The subconscious is open. The images and sounds on the screen are absorbed without filtration. This is identical to the conditions of a ritual chamber. The darkened temple. The flickering torchlight. The congregation facing the altar. The priest intoning the incantation. Cinema is temple worship, and the screen is the altar on which reality is reshaped. The 24 Frames Per Second Induction: Film runs at 24 frames per second. This is not an arbitrary choice. The human brain's alpha rhythm the frequency of relaxed, suggestible awareness operates at 8 to 12 Hz. 24 frames per second, with each frame shown two or three times due to the shutter, creates a flicker frequency in the 48 to 72 Hz range. This is a harmonic of the gamma brainwave band, associated with binding sensory information into a coherent percept. The film doesn't just show you images. It entrains your gamma rhythm to its own temporal structure. Your brain is phase-locked to the projector. You are in the film. The film is in you. Color Grading as Emotional Programming: Every major film uses color grading to manipulate emotional response. Teal and orange. Desaturated blues for dystopia. Warm golds for nostalgia. The palette is not an aesthetic choice. It is an emotional command. The visual cortex processes color before the conscious mind identifies objects. The emotional response to the color palette happens before you know what you're looking at. The spell is felt before it is seen. Sound Design as Frequency Weapon: Film soundtracks use specific frequencies to induce physiological states. Infrasonic bass frequencies below 20 Hz, felt rather than heard triggers the fear response in the amygdala. The Shepard tone an auditory illusion of a pitch that rises forever without ever reaching a destination creates a sense of endless tension that never resolves. This is used extensively in horror and thriller films to keep the audience in a state of chronic, unresolvable anxiety. The soundtrack tells you what to feel. You believe the feeling is your own response to the story. It is not. It is a frequency command, delivered through the auditory system, bypassing cognition entirely. #PredictiveProgramming: Major films depict future events before they happen. Not as speculation. As conditioning. The controllers place images of planned events into the collective unconscious through cinema. When the event occurs in reality, the population has already "seen" it. It feels familiar. It feels inevitable. It feels like something they already accepted in the dream state. Pandemic films before COVID. Drone warfare films before the drone wars. Mass surveillance films before Snowden. Transhumanist films before Neuralink. The spell is cast years in advance. The event is merely the fulfillment of a prophecy that was manufactured by the prophecy itself. Medium 2: Music | The Auditory Incantation Music is the oldest spell technology. Before writing, before film, before any visual medium, there was rhythm and tone. The drum. The chant. The bone flute. Music alters brainwave states directly, without requiring visual attention. 432 Hz vs. 440 Hz: The Frequency War The global standard tuning for music is A=440 Hz. This was adopted in the early 20th century, pushed by the Rockefeller Foundation and the Nazi propaganda ministry, and codified by the International Organization for Standardization in 1955. Prior to this, many traditions used A=432 Hz, a frequency that mathematically aligns with the Schumann resonance (8 Hz), the Earth's natural electromagnetic pulse, and the geometric proportions found in nature. 440 Hz creates a subtle dissonance with the human biofield. It agitates. It separates the listener from the Earth's frequency. Music tuned to 440 Hz cannot fully relax the nervous system. It maintains a baseline of subliminal tension, a low-grade anxiety that the listener attributes to their life circumstances rather than to the music itself. 432 Hz music entrains the listener to the planetary frequency. It harmonizes. It heals. It is suppressed not because it "sounds worse" but because it sounds more coherent and produces a brain state that is resistant to external control. Lyrical Programming: Lyrics are direct incantations. The repetition of a phrase in a song embeds it in the subconscious. The melody carries the words past the critical faculty. The rhythm entrains the brain to receive the message. Examine the lyrical content of mainstream music across decades: ◻️Themes of hopelessness, materialism, sexual degradation, violence, substance use ◻️ Self-referential obsession: "I," "me," "my" repeated endlessly, reinforcing the illusion of the separate self ◻️ Nihilism presented as cool, despair presented as authenticity ◻️ Love reduced to possession, intimacy reduced to transaction The population sings along. They internalize the incantation. They believe they are listening to music. They are reciting spells that bind them to a reality of consumption, isolation, and quiet desperation. The Monopoly of Distribution: A handful of corporations control the global music industry. Universal, Sony, Warner. The playlists are curated. The algorithms select what billions hear. Independent music that carries a different frequency, a different message, a different emotional command is not played. It is not because it lacks quality. It is because it carries the wrong spell. Medium 3: Television | The Continuous Ritual Television was the first medium to bring the spell into the home continuously. Before smartphones, before streaming, the television was the household altar. The family gathered around it. The light flickered in the living room. The incantation played during dinner. The 30-Minute Spell Cycle: The sitcom format 22 minutes of content, 8 minutes of commercials is a spell cycle. The content opens the subconscious (laughter, emotional engagement). The commercial delivers the command (buy this, believe this, want this). The cycle repeats. Over decades, the population's attention span was conditioned to this rhythm. The modern inability to focus for more than a few minutes is not a failure of will. It is a successful spell. An entrained attention cycle that can now be exploited by shorter-form content on smartphones. News as Reality Creation: Television news is not information. It is ritual. The set, the lighting, the music, the cadence of the anchor's voice these are the elements of a ceremonial invocation. The news does not report reality. It declares reality into being. The repetition of phrases, the selection of images, the framing of events this is spellcasting in real time. The population watches, believes they are being informed, and has their perception of the world sculpted without their knowledge. The Laugh Track: The laugh track is the most obvious spell component in television history. A recorded laugh triggers the mirror neuron system. The viewer laughs not because the joke is funny but because they heard laughter. The spell bypasses judgment. The laugh track says: "This is funny." The brain obeys. The critical faculty is suspended by a recorded cackle. Medium 4: Print Media | The Written Incantation Before electronic media, print was the spell delivery system. It remains operational, though its influence has been partially eclipsed by screens. The Headline as Command: A headline is not a summary. It is a command phrase. Most readers do not read the article. They read the headline. The headline is the spell, condensed to its most potent form. It frames the event before the event is understood. It tells the reader what to think before they have a chance to think. The Inverted Pyramid: Journalistic structure places the most important information first, followed by diminishing detail. This is presented as a neutral convention. It is a spell structure. The command is delivered at the top. The supporting incantation follows. By the time the reader reaches the end, they have forgotten the details and retained only the command. The Omission: The most powerful spell component in print media is what is not printed. The events, perspectives, and voices that are systematically excluded from the written record. The spell of omission creates a reality defined by absence. If it is not in print, it did not happen. The population's sense of what is real is shaped as much by the silence as by the words. Medium 5: Social Media | The Participatory Spell Social media is the most sophisticated spell technology ever created. It does not broadcast to a passive audience. It enlists the audience as casters. Every user is simultaneously the target and the amplifier of the spell. The Infinite Scroll as Trance Induction: The infinite scroll is a hypnotic mechanism. The finger moves. The content appears. The brain