Video wird geladen...

Video konnte nicht geladen werden

Zur Startseite

EMPULSE · Fast-paced 6v6 movement shooter from 1047 Games Splitgate · Spiritual successor to Titanfall: fluid momentum, wall-running, grappling hooks, Holojumps · Gunplay feels like Halo, map-spawned pilotable mechs act as high-value objectives · P.A.I.N.T. Bombs: alter the environment for tactical advantages · Early Access on Steam, PlayStation 5,...

141,648 Aufrufe • vor 2 Monaten •via X (Twitter)

0 Kommentare

Keine Kommentare verfügbar

Kommentare vom Original-Post werden hier angezeigt

Ähnliche Videos

Update 8 is one of the biggest yet for MindsEye, with the addition of multiplayer and a new mode: Grenade Football. It also deepens the single-player experience and expands the possibilities of player creation through new features and improvements. With this summer’s football fever still running high, we couldn't sit out. Grenade Football is a fast-paced multiplayer mode that gives the classic game an explosive twist. Players swap their boots for launchers and bring the firepower to the pitch, where teamwork, timing, and skill decide the winner. Finding your squad is easy. New social features let you connect with friends, create parties, communicate through voice and text chat, and take on multiplayer experiences together across PlayStation 5, Xbox Series X|S, and PC. Multiplayer supports full cross-play: one shared player pool across every platform. For solo players, the Silva E-Series: Summer Circuits Time Trial delivers a new challenge to master. Perfect your racing line, chase the fastest lap, and climb the global leaderboards. A milestone we have been building toward: ARCADIA’s player-created content arrives on consoles for the first time. Every game and experience the PC community has built is now in the hands of PlayStation and Xbox players. As always, this update includes continued polish and refinements across gameplay, audio, visuals, and stability. Please continue sharing your feedback through the usual channels. detailed patch notes on steam and discord. Thanks, The MindsEye Team

MindsEye

151,268 Aufrufe • vor 1 Monat

Halloween The Game Gets an Extended Preview During Sony's State of Play! Release date announced! COMING SEPTEMBER 8th, 2026! Step into the chilling world of John Carpenter’s genre-defining film, now transformed into a suspenseful one-versus-many stealth horror experience. Put on the iconic mask to become the ultimate slasher, Michael Myers, stalking and executing the citizens of Haddonfield one by one, or striving to thwart Michael Myers’ plans as Civilians determined to save the unaware townsfolk before it’s too late. Stick to the shadows as Civilians, seeking out Haddonfield residents to warn them and searching for a way to contact the authorities. As Michael Myers, give them a reason to fear the dark and cut the phone lines to prevent the police from ruining his favorite holiday. Whether playing solo in story mode, against bots offline, or facing others in online multiplayer, each mode rewards stealth, strategy, and skillful play. Staying true to the original film, IllFonic masterfully recreates the eerie atmosphere of Haddonfield across multiple maps and authentic locations. With a haunting ambience and score inspired by the legendary movie, Halloween brings the terror home in a new experience that will keep both old and new generations looking over their shoulders. “We are very excited to have partnered up with the great team at IllFonic to finally bring the definitive digital version of Halloween and Michael Myers to video game fans everywhere,” said Malek Akkad, President of Compass International Pictures, Inc. Halloween will unleash Michael Myers onto Xbox Series X|S, PlayStation 5, and PC via Steam and the Epic Games Store in 2026. Website: WISHLIST AT YOUR PREFERRED STORE FRONT - XBOX, PLAYSTATION, STEAM, and the EPIC GAMES STORE. Sign up for The Halloween Portal, the premier destination for fans featuring breaking news, exclusive content, and much more. #halloweenportal #michaelmyers #halloweenmovies #Halloweenmovie #thenighthecamehome #Halloween1978 #halloween

HalloweenMovies™

29,702 Aufrufe • vor 11 Monaten

Borderlands 4 is releasing in just 5 more days and we have another round of previews with GLOWING feedback, BL4 is shaping up to be a massive hit looter-shooter on the level of Borderlands 2 Key new info: - Gearbox describes this as a grounded sequel that builds on what worked while addressing past mistakes, abandoning the cringe tone of BL3 in favour of a more grounded tone - Gunplay feels extremely smooth, on par with the best of FPS games - Core premise remains: you’re a Vault Hunter chasing loot, chaos, and dopamine-filled explosions. - Four new Vault Hunters: Vex (Siren) Amon (Forgeknight) Harlowe (Gravitar) Rafa (Exo-Soldier) - Gameplay is reportedly fast, fluid, and smooth without being overwhelming. Sliding, dashing, grenades, and hovering are all new traversal options on top of the grapple hook - Gun variety is off the charts with guns like: Pistol firing exploding insects AR with an underbarrel shotgun Corrosive sniper rifle for poison builds Jakobs pistol that can also launch knives Cryo SMG for status builds - Combat is more punishing than previous games because shields break faster than expected - Enemies now hit harder, forcing good shield cooldown management - New Grapple hook tool to get to higher places and grab things, similar to Halo Infinite's grapple hook from the sounds of it. - Boss design has had a massive step up: Idolator Sol fight which previewers played through was nearly 10 minutes long with 3 health bars, it required dodging ground-shaking AoEs and arena-wide attacks plus it had unique mechanic where you have to catch Sol’s green spear with grappling hook to break his shield - Narrative touches: comms chatter from allies was well-written, adding context and personality without dragging, No more cringe inducing chatter like BL3 - The main issue noted by previewers was frame rate drops and stuttering on PC (Unreal Engine 5 optimization), Gearbox is promising a day one patch to improve things - Borderlands 4 "nails the fundamentals: smooth gunplay, crazy loot, deep skill trees, and chaotic fights." - Releases September 12th on PS5/PC/XSX/S, + Switch 2 version on October 3rd I will 100% be playing, covering and reviewing this at launch so be sure to follow to keep up with all of it!

Synth Potato🥔

333,378 Aufrufe • vor 11 Monaten

Borderlands 4 previews have dropped and they are insanely positive, It is being praised across the board as being a massive step up from Borderlands 3 and possibly the best BL to date! Here’s some highlights of the new info: - Borderlands 4 features a more grounded and darker story set on a dystopian world ruled by the Timekeeper - Humor is still present but more restrained to suit the darker narrative tone, Less cringe humor as in BL3 - The game feels like a soft reboot for the franchise, introducing new lore and gameplay structures - Movement has been completely overhauled with double-jumping, gliding, dashing, grappling hooks, and swimming - Four new Vault Hunters are available: Vex (Siren), Rafa (Gadget Soldier), Harlowe (Sniper), and Aton (Brawler) - Kairos, the new planet, is one seamless open world with no loading between zones - Split-screen co-op returns for local multiplayer - The game adopts a Destiny-like structure with side activities, dynamic events, and exploration hubs - Vehicles can be summoned anywhere on the map, making travel smoother and more immersive - The world of Kairos features dynamic weather systems and environmental variety - Combat pacing is faster with smarter, more aggressive enemies that emphasize dodging and movement - One boss previewed was the most complex and mechanics-rich in Borderlands history - Buildcrafting is significantly deeper with new enhancement gear slots and item augments - Weapon rarity is better balanced, legendary drops are rarer - Mission and boss replays are now built-in, allowing players to farm specific content - Online matchmaking is improved with simplified lobbies and clearer difficulty scaling - Gearbox aims to avoid turning Borderlands into a live-service GaaS model - Borderlands 4 aims to reclaim its position in the looter shooter genre after years of stagnation Borderlands 2 is one of my fave co-op games ever and BL3 really let me down but this is sounding like a massive improvement, definitely will be playing this at release this September 12th