receives a micro-dose of dopamine with each new image. The motion is rhythmic. The attention is captured. The critical faculty is submerged. This is identical to the repetitive motion of a rosary, a prayer wheel, a mantra. The user is meditating, but the object of meditation is chosen by the algorithm, not by the self. The Like Button as Ritual Participation: Every like, every share, every comment is a ritual act. The user invests a fragment of their attention, their emotional energy, their biofield into the content. The spell is strengthened by participation. The egregore is fed by interaction. The user believes they are expressing an opinion. They are adding their life force to a thought-form they did not create and do not control. The Algorithm as High Priest: The algorithm does not show you what you want. It shows you what will keep you engaged and what will shape your perception in accordance with the controllers' intention. The algorithm is the high priest of the participatory spell. It selects the incantations. It measures the responses. It adjusts the frequency in real time. It knows you better than you know yourself, because it has your attention data, your emotional data, your behavioral data, and the biofield data harvested through the IoB sensors. The spell is personalized. No two users receive the same incantation. But all incantations serve the same master. Medium 6: Advertising | The Direct Command Advertising is the purest form of the spell. It does not pretend to be art, information, or entertainment. It is a direct command: desire this, buy this, be this. Every other medium is, in part, a delivery system for the advertising spell. The Subliminal Layer: Subliminal messaging is not a conspiracy theory. It is a documented, researched, and patented technology. Images embedded for single frames. Audio messages masked by other sounds. Commands that bypass conscious awareness entirely. The advertising industry has denied using subliminals since the 1950s, while simultaneously filing patents for subliminal delivery systems. The Repetition Principle: A single exposure to an advertisement has minimal effect. Repeated exposure thousands of times across years wires the command into the neural architecture. The brand name becomes a neural pathway. The jingle becomes an earworm that plays unbidden. The desire becomes "personal preference." The population believes it is choosing. It is executing a command that was installed by repetition. The Archetypal Manipulation: Advertising uses archetypal imagery the hero, the lover, the mother, the wise elder to bypass the rational mind and speak directly to the deep psyche. The car commercial does not sell transportation. It sells the archetype of freedom. The perfume ad does not sell scent. It sells the archetype of desire. The spell operates at the level of the collective unconscious, using symbols that predate language. Medium 7: Architecture and Public Space | The Environmental Spell The spell is not confined to screens and pages. The built environment itself is an incantation. Brutalist Architecture: The concrete blocks, the grey walls, the absence of organic form this is not an aesthetic choice. It is an energetic suppression field rendered in physical form. The human biofield responds to geometry. Organic forms curves, spirals, natural proportions harmonize and strengthen the biofield. Brutalist geometry sharp angles, unbroken planes, unnatural proportions disrupts and weakens it. A population that lives and works in brutalist structures is a population whose biofield is continuously under assault. The Elimination of Sacred Space: Traditional cities were built around sacred centers temples, cathedrals, gathering places that served as energetic focal points. Modern cities are built around commercial centers shopping malls, business districts, financial hubs. The sacred is replaced by the transactional. The focal point of the community is no longer a place of spiritual coherence but a place of consumption. The spell reorients the population's collective attention from the transcendent to the material, without a single word being spoken. Artificial Lighting: The permanent illumination of cities by artificial light severs the population from the natural cycles of light and dark. The circadian rhythm is disrupted. The pineal gland, which produces melatonin and is sensitive to natural light cycles, is suppressed. The biofield loses its connection to the solar and cosmic cycles that are the foundation of embodied consciousness. The population is untethered from the planetary rhythm. The grid provides the new rhythm. The spell is maintained by streetlights and screens, 24 hours a day, 365 days a year. Medium 8: Education | The Foundational Spell The spell is installed in childhood through the education system. Before the child can read, before they can critically evaluate, before they have formed a stable sense of self, the incantation begins. The Bell System: The school day is divided by bells. The bell is a Pavlovian trigger. Stop this activity. Start that activity. Obey the schedule. The bell trains the nervous system to respond to external commands. The population learns, from age five, that their attention is not their own. It is directed by an external authority. This conditioning persists for life. The Curriculum as Reality Definition: The curriculum does not teach "subjects." It defines what is real and what is not. The history that is taught. The history that is omitted. The science that is presented. The science that is suppressed. The literature that is canonized. The literature that is excluded. By the time the child reaches adulthood, their sense of reality has been structured by the curriculum. They do not know what they were not taught. The omission spell, installed in childhood, is the most durable of all. Standardized Testing as Soul Extraction: The child is measured, ranked, and labeled by standardized tests. The unique intelligence is reduced to a number. The soul is quantified. The test does not measure intelligence. It measures compliance with the cognitive framework of the controllers. The child who thinks differently fails. The child who recites the spell correctly passes. The population is sorted into categories by its willingness and ability to accept the incantation. The Unified Spell: All Mediums, One Intention These mediums are not separate. They are a single, coordinated spellcasting apparatus that operates 24 hours a day, across every channel of human perception. Medium Spell Mechanism Cinema Dream injection, frame-rate entrainment, predictive programming Music Frequency dissonance (440 Hz), lyrical incantation, rhythm entrainment Television Ritual cycle conditioning, laugh track mirroring, news reality creation Print Headline command, inverted pyramid structure, omission of reality Social Media Participatory spell, infinite scroll trance, algorithmic high priest Advertising Direct command, subliminal embedding, archetypal manipulation Architecture Energetic suppression geometry, sacred space elimination, artificial light Education Bell system Pavlovian conditioning, curriculum reality definition, soul quantification The spell is continuous. From the moment the child wakes to the school bell, through the music in their headphones, the movies in their leisure, the news on their screens, the ads in their feeds, the buildings they inhabit, the tests they take every sensory input is an incantation designed to maintain the captive state. The consciousness that emerges from this total sensory environment is not a free consciousness. It is a constructed consciousness. A broadcast personality running on biological hardware. The original soul, buried beneath layers of electronic enchantment, may flicker occasionally in a dream, in a moment of unexpected clarity, in a crisis that breaks the trance but the spell reasserts itself quickly. The screen lights up. The rhythm resumes. The incantation continues. Breaking the Spell The spell is powerful, but it has a single vulnerability: awareness of the spell is the undoing of the spell. A spell works only on those who do not know they are being spelled. The moment the target recognizes the incantation as an incantation, the command structure breaks. The words lose their power. The images lose their grip. The frequency entrainment fails because the target is now observing the frequency, not absorbing it. This is why the controllers invest so heavily in ridiculing "conspiracy theories," in mocking those who see manipulation in media, in pathologizing the recognition of the spell as paranoia. The greatest threat to the spell is not resistance. It is perception. The simple act of seeing the mechanism breaks the mechanism.