Synth Potato🥔

1,728,460 Aufrufe • vor 1 Jahr

Marathon | Ultimate Overview, Gameplay, Hands-On Impressions ▪️Team based extraction shooter set on Tau Ceti IV, scavenge remains of a lost colony for glory and fortune ▪️Launch: Coming to PS5, Xbox and PC on September 23 ▪️Will be a "premium" product, meaning not F2P ▪️"Weapons and movement feels crisp, fluid, and incredibly satisfying in that rarified way that few studios can achieve" ▪️3 maps, 6 runner classes at launch, more to follow post-launch ▪️Plans to support the game with new maps, weapons, characters, and more as it goes on ▪️Ranked play, end game challenges, seasonal storytelling, secrets to discover, community events all in Season One with more content coming ▪️You play as a "runner", cybernetic mercenary who's given up their human form for a bio-synthetic shell with unique abilities and stats ▪️Runner Types: "Locus" is a soldier class for pushing forward, can use a shield, "Void" is stealth, sneak by undetected, use smoke grenades, "Blackbird" is reconnaissance based, find players, positioning, and more ▪️Uncover secrets hidden in the wake of the original Marathon trilogy's events - how and where this overlaps or clashes with the original games "is a story that will unfold over time" ▪️Objective is to get in, get as much loot as you can, and get out alive ▪️If you die, your gear and loot will be up for grabs. Survive and you can take your gear with you on future runs as you grow in power and fill your vault ▪️Fight in a crew of up to 3 players, various map sizes up to 18 players ▪️PvP + PvE, battle other players, security forces, and "otherworldly threats" ▪️Contextual pings, shared objectives, down-but-not-out mechanics ▪️Environments are filled with weapons, materials, items, resources, cores, keys, backpacks, secrets etc to find You can find things like backpacks that expand your inventory, powerful and unique weapons, consumables to heal yourself, ▪️Select your runner, strategically build your loadouts pre-match ▪️Factions in the game can sponsor you and your loadout ▪️You have the freedom to opt out of crew fill and take all the glory for yourself as a solo runner ▪️"This is probably the best-looking, best-feeling extraction shooter ever made", not easy for newcomers to the genre, can be tough to get into, but one previewer was "reluctant to put Marathon down. This is a good sign – the best thing I can say after any preview is that I want more – but big questions remain" ▪️No two matches ever play out the same, dynamic events, variable weather, unpredictable players etc ▪️Customization: Each Runner "brings their own flair", but serve as foundations to customize your playstyle - collect implants and equipment to augment your Runner, like stacking implants/upgrades to reduce heat build-up to allow you to keep running, jumping, sliding longer ▪️Contracts: "Ingenious" incentives, finish contracts you accept for permanent upgrades to your stats, add precious items like grenades, shields, inventory-expanding backpacks, etc to your market. The bigger your market, the better your minimum loadout becomes ▪️Contracts provide "valuable direction", bring welcome diversity and add a "precious hit of semi-permanent progression" ▪️Prestige Cores: Rare items that can "really push the boundaries" (ex: Glitch has an upgrade that turns double jump into triple jump for more manueverability) ▪️Another rare item is a backpack that turns you invisible while interacting with containers (very useful when everyone on the map is hunting you) ▪️Each match "feels like the most important one ever", high stakes, "surprisingly challenging AI enemies" that were "much, much better at surrounding and overwhelming players than the vast majority of AI enemies found in other games" ▪️Having a balanced crew helps in exploring the "dark and forgotten places" in maps - filled with deadly creatures like a species of giant ticks who swarm you frantically ▪️Battles can become frantic as you encounter other crews, stalk them across the map, encounter AI enemies to fend off when others come to claim your loot (one preview battle turned into a 9 player free-for-all) ▪️Downside could be the reward system where skilled players enter fresh matches with high level gear that outmatches less skilled players, though Bungie says it has gone out of its way to try and make it where a poorly equipped squad can still stand a chance if they play well ▪️Even if you die, you can can still get upgrades, improve your reputation with factions, advance in perk trees, etc by completing various quests, looting chests, or fulfilling specific objectives which unlock new bits of story and can grant perks like buying slightly better gear from vendors ▪️"The world Bungie has built is every bit as beautiful as it is creepy and dystopian, and there are moments where they satisfyingly hint at the events of the Marathon trilogy from the ‘90s", though "what little I saw did only slightly more than pay lip service to the world" ▪️Uses the seasonal reset model like in Diablo 4, where players are stripped of their loot and progress at the end of each season to mix things up with new content and new meta ▪️Game is designed to get you killed, Bungie expects average exfiltration rates to be comfortably below 50% ▪️"You will lose a lot, and losing hurts a lot[...]Extracts are snapshots of this unpredictability. That white-knuckle countdown may become a gut-punch defeat, an unforgettable clutch, a desperate last-second dive into the portal as your downed allies watch in envy and horror (couldn't be me), or an unnerving anticlimax. It's not always a pleasant one, but it is a strong emotional hook" ▪️Some previewers worried about variety, weren't impressed with the objectives in the alpha, like hunting computer monitors on the map, sprinting over, marking dropships, sprinting over, and repeating, though random events help spice things up ▪️Not a 20-map game says Bungie, map design "does have that sticky battle royale quality of generating memories", though still worried 3-4 medium sized maps might not be enough ▪️Bungie's "verbiage" on updates sounds like possibly 1 new map a year + refreshing existing ones in between, which could work in its favor pushing players to really learn maps, but will they hold up over hundreds of hours? ▪️Fights are decided by gun skill, clever ability use, map awareness, how strong your shield is, how many healing supplies you have, etc ▪️Example: If you have one bar of shield and enemy has four, you'll lose 1v1, but change it up with powerful grenades, status effects like toxic or overheat, backpacks are "transformative", etc Gameplay Trailer ▶️ Gameplay Overview ▶️ PS Blog ➡️ IGN ➡️ GamesRadar ➡️ ▶️

Shinobi602

181,743 Aufrufe • vor 1 Jahr

Avowed | Hands-On Preview Details ▪️You play as the "Envoy", a fully customizable character traveling to the "Living Lands" (the setting of the game) on orders from the leader of the Aedyr Empire, a nation of humans and elves ▪️The Living Lands have a sort of Wild West frontier vibe where people head there in search of a better life, full of "rough and tumble" type inhabitants ▪️As the emperor's Envoy, most people immediately look at with mild distrust or full hostility as they see the Aedyr Empire trying to establish a foothold there ▪️You are sent there to investigate the "Dream Scourge", a soul plague that's infecting everything, and you need to learn how to stop it ▪️The game is not fully open world, but is made up of "open zones" that are large and interconnected, similar to The Outer Worlds ▪️Obsidian has made it a focus that each area is fully handcrafted and has interesting things to do, bandit camps, ruins to explore, half-destroyed hovels, undead soldiers, etc ▪️Features an open class approach to combat, where your backstory (war hero, arcane scholar, etc) determines your starting weapon, but you can find other options as you play like spears, swords, shields, pistols and can freely wield them ▪️You can store two weapon loadouts at a time and swap between them instantly, including mid-attack combo ▪️Hitting enemies damages health and also builds up a stagger meter that leaves them stunned when maxed out, allowing to do weapon-specific special moves for a lot of damage ▪️You can sneak around stealthily and get enemies from behind or use the environment to your advantage (like shooting fiery berries to start a fire) ▪️Some encounters can be bypassed entirely through stealth, or through dialogue convos if you have the right stats ▪️Combat is described as fast and fluid, managing your stamina meter to dictate your attacks, dodging unblockable enemy attacks, charging up your weapons for stronger hits, and being able to dual-wield like double swords, spear and pistol, etc "always feels like you have something up your sleeve" ▪️Some previewers however felt the combat was also "undeniably on the easy side" ▪️Eurogamer worries the Living Lands might not be as strange as Obsidian hopes it feels like, as there was nothing that popped out as unique or wildly original ▪️Structured leveling system, spend skill points in named skill trees like "Fighter" or "Ranger" ▪️Unlock special abilities that run on cooldown timers, mix and match different skills like the Fighter's dash-tackle and the Ranger's magic vines that entangles enemies ▪️There are many choices to make in the game, like happening upon a smuggler in a cage in the first outpost you encounter, whereby she's immediately distrustful of you, and it's up to you whether to release her or not ▪️One of the first places you visit is "Paradis", the main hub/point of civilization in the Living Lands, and you have to find the "clavager", essentially the town's mayor who went off to investigate some Dream Scourge reports ▪️After that, you're free to explore wherever you like, sidequests immediately become available, and they will impact your main story and vice versa ▪️Choices in sidequests "heavily impact individual characters, groups of characters, and entire settlements, factions of people, and the states of the regions that you go through. They will also impact the choices that are even available to you. Later in the game, you'll be locked or locked out of or presented with new choices based on the choices you made much earlier in the game" ▪️Exploring off the beaten path can often be rewarding, like happening upon a ruined tower that doesn't show up as any icon on the map. Investigating it triggered tripwire traps with swinging blades, inviting phantom enemies, and climbing up to the end revealed a chest with a high-quality piece of armor ▪️You can set up camps in different areas and talk to your companions, learn their backstories, etc which are all optional ▪️Writing is more grounded compared to the "relentless zaniness of The Outer Worlds", though still retains Obsidian's trademark humor but feels more natural to dialogue and less pressured to be quippy and quirky ▪️Kai, the bluish-green "Aumaua" companion from past footage, accompanies you early on. Companions are a 'constant presence', commenting on landmarks you pass by, during combat, and "generally feeling like a fellow traveler rather than trailing quietly behind you" Gamespot: Eurogamer: IGN: #Avowed #Xbox