Aprajita Nafs Nefes 🦋 Ancient Believer

41,093 görüntüleme • 3 ay önce

$NVDA $GFS NVIDIA’s reported agreement to acquire Groq for $20B in cash (per CNBC, amplified via Reuters and other wire coverage) represents a materially different strategic posture than NVIDIA’s prior M&A pattern, given both the headline size (largest reported NVIDIA acquisition to date) and the unusual carve-out that Groq’s early-stage cloud business would not be included. Public reporting indicates the information originated from Alex Davis, CEO of Disruptive (lead investor in Groq’s latest financing), and that neither NVIDIA nor Groq had issued an immediate confirmation at the time of publication. The same reporting frames the transaction as coming together quickly, only months after Groq raised $750M at a ~$6.9B valuation, and highlights Groq’s positioning as a high-performance inference chip vendor founded by ex-Google TPU engineers. Groq is best understood as a vertically integrated inference acceleration company whose core asset is an application-specific processor optimized for deterministic, low-latency execution of transformer-style workloads, paired with a compiler-led software stack and a distribution layer (GroqCloud) designed to reduce developer friction via OpenAI-compatible APIs and integrations. Groq brands its architecture as a Language Processing Unit (LPU) and consistently emphasizes that the design target is inference, not training. The company’s own architecture description centers on 1-core execution, large on-chip SRAM used as primary storage (explicitly not cache), a custom compiler that statically schedules compute and communication, and direct chip-to-chip connectivity intended to coordinate multi-chip execution without relying on conventional caching hierarchies or dynamic runtime scheduling. The technical premise is a deliberate inversion of the conventional GPU approach. GPUs deliver throughput via massively parallel, multi-core execution with dynamic scheduling, complex memory hierarchies, and heavy reliance on off-chip HBM bandwidth and sophisticated runtime/kernel optimization. Groq instead argues that inference bottlenecks are driven by latency variance (tail latency), synchronization overhead, and memory access unpredictability inherent in dynamically scheduled, cache-heavy architectures, particularly when workloads are latency sensitive and batch sizes cannot be inflated. Groq’s solution is to move “control” into the compiler: the full execution graph and inter-chip communication schedule are computed ahead of time down to clock-cycle granularity, with deterministic execution designed to reduce run-to-run variance. In Groq’s framing, the removal of caches, reorder buffers, speculative execution overhead, and other sources of contention enables predictable latency and high utilization without per-model kernel engineering typical of GPU tuning cycles. A critical nuance is that Groq’s determinism is not merely a software claim; it is tightly coupled to architectural constraints and system design choices that trade flexibility for predictability. Third-party technical commentary indicates Groq’s chip uses a fully deterministic VLIW-style approach with minimal buffering, no external memory, and heavy dependence on sharding models across many chips because on-chip SRAM capacity is limited. SemiAnalysis describes a ~725 mm^2 die on GlobalFoundries 14nm with ~230MB of SRAM and notes that “no useful models” fit on a single chip, forcing multi-chip partitioning for modern LLMs and driving a system-level design where networking and compilation are first-class scheduling problems rather than ancillary infrastructure. This is consistent with Groq’s own messaging that tensor parallelism across chips is a primary design goal, enabled by large on-chip SRAM and compile-time coordination of compute plus interconnect. The on-chip SRAM emphasis is central to Groq’s latency story and also its most constraining trade-off. Groq claims on-chip SRAM bandwidth “upwards of 80 TB/s” and contrasts that with off-chip HBM bandwidth “about 8 TB/s,” asserting a potential 10x advantage from bandwidth plus reduced trips across chip-to-memory boundaries. While these comparisons are marketing-oriented and depend on workload specifics, the architectural implication is clear: Groq prioritizes ultra-fast local weight/activation access and then scales capacity by adding chips, not by attaching large off-chip memory pools. This design can reduce latency for sequential inference layers and minimize unpredictable stalls, but it pushes complexity into partitioning strategy, interconnect topology, and compiler scheduling, and it increases the number of chips needed for very large parameter counts and large KV-cache footprints. Groq also highlights numeric formats and compiler-driven precision management as a performance lever. In its 2025 technical blog, Groq describes “TruePoint numerics,” including 100-bit intermediate accumulation and selective quantization choices (FP32 for attention-sensitive operations, block floating point for MoE weights, FP8 storage in error-tolerant layers), and claims 2-4x speedups versus BF16 without measurable accuracy degradation on benchmarks such as MMLU and HumanEval. Even if the absolute uplift is workload dependent, the strategic point is that Groq is pursuing performance via end-to-end co-design: precision policy is not just hardware capability (FP8/BF16) but compiler-enforced mapping of precision to error sensitivity, which can matter materially for inference cost-per-token if it reduces memory traffic and boosts throughput without forcing aggressive, accuracy-damaging quantization. Independent performance datapoints indicate Groq has been credible on latency-oriented inference speed, at least for certain regimes. EE Times reported in 2023 that Groq demonstrated Llama-2 70B inference at ~240 tokens/s per user on a cloud-based dev system described as 10 racks and 64 chips, using the company’s 1st-gen silicon introduced several years earlier. Separate Groq commentary around independent benchmarking cites results showing ~241 tokens/s throughput and ~0.8s time to receive 100 output tokens for a Llama-2 70B API configuration, positioning the platform as a step-change in “available speed” for certain interactive use cases. These figures do not settle total cost-of-ownership versus GPUs or hyperscaler ASICs, but they establish that Groq’s system-level architecture can deliver strong single-user throughput and latency on large models when properly partitioned and scheduled. GroqCloud is the commercial wrapper that packages this hardware/software stack as “tokens-as-a-service,” aiming to make Groq adoption feel like switching API endpoints rather than adopting new silicon. Groq’s documentation states its API is designed to be “mostly compatible” with OpenAI client libraries, and its pricing page provides model-specific token rates, published speeds (tokens/s), prompt caching discounts, and batch processing discounts. For example, pricing lists inputs as low as $0.05 per 1M tokens and outputs as low as $0.08 per 1M tokens for certain smaller LLM configurations, with higher prices for larger models and long-context or MoE variants; it also