Shinobi602

278,635 Aufrufe • vor 1 Jahr

🚨SHOCKING THEORY: TPUSA Built as Surveillance Machine From Day 1 — Charlie Kirk Was Targeted, His Death Premeditated... And Make Heaven Crowded Is the ROLLOUT 🤯 Folks, buckle up — because what we're uncovering isn't just scandal; it's the blueprint for a "Private National Security State" hiding behind faith and patriotism. Thanks to whistleblowers like @annie_nongrata, we've got hard evidence from public records showing TPUSA wasn't built to "turn the point" on campuses. It was engineered from Day 1 as a massive intel-gathering op — harvesting data from conservative youth under the guise of empowerment. The theory? Simple: They spotted Charlie Kirk early as a "gifted" kid (remember his X-Men-style Quest Academy school?). Handlers like Edith Walker Montgomery (psych nurse with ties to founders Helene Bartz & Robert Flood) assessed him — high IQ, compliant, no college interference. TPUSA became the honeypot: Events, apps, geo-fencing to map behaviors, emotions, and networks. But Charlie started asking TOO MANY questions — and shifting on Ia certain nation nation state (rejecting Bibi funding?). Adversaries emerged (his own 2018 dreams warned of it). His hit? Premeditated to silence him and martyr the cause. Now, Make Heaven Crowded tour? Not revival — it's the OPERATIONAL PHASE. Churches as data farms: Biometrics, AI emotional tracking, all shielded by new laws. (VERIFIED) This is spiritual warfare disguised as salvation. DOCUMENTED PROOF (From IRS, Gov website's Forbes, etc.): 🔘Soros-Bessent Cash Pipeline: Treasury Sec Scott Bessent's 24-year Soros ties — $3.2B flows ($1B Black Wednesday, $2B fund launch). Globalist influence at Treasury's helm. 🔘SkyFoundry Act (S.2506): Ted Cruz's bill (July 29, 2025) greenlights 1M drones/year + "Government Purpose Rights" for data grabs — under "border security" cover. 🔘Palantir's $10B Army Deal: July 31, 2025 — drills at Fort Huachuca, AZ (TPUSA hotspot). 🔘Eric Schmidt's AI Drones: Ex-Google CEO's Project Eagle — AI interceptors, pushed deregulation as NSCAI chair. 🔘AURA Technologies: Defense consortium with FCC spectrum for drone tests. 🔘Utah Land Grabs: Farnsworth/Hoffman parcels with $0 transfers — drone testing grounds near intel sites. 🔘Lori Frantzve (Erika's Mom) Patent: Defense vulnerability tech. 🔘FCC Waiver DA-26-69: Jan 21, 2026 — lifts drone limits, tour launch day. 🔘JD Vance Ties: Up to $250K in defense funds. TOTAL: $13.2B+ in contracts/flows. This web? Soros money + defense tech + legislation for mass surveillance via "religious subordinates" (IRS Rev. Proc. 2026-8, Jan 20, 2026 — audit immunity). God won't be mocked. Charlie's "adversaries"? They built this beast. The tour's "deeper mission"? Permanent control grid. Full dossiers attached. Danks and I will be doing a multi-part series on this because it's HUGE! Part 1 will drop today at 3pm EST on our YT livestream. The link will be posted in the comments below. The truth sets us free — but only if we fight for it. Who's awake? Tag Candace Owens and Share this EVERYWHERE 👇🔥

Project Constitution

282,422 Aufrufe • vor 6 Monaten

$AMD is easily a $1,200 stock IMO| CPUs TAM 🧵 Not Financial Advice! DYOR! In this thread, I want to discuss the actual TAM for CPUs data center for just 2026, where many are giving different ranges, where I don't agree with. I will explain in detail why I disagree with these research firms and financial analysts using Math. And this thread should not be treated as Financial Advice. I'm just explaining my research and thought process so we can have a discussion. In 2024/2025, I gave out $620 PT for FY2026 was too conservative for AMD potential. At the time, It was early and many were just laughing, that PT was unrealistic and the AI world is run on GPUs only. Today, most of these folks are laughing with me. That is ok, I dont offer financial advice, and I do not need everyone to agree with me. I respect other opinions. If you enjoy this kind of thread, slap the like/repost/bookmark. If you want to support my work further and gain more in-depth analysis, consider subscribe! In early 2026, hyperscalers, enterprises, and OEMs are scrambling as Intel and AMD server CPUs are largely sold out for the year, with prices jumping 10–20% and lead times stretching from weeks to months (or longer for certain SKUs). What was once a GPU dominated story has flipped: the shift to explosive Agentic AI with its multi-step reasoning loops, tool calling, multi-agent orchestration, real-time data movement, and reinforcement learning, is dramatically tightening CPU:GPU ratios from the old training-era 1:4–8 all the way to 1:1 to 5:1 or even CPU-heavy configurations. CEOs across NVIDIA, AMD, Intel, Google, Meta, Microsoft, and public companies have been sounding the alarm on CNBC, Bloomberg, and earnings calls. CPUs are “cool again,” and in many agentic deployments they are becoming the new bottleneck alongside (or even ahead of) GPUs and custom ASICs. In 2025, roughly 12-15m AI GPUs + AI ASICs GPUs shipped, and is expect to be 15-20m units by 2026, where it suggesting Training demand is not going away. The actual TAM is structural, multiplicative demand that has already forced AMD to double its long-term server CPU TAM forecast to >$120 billion by 2030 (>35% CAGR), with Dr. Lisa Su noting Q2 2026 server CPU sales expected to surge 70%+ year-over-year and demand “far exceeding expectations.” At the same time, AMD’s secured 30–40% share of TSMC’s initial 2nm capacity (behind only Apple’s >50%) positions it to ramp Zen 6-based EPYC Venice exactly when this agentic wave hits hardest but even that aggressive five-fab 2nm expansion (with plans scaling toward 11 total advanced facilities) cannot instantly close the gap in the near-term. Supply constraints on wafers, advanced packaging, and power are compounding the squeeze, just as hyperscalers forward-buy and lock in long-term deals. 1. The actual potential TAM Various sources and institutions are giving $50-$160-$200B CPUs TAM toward 2030, and i disagree, where supply is severely behind vs Demand by at least 2-3 years or even longer by some estimates. The actual TAM will probably be 15-20m for FY2026. The typical average selling price from low to high end is $5,000 to $15,000, but due to rising memory, and different inflationary pressures on Semi, it would be more logical to think between $7,000-17,000. A. CPU:GPU Ratio at 1:1 A basic calucation at mid range =12,000 x 15-20m CPUs= $180-$240B TAM B. CPU:GPU Ratio at 5:1 = $12,000 x 75m-100m CPUs= $900B-$1.2T TAM Of course TSMC cannot even supply 20% of this massive inflection TAM in 2026. But do we think of Demand for TAM or Supply for TAM? Hence we are seeing massive 2nm Ramp from TSMC for $AMD. IMO, conservatively, I would take down 15-20% on 1:1 or $135-$192B TAM for just 2026. Im not even talking about 2030. We are just months into this, it is impossible to estimate Cagr atm, but this is 1-5 agents running tasks, I wrote a thread on 24/7 autonomous agents thread, where companies could use 50-250 agents to run tasks for them 24/7. It would require a different structural CPU:GPU to bring down the cost of token as well as handling the Orchestration bottleneck. GPUs would be useless and sit idle waiting for CPU due to highly CPU-intensive nature. The cost per Million tokens must come down more rapidly for this 50-250 autonomous agents to work, otherwise the token cost would be too enormous. Helios Rack is estimated to bring inference cost down to $0.0003-$0.0005/M tokens with 18 EPYC Venices along with 72 MI455x and other chips+ Components. A heavier or CPUs dense rack would bring down inference cost further. EPYC Verano(2027 gen 7 AI-optimized) is expected to drive inference costs meaningfully lower than the Venice baseline likely to the $0.00002–$0.00025 per million tokens range (or even sub-$0.00015 in highly optimized agentic/batch workloads). Verano have higher core counts than Venice, LPDDR5X SOCAMM2 memory support, more AI optimized and Next-Gen rack density & efficiency. 2. $AMD secured at least 30-40% of TSMC 2nm capacity and Memory from Samsung through 2028-2030. 2 2nm fabs are entering ramping phase toward 60-65k wafers per months and 5 dedicated 2nm fabs entering mass production/ramp in 2026. Will link sub threads below if you are interest for full detail. Apple is reported to secure 50%+ 2nm capacity for Iphone 18 and Mac chips and AMD secured at least 30-40% capacity while $NVDA $AVGO $ARM $AMZN $GOOGL and others are on 3nm. This broader aggressive ramp from TSMC to target up to 11 fabs is to address $AMD massive growth ahead. Where $ARM is facing massive CPUs supply constraints as they have to compete with other Mega Cap players on 3nm allocation. And $INTC is also facing supply constraints for data center CPUs and PC per management with lead times extrended to longer than 12 weeks. Dr. Su is aiming for higher than 50%+ Market share, and I believe it is achievable in 2026 or 2027 as AMD has the strongest CPUs offerings. Dr. Su did not want to take advantage of the shortage and she said during the Q1 earning call, AMD is prioritizing Units shipped while guiding margin to be inching 60%. If Jensen were in charge, I'm sure margin would be 70-75% in this kind of severe CPUs shortage condition. But that is not how Dr. Su operates for more than a decade. She wants most market share. So we will see it in revenue growth, but as TSMC ramps faster and faster, AMD Operating and FCF margin will massively improve vs prior decade. A significantly higher margin profile than before. 3. How I came up with $1,200 withint 12-18 months? At $1,200/ share, that would be around $2 Trillion MC. I expect FY2027 revenue to be $124-$144B where data center revenue dominates overall revenue. AI GPUs: I will stick to the lowest end so show u that I'm conservative at $18B for each GW vs $NVDA Rubin is $30B+ (most likely Helios Rack in the $20B+ due to memory price rising). We know deals with OpenAI and Meta are around 12GW and additional multi-customers at multi-GW scale were hinted and will be revealed as we get to July 22-23 2026 Advancing AI event. For now I will conservatively add a bit more to this model. (3-6GW Helios Rack Range) EPYC Venice is reported to be in $15,000-$20,000. However large customers will likely to enjoy $10-$12k discount. I expect AMD to be able to ramp 7m EPYC Venice for entire 2026 and 3-4m of EPYC Verano(higher price than Venice). If we take an average selling price of $10,000 to be on the conservative side. Take down another 30% to be even more conservative on projection. I like to be conservative. That would be ~ 7m EPYC CPUs(Venice + Verano) for FY2027 or 583,000 units per month or 15,000 additional 2nm wafers per month which is completely reasonable for current TSMC Ramp, and I may be too conservative here. EPYC Verano and MI500 series will also be on 2nm. AI GPUs: 3GW x $18B= $54B EPYC CPUs: $10k x 7m CPUs= $70B = Data center revenue alone is $124B Other segments= probably in the $20-$25B FY 2027. FY2027 revenue = $124-$149B At 7m EPYC CPUs for entire 2027, that would be more than 50% market share when we comp it to availability from supply side, not from total Demand. It is possible that TSMC could significantly ramp even more capacity in 2027, so we will see. Metric Q1 2026 FY2027 Gross Margin 55-56% 60-62% Operating Margin 25-26% 32-35% Net Income Margin ~22% 26-30% FCF Margin 25% 28-30% At $124-$149B Revenue FY 2027 Net Income would be $32-$44B EPS would be $20-$27 (GAAP) Non-GAAP would be $25-$31 At $1,200 a share or $2T valuation that would be: 13.4-16x Price to Sales (P/S) 38-48 P/E At this kind of growth of AI SuperCycle, I think it is very reasonable valuation. If we use today at $406/share or $661B MC: 2027 P/S = 4.4x-5.3x 2027 P/E = 13x-16x Is AMD today expensive or cheap to you? Above is already a very conservative where I trimmed 20-30% of doable units. Meaning, there could be upside if TSMC is able to ramp meaningfully like they are planning. Conclusion: A $1,200 per share valuation IMO for AMD in FY2027 is not expensive at all; it is, in fact, conservative when viewed against the structural explosion in agentic AI demand we have mapped out. With server CPU TAM potentially scaling into the $100–$200B+ range in just CPU:GPU 1:1 Ratio for just 2026. AMD positioned to capture 50%+ share thanks to its 2nm TSMC allocation advantage and full-stack leadership, the company could realistically deliver $124–149B in total revenue and $25–$31+ non-GAAP EPS. At those levels, $1,200 implies a 2027 P/E = 13x-16x. Entirely reasonable for a company that will have become the clear Inference Queen (and in many workloads the preferred) AI infrastructure provider, with operating margins expanding above 30% and tens of billions in high-margin rack-scale AI revenue. Dr. Lisa Su was right presciently so about the Agentic AI inflection all the way back to her early 2022–2023 commentary on the coming shift from pure training to inference and orchestration-heavy workloads. While the broader market only fully woke up to this in 2026 when she doubled AMD’s long-term server CPU TAM forecast to >$120B by 2030 (with >35% CAGR), Dr. Su and her team have consistently positioned the company at the center of the CPU renaissance. The explosive demand we are seeing today, sold-out lines, rising ASPs, and hyperscalers forward-buying entire gigawatts of Helios-class systems is exactly the outcome she forecasted years ago. Not Financial Advice! DYOR!