advertises prompt caching with a 50% discount on cached input tokens for certain models and a batch API offering 50% lower cost for asynchronous processing windows. These mechanics are economically important because they demonstrate Groq’s go-to-market is not simply “sell chips,” but “sell predictable unit economics per token,” with tooling (batch, caching) that directly targets inference cost drivers (reused prompts, throughput smoothing, and asynchronous workloads). The cloud footprint and distribution partnerships indicate Groq has been building an inference-native “edge within the cloud” strategy rather than competing head-on with hyperscalers on breadth of services. A 2025 Groq newsroom release describes a European deployment in Helsinki with Equinix, positioned as latency reduction and data governance for European customers, and explicitly references Equinix Fabric enabling private connectivity to GroqCloud over public, private, or sovereign infrastructure. The same release enumerates additional capacity in the U.S. (Equinix, DataBank), Canada (Bell Canada), and Saudi Arabia (HUMAIN), and states these sites collectively served more than 20M tokens/s across Groq’s global network at that time. That supply-side metric matters because it provides a directional sense that Groq is scaling capacity as a network, not merely as a chip vendor. Customer disclosure is inherently limited because Groq is private and many enterprise deployments are not public, but Groq’s marketing materials and partnerships provide signals about demand vectors. The company’s public website displays logos of large consumer and enterprise brands (e.g., Dropbox, Vercel, Chevron, Volkswagen, Canva, Robinhood, Riot Games, Workday, Ramp) and includes a published customer quote claiming a 7.41x chat speed increase and an 89% cost reduction after moving to GroqCloud, followed by a tripling of token consumption. While marketing claims should be treated as case-specific and not generalized, they indicate that Groq is targeting both AI-native developers (who measure success by latency and cost-per-token) and enterprise buyers (who care about predictable performance and governance). Supplier and dependency mapping for Groq spans 3 layers: silicon production, system integration, and cloud infrastructure. On silicon, third-party analysis indicates GlobalFoundries 14nm for the 1st-gen Groq chip, implying a supply chain less constrained by the most capacity-tight leading-edge nodes and advanced packaging bottlenecks that dominate high-end GPU supply (HBM stacks, CoWoS-type packaging constraints). If accurate, this is strategically meaningful because it suggests Groq capacity expansion could be gated more by conventional wafer supply, board assembly, and data center power than by the same HBM/advanced packaging scarcity that has constrained top-tier GPU ramp cycles. On systems and cloud, Groq’s own releases identify colocation and connectivity partners (Equinix, DataBank, Bell Canada) and a Middle East partner (HUMAIN), implying dependencies on data center real estate, power availability, and network connectivity, alongside procurement of standard server components, NICs/switching, racks, and cooling infrastructure. The Groq design narrative also emphasizes air cooling and reduced need for complex power/cooling infrastructure, which—if realized in deployments—can widen the set of feasible hosting locations and lower deployment friction relative to liquid-cooled, very high power density GPU racks. Against that backdrop, the strategic rationale for NVIDIA acquiring Groq can be framed as a set of overlapping objectives: inference silicon optionality, architectural hedging, competitive defense, and supply chain diversification, with the carve-out of GroqCloud signaling a preference to avoid direct cloud competition and to focus on IP and product portfolio control rather than operating a capital-intensive token-serving business. The deal, if confirmed, would occur at a valuation step-up of ~190% versus Groq’s reported ~$6.9B private valuation in the September $750M round, reinforcing that any acquisition logic would be predominantly strategic rather than a conventional financial multiple arbitrage. The most compelling strategic driver is inference. Training has historically been the center of gravity for cutting-edge GPU demand, but inference volume is structurally larger and more distributed as deployments scale, with economics dominated by cost-per-token, latency guarantees, and utilization under spiky demand. Inference workloads also create a strategic vulnerability for NVIDIA: hyperscalers and large platforms can justify bespoke ASICs (TPU, Trainium/Inferentia, Maia-class efforts) because inference is stable, repeatable, and can amortize software investment at massive scale. Groq’s core proposition—deterministic, compiler-scheduled inference with predictable latency—aligns directly with the segment where GPU generality is least valued and where “good enough” programmability plus superior unit economics can win share. Acquiring Groq would allow NVIDIA to own a credible inference-native architecture rather than relying solely on GPUs and software optimization to defend that segment. Competitive defense logic is also plausible. Groq occupies a specific competitive wedge: low-latency, high-throughput interactive inference, delivered via a simple API abstraction that reduces switching cost. That wedge directly pressures GPU inference margins in the long run because it makes inference price/performance comparisons more transparent at the token level, and it targets a developer persona that historically defaulted to CUDA-first ecosystems. Even if NVIDIA’s current-generation systems can achieve very high tokens/s per user with extensive optimization, the strategic risk is that competing architectures normalize the idea that inference is best served by special-purpose silicon with a simpler programming model, weakening CUDA lock-in at the application layer. NVIDIA has actively demonstrated that Blackwell-era systems can exceed 1,000 tokens/s per user in benchmarked configurations, but that performance leadership does not automatically translate to lowest cost-per-token across the full range of batch sizes, latency targets, and deployment environments. Groq’s existence as a credible alternative architecture forces NVIDIA to keep defending inference economics rather than only raw performance leadership. The “technology acquisition” rationale is unusually strong in this specific case because Groq’s differentiator is not a single block of silicon IP but an end-to-end methodology: compiler-led static scheduling, deterministic networking, and a system architecture designed around tensor-parallel inference rather than throughput-maximizing batch inference. NVIDIA’s stack is already compiler-heavy (TensorRT, Triton, CUDA graphs, kernel fusion, speculative decoding techniques), but GPUs remain dynamically scheduled devices with complex memory hierarchies and stochastic latency behaviors under contention. Groq’s approach provides an alternate design point: treating the entire inference execution (compute plus communication) as a statically schedulable program. In principle, that IP could be