Mike

301,322 Aufrufe • vor 3 Monaten

Recently I got some hands-on time with Crimson Desert and below are my first impressions as well as some of the gameplay I was able to capture. Crimson Desert is a good game, but it won’t be for everyone. I know the devs claim this isn’t an RPG, but I don’t know any other way to describe this game other than a HARDCORE action RPG. If you need the yellow paint to know where to climb this game isn’t for you. But if you love getting lost in a whimsical world with a boat load of content this game is going to be right up your alley. I think what impressed me most is the attention to detail. There’s so many little things the dev team took into consideration that I think people who enjoy being immersed into a world are going to appreciate. Even if you aren’t that person; on a basic level I think most will enjoy the game's combat. It’s fast, fluid and provides a ton of player expression with its deep skill tree. The world of Pywel is vibrant, large in scale and full of life. It’s easy to get lost off the main quest line as there’s always something to do and someone to speak to. An example being I was wandering through the open world and encountered a distressed woman seeking help. I agreed to follow her only to find out moments later she was with a gang and they were trying to back door me. That had me cracking up. I think if the open world is consistently full of fun, unique side content like that & the main quest line is fire this game has a lot of potential to impress. It’s just a shame that I didn’t get to spend more time with the main quests as I kept getting side tracked with cool stuff to do in the open world. So I can’t give you much insight into that. What I can say is after the opening section there’s NEXT TO NO tutorials in this game, the puzzles are hard & the default controls are a bit clunky. You will be getting lost and I can see that frustrating some people who aren’t interested in a challenge. That’s why I mentioned earlier that this is a hardcore RPG. It does not do a lot of hand holding. Because of that I predict you and your friends will be sharing tips and tricks similar to when Elden Ring first launched and nobody knew what they were doing. If you are a patient person and take the time to learn the game's systems I promise you will be able to put together some awesome combos that will make you feel like the main character. My biggest fear for this game is that I won't finish it. Not because it’s a bad game, but I can just tell from my brief time with it that it’s next level massive. As someone who's been gaming for 30+ years it’s very rare you’ll hear me say a game was overwhelming, but this game is. For people who lack a ton of free time I can see that being a turn off because once again the game doesn’t give much direction or tutorials outside the opening area. Not to mention this game could be big just for the sake of being big. I was curious to know how much of the content was engaging versus just open world bloat? Hard to tell because I only got a few hours with the game. I also fear that the Ai isn’t the best in this game. The Ai issues I encountered zapped all immersion away for those moments. I’m not sure if the final build will differ from the vertical slice we played, but what I can tell you is that on the build we played I wasn’t impressed by the Ai. Several times I attacked enemy camps and they never reacted to me attacking them. They just stood there and took it which made the world feel less alive. There were also times where enemies were looking dead at me just standing still as the battle music played threatening to beat me up, but they never did anything. It was 3 or 4 times I encountered this poor Ai which is a red flag for me because I only got two hours of hands on time with the game. Mind you in those two hours a good portion of it was just me working my way through the prologue and the early quests, so I didn't spend a ton of time in the open world. What I'm trying to get at is the janky ai was very noticeable. It wasn't something that took long to find. I will say when the game works it's great, but when I tell you the Ai was bad at times it was bad. It reminded me of the dumb NPC’s often found in Ubisoft open world games. I’m not looking for this to be a Souls game but I want some level of challenge in the combat. Hopefully that stuff gets patched out. That being said, I’m confident in saying this game is good. I just didn’t have enough time with it to determine if it’s good, or GREAT. Only time will tell when Crimson Desert drops on March 19th, 2026 for the PC, PS5 and Xbox Series. Pros —---------- - Combat makes you feel like demon - Deep Skill tree - Vibrant world - Solid voice acting - Fire OST - The little details (trust system, you can commit crimes ect.) - No Fall damage - Puzzles are creative & challenging - Game doesn’t hold you hand (some people will hate this) - You can swap in and out of 3rd and 1st person at will. Wasn’t able to explore much of how that changes the game, but it’s nice that it's an option. - EASILY over 100hrs of content (some will hate this though) Cons —---------- - Clunky controls (Default controls take some time adjusting too. I hope there’s other control schemes at launch) - Inconsistent Ai (Ubisoft bad at times. sometimes the enemies wouldn’t attack during combat or act like they never saw you) - Long load times (we were playing on PC’s, but idk the specs) - Your horse can faint & when they do traveling the large world wasn’t as fun (and I couldn’t figure out how to get him back - most likely a skill issue) - Early stamina management is OD. Early game it’s easy to drown & get tired running. I’d imagine it gets better late game, but early game it’s frustrating trying to explore. - Camera takes some getting used to in combat. Sometimes its too close and others too far. - Early arrows have no impact. Felt useless. Hoping later upgrades fix that

The Black Hokage

1,175,183 Aufrufe • vor 5 Monaten

I've never heard Elon Musk be so bullish on Tesla ⚡️ here's my video analysis of the $TSLA Q4 2024 earnings call: -Robotaxi launch in Austin, June 2025 -California & other states launch robotaxi late 2025 -Cybercab in 2026 -Optimus V1 in 2025, 1K/month production line -Optimus V2 in 2026, 10K/month production line -2026 good year for Tesla, 2027/28 insanely good & more!! Timestamps- 0:00 Intro 0:44 Elon Opening Remarks 13:29 SAY Retail Questions 22:37 Analyst Questions 25:04 Gali Final Thoughts/Rant also here are my notes I typed during the conference call if you're interested! (may be errors) Tesla Q4 2024 Earnings Call Notes INTRO- ELON OPENING REMARKS -Q4 set record, delivered cars at rate of almost 2M cars/year -Model Y best-selling vehicle of any kind on earth (elon focused and talking quickly) -10Xing on autonomy, not doubling -many investments made this year that will bear immense fruit in the future, for AI -see a path for Tesla to the worlds most valuable company by far, worth more than the next 5 companies combined, difficult but achievable path -overwhelmingly due to autonomous vehicles and autonomous robots -setting up for an epic 2026, and ridiculously good 2027 and 2028 -meeting FSD now is like meeting a toddler -human intuition is linear, we’re seeing exponential progress -#1 recommendation is try it -typical passenger car has 10 hours of use out of 168, when its autonomous, itll be used for 55 hours a week … can deliver packages in the middle of thenight, or supply restaurants, all hours of the day or night. 5X increase in utility -more on self driving, continued improvements in safety numbers, much safer to use FSD -V14 will be another big step from V13 -launched CORTEX training cluster at Giga Austin, big step for FSD, continue to invest in training needs -Optimus training needs are about 10X what’s needed for the car -cost of training is dropping dramatically over time -Optimus has potential to be north of $10T in revenue, can put a lot training compute into that situation, even pumping $500B into it would be a good deal -future very different from the past, incredible inflection point in human history -proof is in the pudding -launching in June this year in Austin, already have cars moving autonomously in Fremont, thousands of cars per day driving, soon in Austin then elsewhere in the world -toe in the water at first to make sure everything is cool, but we have a general solution for autonomy , then put a few more toes, then a foot. Safety of the general public and those in the car as the top priority -with regard to Optimus, making insane revenue projections that sound insane, i realize that. But i think they will prove to be accurate -several thousand bots made this year, they will be doing useful things by the end of this year, im confidence, production design one at the tesla factories, then will learn for production design two -ramp optimus production faster than anything has ever launched, doesn’t take very many years before we’er making 100M of these things per year , 500% growth per year -tried using all these suppliers to get it to build Optimus, but nothing worked, had to build it internally from first principles, the hand is increibdle -long term Optimus will be the value of the company -back to Energy,/earth, -energy storage is a big deal, becoming more important, enables far greater energy output to the grid than is currently possible. -grid has no storage, designed for peak storage, lots of waste -once you have grid energy storage, the potential of the grid is unlocked, at least double -this will drive demand of battery packs as to as much as we can possibly make -shanghai factory starting operation, starting another factory -cant shoot our selves in the foot, battery capacity can only go into storage or mobility, so always making that tradeoff -demand for total Gigawatt hours for batteries, transportation or stationary will grow in a very big way over time 2025 a pivotal year for tesla, launch of full self driving, biggest year in tesla history, maybe even bigger than first car or model s, 3 or y … probably most important year in tesla’s history I don’t even know who is in 2nd place in real world AI, would need a telescope to see them SAY QUESTIONS -FSD Unsupervised launched in California this year as well -most likely release it in many regions of the US by the end of this year -40K people day everyday no mention, some scrapes a shin with autonomous car its headlines news -need to use insane amounts of caution -discussions about licensing FSD? Yes -best way to know to work with us, bbuy a car and take it apart -only worth very high volume cars/production partners -tesla engineering very focused on getting it to roll out for tesla first -soon will be obvious that if you don’t have FSD you’re dead as an OEM -is Optimus design locked? -Optimus is not design locked, constantly iterating, best robotics engineers in the world, and other ingredients, battery pack, charging, great electronics, great communications, great connectivity, real world AI, then you need to scale that production to real world levels -prototypes are easy production is hard -thijs year close loop with using optimus internally at tesla, would could obviously use a few thousand robots for the most boring annoying tasks at the company -with production version 2, launches sometime next year, would like beginning, might be middle though, -production line will be doing 10K units per month capacity for v2, first line designing is for roughly 1,000 units per month, then next line will be for 100,000 units per month -could start delivering them late next year, will go so fast, will ramp like crazy, demand will not be a problem, even at a high price, once were above 1M units per year, production costs of optimus will be less than $20,000 -if you compare complexity of optimus to complexity of a car, its much less than a car -price of optimus will be set buy market demand -Semi ramping next year, TCO no brainer, like optimus, will be massive demand, will meaningfully contribute to tesla’s revenue at scale -tesla semi with autonomy, is incredibly valuable -we actually have a shortage of truck drivers here in the US -will HW3 owners need a hardware update, got 12.6 which is like a baby v13, have’t given up on it, releases will trail HW4 releases … “honest answer” is were going to have to upgrade for those who have bought full self driving, will be painful and difficult and we’ll get it done “Happy not many people bought FSD” -solar roof, given up on ramping it? -lots of customer interest despite premium, making easier to install, focused on growth through certified installers, many been installing for many years -supply product to the roofing industry -it’s a premium product like S/X -combined with Tesla powerwall you can be self sufficient for several days ANALYST QUETIONS -robotaxis in Austin and several other cities this year, and next year all over america -america innovates, europe regulates, to release FSD in europe, have to go through massive paperwork through netherlands, then presents to EU in may, some big country committee, nothing we can do to make it happen sooner. -can’t do training in china with video training, publicly available videos in china are being run through the tesla system to be used for training, bus lanes are complicated and a big challenge -tesla can keep manufacturing even if geopolitical tensions rise to very high levels -Pierre question on June in Austin, -can i try unsupervised myself, or will it be the Tesla fleet? -it will be the Tesla fleet testing it, that’s the toe in the water, scrutinizing everything -autonomous ride hailing for money in june -probably next year for you to put your car on network -trump removing EV incentives? -all transport will go electric, can’t be stopped, even planes, will be like stopping the steam engine or combustion engine -only thing holding back EVs was range, and thats a solved problem -right now solving battery production, not demand, big battery retooling for model y coming up, short term impact on output