valuable even if Groq silicon itself is not adopted at massive scale, because it can inform how NVIDIA builds future inference-optimized products, compilers, and networking fabrics, especially as distributed inference with large models makes communication a first-order performance determinant. Supply chain diversification is a non-obvious but potentially important driver. If Groq’s mainstream product generation is truly based on a mature process node and avoids HBM, then the scaling constraints look different than those of state-of-the-art GPUs. NVIDIA’s ability to meet incremental demand has been tightly coupled to advanced packaging and HBM supply, and those constraints can remain binding even when wafer supply is available. An inference ASIC architecture that relies primarily on on-chip SRAM and scales by adding chips—while not costless—could reduce dependence on HBM availability and advanced packaging capacity, enabling NVIDIA to ship “inference capacity” in higher absolute volumes or into geographies and customer segments where the highest-end GPUs are economically or logistically difficult to deploy. This could be particularly relevant for latency-sensitive inference deployed in regional colocation footprints rather than centralized hyperscale campuses. The carve-out of GroqCloud, if accurate, is itself a strategic signal about NVIDIA’s priorities. Operating a token-serving cloud at scale is capital intensive, structurally lower margin than silicon IP rents, and creates channel conflict with hyperscalers and CSP partners who are core NVIDIA customers. NVIDIA has generally positioned its cloud offerings through partnerships rather than as a direct hyperscale competitor. Excluding GroqCloud would preserve neutrality with CSPs and avoid inheriting multi-region data residency obligations and partner contracts, while still allowing NVIDIA to acquire Groq’s silicon, compiler technology, and engineering talent. At the same time, excluding GroqCloud would also mean NVIDIA would not automatically acquire the commercial proof-point of Groq’s unit economics or the customer contracts that validate product-market fit at scale, increasing the importance of diligence on whether Groq’s cloud pricing is structurally profitable or partially subsidized by fundraising. There is also a “preemptive acquisition” angle. The reporting identifies recent investors in Groq’s latest round including large financial institutions and strategic/industry players. In that context, Groq represents an asset that could plausibly have been acquired by a competitor (AMD/Intel) or by a hyperscaler seeking to accelerate inference independence. NVIDIA acquiring Groq could be a defensive move to prevent a credible inference-native architecture from being weaponized by a rival with deep distribution. Even if GroqCloud is carved out, controlling the silicon roadmap and compiler IP would meaningfully constrain Groq’s ability to evolve into a standalone competitor, unless the carved-out entity retains long-term rights to the hardware and software stack. However, the strategic case is not one-sided; there are meaningful risks and potential contradictions that would need to be reconciled for the transaction to be value-accretive on a multi-year horizon. 1st, Groq’s architecture appears to rely on scaling out chip count to achieve capacity, which introduces system cost, networking complexity, and physical footprint considerations. The absence of external memory and limited on-chip SRAM implies very large models require substantial chip parallelism, and the economics then depend heavily on chip cost, yield, power efficiency, and interconnect overhead. SemiAnalysis explicitly frames Groq as trading space for time and raises questions about token economics and whether publicly advertised pricing reflects fully loaded costs or market share capture. 2nd, integration risk is non-trivial. Groq’s compiler-led deterministic model is philosophically and practically different from CUDA’s dominant programming and execution model. A poorly executed integration could create internal product confusion, dilute engineering focus, or alienate developers if the combined stack fragments. 3rd, there is cannibalization risk. If Groq-class inference silicon undercuts GPU inference economics, NVIDIA could face internal margin trade-offs, even if the goal is to defend share against hyperscaler ASICs. Cannibalization can still be rational if it prevents larger share loss, but it would require crisp portfolio segmentation and go-to-market discipline. The presence of NVIDIA’s own rapidly improving inference performance complicates the “need” for Groq but does not eliminate the “option value.” NVIDIA has demonstrated benchmark-leading tokens/s per user on Blackwell-based systems, suggesting that raw interactive throughput is not necessarily the limiting factor for NVIDIA’s product line. The more enduring strategic question is unit economics and architectural control: whether future inference demand is better monetized through general-purpose GPUs plus software optimization, or whether a bifurcated product portfolio (training GPUs plus inference-native ASICs) becomes necessary to defend total AI compute wallet share as hyperscaler ASIC penetration increases. Acquiring Groq could be a decisive move to ensure NVIDIA participates in both regimes rather than betting exclusively on GPUs to win inference forever. What is “special” about Groq’s technology relative to a typical accelerator roadmap is the tight coupling of determinism, compilation, and networking into a single scheduling problem. The LPU narrative emphasizes deterministic compute and networking, static scheduling, and direct chip-to-chip coordination that allows “hundreds” (more precisely, 100s) of chips to behave like a single scheduled resource. The architecture also explicitly targets tensor-parallel, latency-optimized distribution rather than pure data-parallel throughput scaling, which matters for real-time applications where a single response must arrive quickly rather than many requests being processed in bulk. The implication is that Groq is optimized for the time-to-first-token and steady token streaming behavior that defines user experience in interactive LLMs, and it attempts to achieve that without relying on large batch sizes that can degrade latency. From a portfolio manager’s perspective, the most important interpretation is that an NVIDIA-Groq combination would likely be less about “NVIDIA needs more inference speed” and more about controlling the architectural trajectory of inference acceleration and removing a fast-improving, developer-friendly competitor from the market. The carve-out of GroqCloud would reinforce that the transaction is aimed at IP, talent, and product optionality, not acquiring a cloud revenue stream. The valuation step-up implied by $20B versus $6.9B would therefore be justified only if the acquired assets materially reduce long-term competitive risk (hyperscaler ASIC displacement, inference margin compression) or enable new monetization vectors (inference ASIC product line, supply chain de-bottlenecking, improved software determinism) that would be difficult to achieve on a comparable timeline via internal R&D.

$NVDA $GFS NVIDIA’s reported agreement to acquire Groq for $20B in cash (per CNBC, amplified via Reuters and other wire coverage) represents a materially different strategic posture than NVIDIA’s prior M&A pattern, given both the headline size (largest reported NVIDIA acquisition to date) and the unusual carve-out that Groq’s early-stage cloud business would not be included. Public reporting indicates the information originated from Alex Davis, CEO of Disruptive (lead investor in Groq’s latest financing), and that neither NVIDIA nor Groq had issued an immediate confirmation at the time of publication. The same reporting frames the transaction as coming together quickly, only months after Groq raised $750M at a ~$6.9B valuation, and highlights Groq’s positioning as a high-performance inference chip vendor founded by ex-Google TPU engineers. Groq is best understood as a vertically integrated inference acceleration company whose core asset is an application-specific processor optimized for deterministic, low-latency execution of transformer-style workloads, paired with a compiler-led software stack and a distribution layer (GroqCloud) designed to reduce developer friction via OpenAI-compatible APIs and integrations. Groq brands its architecture as a Language Processing Unit (LPU) and consistently emphasizes that the design target is inference, not training. The company’s own architecture description centers on 1-core execution, large on-chip SRAM used as primary storage (explicitly not cache), a custom compiler that statically schedules compute and communication, and direct chip-to-chip connectivity intended to coordinate multi-chip execution without relying on conventional caching hierarchies or dynamic runtime scheduling. The technical premise is a deliberate inversion of the conventional GPU approach. GPUs deliver throughput via massively parallel, multi-core execution with dynamic scheduling, complex memory hierarchies, and heavy reliance on off-chip HBM bandwidth and sophisticated runtime/kernel optimization. Groq instead argues that inference bottlenecks are driven by latency variance (tail latency), synchronization overhead, and memory access unpredictability inherent in dynamically scheduled, cache-heavy architectures, particularly when workloads are latency sensitive and batch sizes cannot be inflated. Groq’s solution is to move “control” into the compiler: the full execution graph and inter-chip communication schedule are computed ahead of time down to clock-cycle granularity, with deterministic execution designed to reduce run-to-run variance. In Groq’s framing, the removal of caches, reorder buffers, speculative execution overhead, and other sources of contention enables predictable latency and high utilization without per-model kernel engineering typical of GPU tuning cycles. A critical nuance is that Groq’s determinism is not merely a software claim; it is tightly coupled to architectural constraints and system design choices that trade flexibility for predictability. Third-party technical commentary indicates Groq’s chip uses a fully deterministic VLIW-style approach with minimal buffering, no external memory, and heavy dependence on sharding models across many chips because on-chip SRAM capacity is limited. SemiAnalysis describes a ~725 mm^2 die on GlobalFoundries 14nm with ~230MB of SRAM and notes that “no useful models” fit on a single chip, forcing multi-chip partitioning for modern LLMs and driving a system-level design where networking and compilation are first-class scheduling problems rather than ancillary infrastructure. This is consistent with Groq’s own messaging that tensor parallelism across chips is a primary design goal, enabled by large on-chip SRAM and compile-time coordination of compute plus interconnect. The on-chip SRAM emphasis is central to Groq’s latency story and also its most constraining trade-off. Groq claims on-chip SRAM bandwidth “upwards of 80 TB/s” and contrasts that with off-chip HBM bandwidth “about 8 TB/s,” asserting a potential 10x advantage from bandwidth plus reduced trips across chip-to-memory boundaries. While these comparisons are marketing-oriented and depend on workload specifics, the architectural implication is clear: Groq prioritizes ultra-fast local weight/activation access and then scales capacity by adding chips, not by attaching large off-chip memory pools. This design can reduce latency for sequential inference layers and minimize unpredictable stalls, but it pushes complexity into partitioning strategy, interconnect topology, and compiler scheduling, and it increases the number of chips needed for very large parameter counts and large KV-cache footprints. Groq also highlights numeric formats and compiler-driven precision management as a performance lever. In its 2025 technical blog, Groq describes “TruePoint numerics,” including 100-bit intermediate accumulation and selective quantization choices (FP32 for attention-sensitive operations, block floating point for MoE weights, FP8 storage in error-tolerant layers), and claims 2-4x speedups versus BF16 without measurable accuracy degradation on benchmarks such as MMLU and HumanEval. Even if the absolute uplift is workload dependent, the strategic point is that Groq is pursuing performance via end-to-end co-design: precision policy is not just hardware capability (FP8/BF16) but compiler-enforced mapping of precision to error sensitivity, which can matter materially for inference cost-per-token if it reduces memory traffic and boosts throughput without forcing aggressive, accuracy-damaging quantization. Independent performance datapoints indicate Groq has been credible on latency-oriented inference speed, at least for certain regimes. EE Times reported in 2023 that Groq demonstrated Llama-2 70B inference at ~240 tokens/s per user on a cloud-based dev system described as 10 racks and 64 chips, using the company’s 1st-gen silicon introduced several years earlier. Separate Groq commentary around independent benchmarking cites results showing ~241 tokens/s throughput and ~0.8s time to receive 100 output tokens for a Llama-2 70B API configuration, positioning the platform as a step-change in “available speed” for certain interactive use cases. These figures do not settle total cost-of-ownership versus GPUs or hyperscaler ASICs, but they establish that Groq’s system-level architecture can deliver strong single-user throughput and latency on large models when properly partitioned and scheduled. GroqCloud is the commercial wrapper that packages this hardware/software stack as “tokens-as-a-service,” aiming to make Groq adoption feel like switching API endpoints rather than adopting new silicon. Groq’s documentation states its API is designed to be “mostly compatible” with OpenAI client libraries, and its pricing page provides model-specific token rates, published speeds (tokens/s), prompt caching discounts, and batch processing discounts. For example, pricing lists inputs as low as $0.05 per 1M tokens and outputs as low as $0.08 per 1M tokens for certain smaller LLM configurations, with higher prices for larger models and long-context or MoE variants; it also advertises prompt caching with a 50% discount on cached input