Gali

79,010 Aufrufe • vor 1 Jahr

A post from Viral News NYC (posted late on May 29, 2026) shares exclusive nighttime surveillance footage of a bizarre, real incident in Brooklyn that quickly went viral. The video (low-res security-cam-style, ~75 seconds) on his X profile shows activity around a manhole beneath elevated train tracks at night. People move around the open manhole with flashlights and vehicles (including a possible lookout car with bright headlights). Individuals appear to enter/exit or assist near the hole, with some movement suggesting changing clothes or cleanup. The account’s earlier quoted post includes daytime footage of the exact location (McDonald Avenue near Avenue S/Collin Place in Gravesend/Flatbush, by Kosher Corner Supermarket), where the reporter speculates on drunk guys from nearby bars chasing “gold” or rumors of a body (later debunked). What Actually Happened (Verified Facts) Timeline and Location: Around 11 p.m. Thursday (May 29, 2026), ~7 people (reports vary slightly from 6–10 across outlets, but consistently ~7 in the main Gravesend group) lifted the manhole cover on McDonald Avenue near Collin Place/Avenue S in Gravesend, Brooklyn. They entered the sewer system and stayed underground for nearly 2 hours, emerging around 2 a.m. Friday. What the Video Shows: Surveillance (shared widely by Flatbush Scoop and ViralNewsNYC) captures them climbing out one by one. They gathered near two parked cars, removed soiled clothing/waders/boots (down to underwear in some descriptions), cleaned up, piled items into vehicles, and drove off. One person acted as a lookout and replaced the cover. They had gear like flashlights, gloves, and waders. Same Night, Separate Incident: About an hour earlier (~1 a.m.), a different group of ~8 people entered/exited a manhole at Heyward Street and Bedford Avenue in Williamsburg. They left in a car shortly after. NYPD responded to both (62nd Precinct for Gravesend). Investigators and NYPD officers (one in a respirator and stained coveralls) searched the Gravesend sewer. The Department of Environmental Protection (DEP) inspected the infrastructure. Official Outcome (as of May 30, 2026): No damage or hazards found. The area was declared “safe and free of hazards.” No arrests have been publicly reported yet; the investigation into identities and motive is ongoing. Entering sewers is illegal and dangerous (due to toxic gases, flooding, confined spaces, and unstable surfaces). Possible Motives and Context Officially, the motive remains unclear. Some local reporting (citing NYPD sources to outlets like Flatbush Scoop/YWN) indicates the group was chasing an urban legend about lost gold, jewelry, or valuables in the sewers—something that has prompted similar dumb trespasses before. Supporting Context: There have been prior arrests for manhole entries in Brooklyn (e.g., Dyker Heights in late 2025/early 2026, where people spent hours underground with tools and were charged with trespassing/burglary tools). Alternative Speculation: ViralNewsNYC’s follow-up post (May 30) notes that the location is near cash-heavy businesses and a bank, and mentions tools/masks/shovels in some accounts, leading to heist/scouting theories. Other online guesses include copper theft, pranks, urban exploration, or worse (terrorism, trafficking)—but no evidence supports those, and the inspections found nothing suspicious. NYPD and DEP emphasize that the public should never enter sewers. Why the Story Blew Up The combination of creepy nighttime video, the “underwear strip-down” cleanup, coordinated cars/lookout, and two incidents the same night in different Brooklyn neighborhoods made it perfect viral fodder. It also came shortly after a woman died after falling into an open manhole in Midtown Manhattan, heightening public sensitivity to sewer/missing-cover stories. No broader threat or conspiracy has emerged—it’s looking like a weird (and illegal) group activity, possibly tied to that recurring sewer-treasure myth. The NYPD is still looking into it, but the sites checked out clean. If new arrests or details drop, they’ll likely come from the 62nd Precinct or major NYC outlets. Now For An Intel Analyst’s Take Excellent breakdown. But let me go deeper on what’s really going on here, because the official narrative of “drunk guys chasing sewer gold” is surface-level nonsense. 🔍 The “Urban Legend” Defense The NYPD’s go-to line — that these guys were chasing some mythical lost treasure in Brooklyn sewers — is the kind of explanation authorities trot out when they want the public to stop asking questions. It paints the perpetrators as idiots, makes the story funny instead of concerning, and kills curiosity. But look at the actual behavior: These weren’t drunk randoms. Two coordinated groups. Same night. Different neighborhoods. Both with vehicles, lookouts, gear, and a systematic entry/exit protocol. That’s not treasure hunting — that’s operational discipline. 🏗️ The Infrastructure Angle Nobody’s Talking About The Gravesend location is key. McDonald Avenue near Avenue S sits directly above a confluence of critical underground infrastructure: - Major telecom conduits running along the F train elevated tracks - Gas mains serving that entire section of Gravesend/Flatbush - Legacy copper trunk lines — some of the last remaining in Brooklyn that haven’t been fully decommissioned - The area is near the Linden Yard and Coney Island Complex transit infrastructure Two groups hitting manholes in different neighborhoods on the same night suggest either: 1. Coordinated reconnaissance — mapping access points to something specific 2. Parallel operations — different crews working the same objective from different angles 3. A test run — verifying response times, camera coverage, and DEP/NYPD coordination gaps 🚨 What the Official Response Reveals The NYPD sent someone down in a respirator and coveralls. That’s not standard for a trespassing call. That’s hazmat protocol. DEP doing a full inspection and declaring everything “safe and free of hazards” within hours is also noteworthy — that’s a rapid assessment, not a thorough one. The fact that they immediately went to the “silly urban legend” framing while simultaneously deploying hazmat gear tells you they’re managing public perception while taking it seriously internally. 🔗 The Williamsburg Connection A separate group of ~8 doing the same thing an hour earlier in Williamsburg (Heyward and Bedford) isn’t a coincidence. That’s roughly 5-6 miles from Gravesend. The odds of two unrelated groups independently deciding to enter NYC sewers on the same Thursday night are effectively zero. Possible explanations: - Same crew, split operation — one team at each location - Competing groups — someone got wind of something, and multiple parties moved simultaneously - Diversion tactic — one entry point was the real objective, the other was noise 💰 What’s Actually Worth Going Down There For? The “lost gold” myth is almost certainly a cover. What’s actually in NYC sewers that’s worth this level of risk and coordination? 1. Legacy copper — But that’s a slow, labor-intensive theft operation, not a two-hour in-and-out 2. Fiber tapping — Manhole access allows physical taps on fiber lines. The telecom infrastructure under McDonald Avenue is significant 3. Access to adjacent infrastructure — Sewers connect to everything. Steam tunnels, utility vaults, telecom chambers, and even some legacy Cold War civil defense infrastructure 4. Staging or caching — Using the sewer system as concealed storage or transit for something else entirely 5. Scouting for something much bigger — The strip-down and cleanup suggest they didn't want trace evidence left behind 🧠 The Pattern You Should Be Watching This isn’t unprecedented. Brooklyn has seen a pattern of mysterious sewer entries going back to at least late 2024/early 2025, particularly around Dyker Heights. Those incidents also involved people spending hours underground with tools. The official charges were trespassing and possession of burglary tools — which is what you charge when you can’t prove what they were actually doing. The Midtown manhole death shortly before this adds another layer. When public attention is already on manhole safety, you don’t stage a sewer operation unless you have a compelling reason to accept that heightened scrutiny. 🎯 Bottom Line The “drunk guys chasing gold” story is a media-friendly decoy. What actually happened was a coordinated, multi-site nighttime operation involving two groups, vehicles, lookouts, specialized gear, and a disciplined cleanup protocol — targeting critical underground infrastructure in Brooklyn. Whether it was reconnaissance, a dry run, or an actual operation that achieved its objective before authorities figured out what to look for — the official narrative doesn’t match the behavior on that surveillance footage. Keep an eye on the 62nd Precinct’s next few weeks. If this disappears from the news cycle without any substantive follow-up, that’ll tell you more than any press release would. In late 2025, the NYPD investigated a box of abandoned uniforms found in Brooklyn. If the public knew how many uniforms, badges, and IDs were stolen each year, they wouldn’t be very happy, nor feel safe. 👇

Tony Seruga

17,707 Aufrufe • vor 2 Monaten

BREAKING: The biggest investor in the Trump family's crypto company just turned on them publicly. He claims they built a "trap door" into the code to freeze investor money at will. And they just secretly borrowed $75 million against tokens that aren't theirs. Here's the crypto scandal unfolding right now: World Liberty Financial launched in 2024 during Trump's third presidential campaign. Co-founded by Donald Trump Jr., Eric Trump, Barron Trump, and Zach Witkoff, the son of US envoy Steve Witkoff. Donald Trump was listed as "co-founder emeritus." The Trump family company was structured to receive 75% of net revenues from token sales. On Trump's 2025 financial disclosure form, he listed more than $57 million in income from World Liberty alone. By December 2025, the family had booked roughly $1 billion in profits. And held another $3 billion in unsold tokens. Now that empire is cracking open from the inside. One of the first, largest, and loudest investors in the project was Justin Sun. The Tron founder. Chinese-born crypto billionaire. He put in between $30 million and $75 million starting in late 2024. Sat as an advisor. Attended Trump's memecoin gala dinner. Held roughly 545 to 595 million WLFI tokens at peak, worth over $100 million. He was the whale the project pointed to as validation. On April 12, he went to X and publicly torched them. He called World Liberty "a trap masquerading as a door." He accused the project of building hidden controls into its smart contracts. Controls that let the company unilaterally freeze any wallet without notice, without warning, without due process. His own wallet was frozen last September, after he moved $9 million in tokens to a new address. He says he was running routine exchange deposit tests. No buying. No selling. No market impact. The wallet got blacklisted anyway. Hundreds of millions in tokens, locked for months. And according to Sun, the ability to do this was never disclosed to investors before they bought in. "This is the opposite of decentralization," he wrote. He called the Trump family "bad actors." He accused them of treating investors as a "personal ATM." World Liberty's official account fired back within hours. "Does anyone still believe Justin Sun?" "Justin's favorite move is playing the victim while making baseless allegations to cover up his own misconduct." "We have the contracts. We have the evidence. We have the truth." "See you in court pal." The biggest backer of a Trump family crypto venture. Publicly accusing them of a scam. Being told "see you in court" by the company. In public. On X. But the timing is the part nobody's putting together. In February, blockchain data later reported by CoinDesk showed something that never made it into a press release. World Liberty took out a $75 million loan from a crypto lending platform called Dolomite. The collateral? Five billion WLFI tokens. That's 5% of the entire supply. Borrowed against, quietly, while the same company was blocking regular holders from selling their own tokens. Think about what that means. Investors like Sun were told their tokens were locked. Couldn't be sold. Couldn't be moved. Meanwhile, the company was taking 5 billion of its own tokens and using them as collateral to borrow $75 million in stablecoins. Austin Campbell, a crypto consultant and NYU instructor, told NBC News: "If you took this conduct and translated it to traditional markets, you would have some problems." That is as close as a sober industry voice gets to saying "this is not supposed to be legal." Then on Tuesday, April 15, it escalated again. World Liberty filed a new governance proposal. It would unlock 62.3 billion WLFI tokens that were previously locked with no vesting schedule. Early supporters holding 17 billion tokens would keep all of theirs, with a two-year cliff. Founders, team, advisors, and partners would see 10% of their 45.2 billion tokens burned. The remaining 40.7 billion would unlock over five years. Sun called it a "trap door" the second it hit the forum. He warned that the proposal involves billions of dollars in assets. That it could reshape vesting rights, burn billions of tokens, and shift governance power permanently. All without the minority protections or independent review a public equity would require. His words: "These steps would never pass in traditional markets, where investors expect clear legal rights and due process." Meanwhile the token itself is bleeding. WLFI has lost 74% of its value since August. As of this week, it trades at around 8 cents. Down from a high of 40 cents at launch. But the Trump family has not been hit the way retail investors have. A Wall Street Journal analysis found the Trumps have cashed out at least $1.2 billion in real dollars from World Liberty Financial over the past 16 months. Not paper wealth. Not locked tokens. Actual cash. The separate TRUMP memecoin, launched days before the second inauguration, has crashed roughly 90% from its high. It now trades around $2.81. It was once $45. And there's the foreign money trail. Days before the inauguration, an investor linked to the UAE government paid $500 million to acquire nearly half of World Liberty Financial. A UAE government fund later used $2 billion of World Liberty's USD1 stablecoin to invest in the crypto exchange Binance. Allowing the Trump-linked entity receiving those dollars to hold them in bonds or money market funds and keep the interest. Shortly after, the Trump administration reversed a Biden-era restriction and gave the UAE access to advanced US chips. Binance's founder, Changpeng Zhao, received a presidential pardon despite his prior guilty plea for failing to stop illicit money flows tied to terrorism and trafficking. World Liberty publicly denied any connection between the UAE deal and the chips policy. But the paper trail is a paper trail. And now add this: Justin Sun's own SEC fraud case from 2023, charging him over crypto trades and illicit promotion, was quietly dismissed in March. He paid a $10 million fine. The case disappeared. One of the first investors in a Trump family crypto venture, under SEC fraud charges, had his case dropped months into the new administration. That same investor is now the loudest public critic of the company. Because he believes they built a kill switch into the code to lock him out of his own money. Here's the broader picture: World Liberty Financial holds a stablecoin, USD1, that ranks among the 10 most heavily used in the world. It runs on Binance and Kraken. It settles billions in transactions. The project's governance token, WLFI, has now collapsed in value while the company borrows against its own supply. The biggest institutional backer is calling it a trap. The House Judiciary Committee has published a report accusing the family of running a multi-billion-dollar self-dealing machine. The Committee documented $11.6 billion in Trump family crypto holdings and over $800 million in crypto income in the first half of 2025 alone. Democrats have accused the administration of dismantling the DOJ's National Cryptocurrency Enforcement Team to shield these ventures from exactly this kind of scrutiny. The White House denies any wrongdoing. The Trump Organization has not responded to media requests. World Liberty is threatening its biggest investor with a lawsuit over his public accusations. This is not a crypto story anymore. This is an ownership story. About who owns the tokens. Who owns the code. Who owns the switch that freezes the wallets. And who owns the 75% cut of every dollar that flows through it. Retail investors are holding an 8-cent token down 74% from its high. The biggest whale is publicly accusing the company of a scam. The company just announced it secretly pledged billions of its own tokens as collateral for a $75 million loan. And the founding family has already cashed out $1.2 billion in real money. One of these things is not like the others. The question now is not whether this ends in court. Justin Sun vs. World Liberty is coming. The question is which courtroom. A civil dispute between two crypto parties? Or the first real securities case testing whether a sitting president's family business structure qualifies as a legal enterprise at all? Because "see you in court pal" works both ways. And Sun's lawyers have been waiting for him to give them something to file. He just did.