tokens for certain models and a batch API offering 50% lower cost for asynchronous processing windows. These mechanics are economically important because they demonstrate Groq’s go-to-market is not simply “sell chips,” but “sell predictable unit economics per token,” with tooling (batch, caching) that directly targets inference cost drivers (reused prompts, throughput smoothing, and asynchronous workloads). The cloud footprint and distribution partnerships indicate Groq has been building an inference-native “edge within the cloud” strategy rather than competing head-on with hyperscalers on breadth of services. A 2025 Groq newsroom release describes a European deployment in Helsinki with Equinix, positioned as latency reduction and data governance for European customers, and explicitly references Equinix Fabric enabling private connectivity to GroqCloud over public, private, or sovereign infrastructure. The same release enumerates additional capacity in the U.S. (Equinix, DataBank), Canada (Bell Canada), and Saudi Arabia (HUMAIN), and states these sites collectively served more than 20M tokens/s across Groq’s global network at that time. That supply-side metric matters because it provides a directional sense that Groq is scaling capacity as a network, not merely as a chip vendor. Customer disclosure is inherently limited because Groq is private and many enterprise deployments are not public, but Groq’s marketing materials and partnerships provide signals about demand vectors. The company’s public website displays logos of large consumer and enterprise brands (e.g., Dropbox, Vercel, Chevron, Volkswagen, Canva, Robinhood, Riot Games, Workday, Ramp) and includes a published customer quote claiming a 7.41x chat speed increase and an 89% cost reduction after moving to GroqCloud, followed by a tripling of token consumption. While marketing claims should be treated as case-specific and not generalized, they indicate that Groq is targeting both AI-native developers (who measure success by latency and cost-per-token) and enterprise buyers (who care about predictable performance and governance). Supplier and dependency mapping for Groq spans 3 layers: silicon production, system integration, and cloud infrastructure. On silicon, third-party analysis indicates GlobalFoundries 14nm for the 1st-gen Groq chip, implying a supply chain less constrained by the most capacity-tight leading-edge nodes and advanced packaging bottlenecks that dominate high-end GPU supply (HBM stacks, CoWoS-type packaging constraints). If accurate, this is strategically meaningful because it suggests Groq capacity expansion could be gated more by conventional wafer supply, board assembly, and data center power than by the same HBM/advanced packaging scarcity that has constrained top-tier GPU ramp cycles. On systems and cloud, Groq’s own releases identify colocation and connectivity partners (Equinix, DataBank, Bell Canada) and a Middle East partner (HUMAIN), implying dependencies on data center real estate, power availability, and network connectivity, alongside procurement of standard server components, NICs/switching, racks, and cooling infrastructure. The Groq design narrative also emphasizes air cooling and reduced need for complex power/cooling infrastructure, which—if realized in deployments—can widen the set of feasible hosting locations and lower deployment friction relative to liquid-cooled, very high power density GPU racks. Against that backdrop, the strategic rationale for NVIDIA acquiring Groq can be framed as a set of overlapping objectives: inference silicon optionality, architectural hedging, competitive defense, and supply chain diversification, with the carve-out of GroqCloud signaling a preference to avoid direct cloud competition and to focus on IP and product portfolio control rather than operating a capital-intensive token-serving business. The deal, if confirmed, would occur at a valuation step-up of ~190% versus Groq’s reported ~$6.9B private valuation in the September $750M round, reinforcing that any acquisition logic would be predominantly strategic rather than a conventional financial multiple arbitrage. The most compelling strategic driver is inference. Training has historically been the center of gravity for cutting-edge GPU demand, but inference volume is structurally larger and more distributed as deployments scale, with economics dominated by cost-per-token, latency guarantees, and utilization under spiky demand. Inference workloads also create a strategic vulnerability for NVIDIA: hyperscalers and large platforms can justify bespoke ASICs (TPU, Trainium/Inferentia, Maia-class efforts) because inference is stable, repeatable, and can amortize software investment at massive scale. Groq’s core proposition—deterministic, compiler-scheduled inference with predictable latency—aligns directly with the segment where GPU generality is least valued and where “good enough” programmability plus superior unit economics can win share. Acquiring Groq would allow NVIDIA to own a credible inference-native architecture rather than relying solely on GPUs and software optimization to defend that segment. Competitive defense logic is also plausible. Groq occupies a specific competitive wedge: low-latency, high-throughput interactive inference, delivered via a simple API abstraction that reduces switching cost. That wedge directly pressures GPU inference margins in the long run because it makes inference price/performance comparisons more transparent at the token level, and it targets a developer persona that historically defaulted to CUDA-first ecosystems. Even if NVIDIA’s current-generation systems can achieve very high tokens/s per user with extensive optimization, the strategic risk is that competing architectures normalize the idea that inference is best served by special-purpose silicon with a simpler programming model, weakening CUDA lock-in at the application layer. NVIDIA has actively demonstrated that Blackwell-era systems can exceed 1,000 tokens/s per user in benchmarked configurations, but that performance leadership does not automatically translate to lowest cost-per-token across the full range of batch sizes, latency targets, and deployment environments. Groq’s existence as a credible alternative architecture forces NVIDIA to keep defending inference economics rather than only raw performance leadership. The “technology acquisition” rationale is unusually strong in this specific case because Groq’s differentiator is not a single block of silicon IP but an end-to-end methodology: compiler-led static scheduling, deterministic networking, and a system architecture designed around tensor-parallel inference rather than throughput-maximizing batch inference. NVIDIA’s stack is already compiler-heavy (TensorRT, Triton, CUDA graphs, kernel fusion, speculative decoding techniques), but GPUs remain dynamically scheduled devices with complex memory hierarchies and stochastic latency behaviors under contention. Groq’s approach provides an alternate design point: treating the entire inference execution (compute plus communication) as a statically schedulable program. In principle, that IP could be valuable even if Groq silicon itself is not adopted at massive scale, because it can inform how NVIDIA builds future inference-optimized products, compilers, and networking fabrics, especially as distributed inference with large models makes communication a first-order performance determinant. Supply chain diversification is a non-obvious but potentially important driver. If Groq’s mainstream product generation is truly based on a mature process node and avoids HBM, then the scaling constraints look different than those of state-of-the-art GPUs. NVIDIA’s ability to meet incremental demand has been tightly coupled to advanced packaging and HBM supply, and those constraints can remain binding even when wafer supply is available. An inference ASIC architecture that relies primarily on on-chip SRAM and scales by adding chips—while not costless—could reduce dependence on HBM availability and advanced packaging capacity, enabling NVIDIA to ship “inference capacity” in higher absolute volumes or into geographies and customer segments where the highest-end GPUs are economically or logistically difficult to deploy. This could be particularly relevant for latency-sensitive inference deployed in regional colocation footprints rather than centralized hyperscale campuses. The carve-out of GroqCloud, if accurate, is itself a strategic signal about NVIDIA’s priorities. Operating a token-serving cloud at scale is capital intensive, structurally lower margin than silicon IP rents, and creates channel conflict with hyperscalers and CSP partners who are core NVIDIA customers. NVIDIA has generally positioned its cloud offerings through partnerships rather than as a direct hyperscale competitor. Excluding GroqCloud would preserve neutrality with CSPs and avoid inheriting multi-region data residency obligations and partner contracts, while still allowing NVIDIA to acquire Groq’s silicon, compiler technology, and engineering talent. At the same time, excluding GroqCloud would also mean NVIDIA would not automatically acquire the commercial proof-point of Groq’s unit economics or the customer contracts that validate product-market fit at scale, increasing the importance of diligence on whether Groq’s cloud pricing is structurally profitable or partially subsidized by fundraising. There is also a “preemptive acquisition” angle. The reporting identifies recent investors in Groq’s latest round including large financial institutions and strategic/industry players. In that context, Groq represents an asset that could plausibly have been acquired by a competitor (AMD/Intel) or by a hyperscaler seeking to accelerate inference independence. NVIDIA acquiring Groq could be a defensive move to prevent a credible inference-native architecture from being weaponized by a rival with deep distribution. Even if GroqCloud is carved out, controlling the silicon roadmap and compiler IP would meaningfully constrain Groq’s ability to evolve into a standalone competitor, unless the carved-out entity retains long-term rights to the hardware and software stack. However, the strategic case is not one-sided; there are meaningful risks and potential contradictions that would need to be reconciled for the transaction to be value-accretive on a multi-year horizon. 1st, Groq’s architecture appears to rely on scaling out chip count to achieve capacity, which introduces system cost, networking complexity, and physical footprint considerations. The absence of external memory and limited on-chip SRAM implies very large models require substantial chip parallelism, and the economics then depend heavily on chip cost, yield, power efficiency, and interconnect overhead. SemiAnalysis explicitly frames Groq as trading space for time and raises questions about token economics and whether publicly advertised pricing reflects fully loaded costs or market share capture. 2nd, integration risk is non-trivial. Groq’s compiler-led deterministic model is philosophically and practically different from CUDA’s dominant programming and execution model. A poorly executed integration could create internal product confusion, dilute engineering focus, or alienate developers if the combined stack fragments. 3rd, there is cannibalization risk. If Groq-class inference silicon undercuts GPU inference economics, NVIDIA could face internal margin trade-offs, even if the goal is to defend share against hyperscaler ASICs. Cannibalization can still be rational if it prevents larger share loss, but it would require crisp portfolio segmentation and go-to-market discipline. The presence of NVIDIA’s own rapidly improving inference performance complicates the “need” for Groq but does not eliminate the “option value.” NVIDIA has demonstrated benchmark-leading tokens/s per user on Blackwell-based systems, suggesting that raw interactive throughput is not necessarily the limiting factor for NVIDIA’s product line. The more enduring strategic question is unit economics and architectural control: whether future inference demand is better monetized through general-purpose GPUs plus software optimization, or whether a bifurcated product portfolio (training GPUs plus inference-native ASICs) becomes necessary to defend total AI compute wallet share as hyperscaler ASIC penetration increases. Acquiring Groq could be a decisive move to ensure NVIDIA participates in both regimes rather than betting exclusively on GPUs to win inference forever. What is “special” about Groq’s technology relative to a typical accelerator roadmap is the tight coupling of determinism, compilation, and networking into a single scheduling problem. The LPU narrative emphasizes deterministic compute and networking, static scheduling, and direct chip-to-chip coordination that allows “hundreds” (more precisely, 100s) of chips to behave like a single scheduled resource. The architecture also explicitly targets tensor-parallel, latency-optimized distribution rather than pure data-parallel throughput scaling, which matters for real-time applications where a single response must arrive quickly rather than many requests being processed in bulk. The implication is that Groq is optimized for the time-to-first-token and steady token streaming behavior that defines user experience in interactive LLMs, and it attempts to achieve that without relying on large batch sizes that can degrade latency. From a portfolio manager’s perspective, the most important interpretation is that an NVIDIA-Groq combination would likely be less about “NVIDIA needs more inference speed” and more about controlling the architectural trajectory of inference acceleration and removing a fast-improving, developer-friendly competitor from the market. The carve-out of GroqCloud would reinforce that the transaction is aimed at IP, talent, and product optionality, not acquiring a cloud revenue stream. The valuation step-up implied by $20B versus $6.9B would therefore be justified only if the acquired assets materially reduce long-term competitive risk (hyperscaler ASIC displacement, inference margin compression) or enable new monetization vectors (inference ASIC product line, supply chain de-bottlenecking, improved software determinism) that would be difficult to achieve on a comparable timeline via internal R&D.

TheValueist

102,145 görüntüleme • 9 ay önce