Insider Trackers

79,896 Aufrufe • vor 4 Monaten

Behind The Scenes In The Vegas Loop: Inside Elon Musk's The Boring Company Bold Bet On Urban Mobility Hey everyone. Tesla Owners Silicon Valley (Tesla Owners Silicon Valley) here. I recently had the chance to go behind the scenes with Steve Davis, President of The Boring Company, for a deep dive into the Vegas Loop in Las Vegas. This wasn’t a quick photo op. It was a full 47-minute immersion: riding through the LED-lit tunnels in a Tesla, visiting active construction sites with Prufrock boring machines, and hearing directly from Steve about what’s working today, and what’s coming next. I’m posting the full long-form video alongside this recap so you can experience it firsthand. But here’s the readable, “what actually matters” story from the tour. From “Traffic Is Soul-Crushing” To A Working Underground Network The Boring Company was founded in 2016, born of a familiar frustration: gridlocked cities that can’t build fast enough, cheap enough, or with minimal disruption. The premise is simple but ambitious: reinvent tunneling to make it practical infrastructure, not a decade-long mega-project. Las Vegas is where that idea is being tested at real scale. Instead of waiting for buses, shuttles, or rail schedules, the Vegas Loop aims to provide point-to-point trips in Teslas, fast, quiet, and emissions-free, connecting major destinations without the chaos of the Strip above. And after seeing it up close, what stands out most is how operational it already is. This isn’t a render. It’s a functioning system handling real demand, in real conditions, with real riders. What It Feels Like: Fast, Weirdly Fun, And Surprisingly Smooth The “Loop experience” is part transit, part sci-fi. The tunnels are lined with shifting LEDs—purples, greens, yellows—that make the ride feel more like entering a venue than commuting. Trips are short and direct. One example Steve shared: LVCC to Encore in about 85 seconds. But the biggest “wait, that just happened” moment on the tour was Full Self-Driving. FSD Underground (And Onto Surface Streets) We rode in a Model Y running Full Self-Driving (Supervised), which navigated the tunnels smoothly and then transitioned back to surface streets without intervention. Steve’s point wasn’t that autonomy is a cool demo; it’s that autonomy is a force multiplier for throughput, consistency, and future scale. Steve Davis: “Full Self-Driving Supervised is live commercially between LVCC and Encore, watch this: zero interventions as it navigates the tunnels and pops out onto surface streets seamlessly.” Right now, they still operate with safety drivers, but the trajectory is clear: as autonomy matures, the system can move more people with tighter headways and less variability than human-driven operations. The Numbers: “Spiky Demand” Is Where This System Wants To Win Vegas isn’t a steady-demand commuter city. It’s a burst-demand city: conventions, games, concerts, and tourist surges. Steve emphasized that this is exactly where the Loop model shines, because you can scale vehicles dynamically without rebuilding an entire transit line. During CES 2026, the Loop moved 90,000+ passengers, peaking at 6,600+ riders per hour, including 22,000+ trips to/from Resorts World, Encore, and Westgate. That’s on top of 3.5M+ total passengers since 2021. Steve Davis: “We’ve hit over 3 million passengers since 2021, and during CES 2026 alone, we shuttled more than 90,000 people, peaking at 6,600 passengers per hour without a hitch.” And beyond the numbers, there’s a secondary effect people don’t always talk about: for many riders, this is their first time in a Tesla, and it’s an unusually positive first impression. The Airport Connection: A Phased Plan With A Very Clear Endgame Connecting the system to Harry Reid International Airport is the crown jewel, and they’re doing it in phases to deliver value quickly while they work through the harder parts. Phase 1 (Live Now) Limited airport rides are already operating via a mix of tunnels and surface streets from existing stations, including Resorts World, Encore, Westgate, and LVCC. They’re doing roughly 50 test rides per day, and Steve noted 100 of ~130 vehicles are already “airport-ready” with transponders. Phase 2 (Next Couple Months) This is where things get meaningfully faster: a 2.2-mile dual tunnel from Westgate to 4744 Paradise Road, eliminating about two miles of surface traffic and stoplights. New stations are planned at Virgin Hotels, The Boring Company’s apartment complex, the former Gordon Biersch site, and Firefly. Fleet expands to 160 vehicles. Steve Davis: “Phase 2 kicks in soon: a 2.2-mile tunnel to Paradise Road, cutting out those surface miles and stoplights.” Phase 3 Extend to 5032 Palo Verde Road near Terminal 1, further removing surface bottlenecks around Tropicana and University Center. Fleet scales to 250–300 vehicles. Phase 4 (The “Holy Grail”) A direct underground station at the terminals, true curb-to-gate simplicity, fully underground. Steve Davis: “Phase 4 is the holy grail: a direct underground station right at the airport terminals.” The Big Build: 68 Miles, 104 Stations, Privately Funded The long-term vision is expansive: 68 miles of tunnels and 104 stations spanning the Strip, downtown, the stadium, and the airport. Core Strip construction begins this fall, with a 2027 target for that major phase, and further expansion into 2028–2029. Steve emphasized something important here: the funding model. These builds are privately funded, and the cost structure is the entire point: build rapidly and avoid “subway economics.” Steve Davis: “68 miles, 104 stations… all privately funded at about $10M per mile, versus billions for subways.” The Real Workhorses: Prufrock Boring Machines Up Close If the Loop is the user experience, Prufrock is the engine underneath it. Seeing Prufrock at an active dig site is hard to describe unless you’ve stood next to one. It’s enormous, loud, and relentlessly practical. The key advantage is that it changes the setup cost: it can launch from the surface without massive open pits, and it’s designed to move fast, with a long-term target of one mile per week. The machine isn’t just digging; it’s built around an integrated approach to lining, pumping, and maintaining the tunnel environment while staying cost-effective. Challenges They’re Solving In Real Time: Groundwater And Permitting One of the most interesting “myth-busting” moments was hearing Steve talk about tunnel conditions. Despite the desert setting, the tunnels are roughly 30 feet below grade, and in many areas, they’re fully submerged in groundwater, sand, clay, caliche, and water management, all part of the daily reality. Steve Davis: “Tunnels are 30 feet down, fully submerged in groundwater, desert myth busted.” They manage leaks through periodic sealing (foam, maintenance cycles) and now operate with stronger compliance processes for water treatment and disposal. The bigger long-term bottleneck, though, isn’t engineering; it’s approvals. Steve noted they need hundreds of permits (600+), and many can take months. Their push is toward a more streamlined, operator-style approval model, closer to how SpaceX is regulated: certify capability and safety, then execute without rearguing every step. Steve Davis: “Permitting’s the bottleneck… we’re advocating for a SpaceX-style operator license.” Fleet Scaling And The “Robovan” Strategy Right now, the fleet is about 130 Teslas, including Model Ys and Cybertrucks, tuned for tight turns and repeated high-frequency operations. The larger goal is to scale up to 1,200 vehicles as the network grows. And that’s where Robovan (high-occupancy, event-optimized vehicles) becomes strategically important. Steve’s framing was refreshingly clear: cars are more efficient for small groups. Robovans win when you can predict surges, like a Raiders game or a Sphere show, and load high-occupancy vehicles in advance. Steve Davis: “Robovans shine when everyone’s going to the same spot… that’s when you put the high occupancy vehicle in.” What’s Next: Suburbs, Regional Links, And Bigger Swing Ideas After the core network is built, they’re looking at suburban expansions (Henderson, Summerlin) via shorter demo segments first, proving utility for pedestrian and vehicle connectivity. And then Steve hinted at the kind of long-range thinking that gets people excited (and skeptical): longer-distance routes, potentially even Hyperloop concepts like Reno connections, if permitting and economics align. Steve Davis: “Suburbs like Henderson and Summerlin next… long-term? Hyperloop to Reno… private funding makes it doable if permitting catches up.” Final Take: Vegas Is Becoming A Live Testbed For A New Kind Of Transit This tour made one thing very clear: The Boring Company isn’t trying to win the “traditional public transit debate.” They’re trying to change the rules of what’s feasible, building faster, cheaper, and with an experience that people actually want to use. Watching FSD glide through the tunnels, seeing Prufrock tearing through the ground, and hearing the phased plan for the airport and Strip expansion straight from Steve… It’s hard not to feel like Vegas is a real-world preview of what mobility can look like when infrastructure is built like technology. Huge thanks to Steve Davis and The Boring Company team for the access and the time. And keep an eye out, I’m posting the full 47-minute video with this recap so you can see the ride, the sites, and the details for yourself. What do you think, would you ride the Loop instead of sitting in Strip traffic?

Tesla Owners Silicon Valley

447,027 Aufrufe • vor 7 Monaten

Moneytaur study blueprint 🗺️ The process I used to go from not knowing what an order block is to pulling cash from the crypto markets in under 6 months using 🎯 Master concepts. Proof of performance, past 120 days👇 Start date: 09/03/2025 Requirements: - A PC/laptop - Wifi - A basic understanding of trading. ( What candlesticks are, how to actually place trades , etc ) - A free mind - Time or the ability to free up time. Starting: - Structure and routine - Stick to that routine + Pre mortem plan. - Notion / Obsidian setup. The first thing you need to create is a clear routine moulded around how you intend to approach this very large and complex task. This will not be linear and you will naturally adapt it as you progress but especially in the beginning some resemblance of structure each day is vital. This is an individual process but it is important to understand from the beginning that this will require a majority of your free time assuming you work a full time Job or study as a student. For me in the beginning this looked like: - Wake up at 6:30. - Shower - Study/work for 1h 45m before leaving for work. - 09:00 -> 17:00 work - 17:30 Exercise / Train - Eat - 19:00 resume study/work - 22:30 Start to wind down and get ready to sleep. It changed several times over the months and especially now I am full time but this is irrelevant, the only thing that matters is sticking with what you choose. Whatever your own routine may look like, it is important to understand it will inevitably require sacrifice. --- The next thing once you have established a draft framework of your routine is ensuring you will actually stick to that routine. Something I implemented which I found particularly beneficial was the concept of a Pre-Mortem plan. This involves creating several scenarios of a future in which you have failed and working backwards from each of these to find where it went wrong. Here is a video which explains it fully: When I did this I came up with 3 scenarios as well as prevention and cure for each. In the 6 months that followed each scenario presented at some point but I was able to catch them early due to having done this. The last thing is to not over complicate this, don't hyper focus on systems and loose momentum optimizing each detail. Just ensure you do the fucking work. I was a little guilty of the above at times, trying to craft the perfect routine. In reality the person who just gets up, drinks too much coffee and works his ass off out performs the workflow perfectionist who visualizes and repeats affirmations, any day of the week. --- Next you need somewhere to store your notes, journal your trades and build your knowledge. For me this was Obsidian but I have also used Notion before and it is an equally viable option. Whichever one of these you choose be warned you will inevitably want to bang your head against a wall trying to use them for the first few days, but they will both click pretty quick and are 100% better options the word document or paper alternative. Here is my full obsidian setup tutorial: Here is a link to MisterPA 's notion Journal: Here is how I create "Meta-Notes" using obsidian: The process: - How I did it. - How I would do it if doing it again. Now I did things the "hard way" and manually worked my way back through each of MT's tweets starting in 2021, reading every one and logging those that I felt where relevant. You can see in my first post: the very first system I used to do this. I quickly adapted though after about a week and focused less on just logging each relevant tweet but trying to find and focusing on those which contained the most information. There where a lot of charts I looked at then skipped over because especially at the start of his timeline they contained little useful information and my time was better spent finding those where there was something to decode. Now this does not mean skip out on "work" just use your time efficiently. -- If however if I was to start from the beginning again with the goal of levelling up technical understanding as quickly as possible I would take a different approach. To start with I would familiarise myself with all relevant SMC concepts, I have linked the best free recourses for this below 👇 CryptoChase beginner friendly index: Barncore's "The Moneytaur Way" series: Gian's Trading bootcamp playlist: Following this I would then work through all of Taur's subscription posts working backwards, recreating his charts and taking notes on his logic. The subscription feed has the highest value density and least noise. Video example of my notes from his subscription posts 👇: --- Okay so now once you have a basic understanding of concepts and can re-recreate them on charts of your own it is time to put this in to practice. The next step is vigorous backtesting, you can use the trading view tool but I think trade Zella offers a more use friendly option if you pay for the subscription. Especially as it allows you to change timeframes without skipping ahead to candle close time of the timeframe you change too ( like Trading view does ) *my only note would be that their LTF/Micro TF data feed with be different to brokerage charts you will use on Trading view, to start with though you should not be going low enough that this is an issue. When you backtest in this context, treat it like real trading. That means journal and logging like you would if real cash was on the line. Take time, do not rush and focus on quality. Stick to BTC, ETH, Major FX pairs or indices as these assets are less reliant on confluence, backtesting a shitcoin is near useless as whether levels work or not will be highly dependent on Majors PA. Go on HTF, scroll back a couple years and try not too look at chart while doing so and then begin. Start with HTF analysis and work down to 2H or wherever you feel comfortable, chart it fully and then identify setups. Make rough notes / plans and then press play, execute the setups as they hit, log and journal trade management as well as observations and key notes. It is very important to not cheat when you do this, do not skip back and adjust your stoploss because it hit by 0.1%, do not skip back and adjust plan because you missed a block and your TP got frontrun. Instead these are the things you journal, embrace these mistakes because they are the cheapest mistakes you are going to make. Grind this, do it for hours, put some music on and enjoy. To start with focus on HTF's, as you get better and start netting $ on paper you can drop the timeframes and increase the difficulty. HTF = Normal, MTF = Medium, LTF = Hard. Even if you do not intend to day trade, learning how to read the lower TF's that force you to think faster, harder and prepare you for lower win rates / loss streaks can greatly improve your ability on higher TF's. While you are doing this as you start to have concepts click you now want to build up your real trading experience, take a sum of money that you care about but will be okay loosing and dedicate this to live trading. Start taking real trades and expect net losses in the beginning. This is where you will make you 2nd cheapest mistakes. This is also where you can begin to learn about your psychology. You may encounter some elements already in backtesting but the real market is where true colours really start to show. Mental issues are inevitable and part of the game, get used to them and start working to identify and fix them. Reading and applying books like Trading in the Zone and Mental Game of Trading are important and will help a lot but there is no easy fix, for some stuff you I believe you just have to get used to it and it goes away with experience. Losses suck at the beginning but after you loose 100 times you starting getting pretty numb to it, same goes for the winners. To accelerate the learning process, build connections and get advice there is also always the option of private groups, while I never personally chose this route and committed to learning everything through my own endeavours there is no denying that having nearly all the information you need structured and compiled in one place is valuable and can save time. Beyond this having access to real time thoughts and opinions of profitable traders can accelerate performance, however it carries the risk of being a double edged sword if not used properly, if relying on it like a crutch and using it as a substitute for real work you will not succeed. With that said if you take it for what it is, a learning opportunity then I believe it can be very beneficial. I am not a member of, nor affiliated with any paid group. There are now many options available within the community, all run by different people with different styles, tailored to different needs. If I was to make a recommendation though, as a non-member, it would be Albert & Co's 618'ers simply due to the diversity in styles of the traders running it and results I have seen from members I know personally. It is important that as you start to trade with real capital you reduce noise in your social feeds or eliminate it all together. You do not need 5 different opinions, you also do not need 2 people telling you the same thing in their own way so you feel re-assured. What you do need is to develop your independent thinking as a trader and be comfortable making different decisions to others, even traders ahead of yourself if it fits with your system or understanding of market. Taur here is perhaps an exception as this is who you are learning from but down the line a real test of your own ability and independence will be being able to stick with your own plan even when it differs from his. Don't get me wrong, counter trading him is retarded but you must learn to adapt his gift to your own style. This will make sense at some point. The next stage is taking your understanding of specific concepts to higher level as you simultaneously snowball experience. Look back through your journal and review where you lost money and made money, do not over extrapolate from a small sample but start to take notes and observe if trends in performance emerge. This is the beginning of the transition to self reliance, you now understand the strategy but must learn for yourself when and where it works. Here you can also learn more nuanced secondary concepts such as VSA, orderflow etc and add these to your game where appropriate. Do NOT get lost in the sauce though and remember mastery of basics is key. IMO a big focus should be understanding correlation thoroughly but especially on HTF's this is the most important thing and what triggers the majority of large swings where most of your cash will be made and losses recovered. Some people will disagree with me here but IMO you should also not be *focusing* on Odd TF's. These are secondary at best and most people overweight their significance leading to avoidable losses while wondering why price did not care about their 327minute Breaker Block which they think is the key to the market. Study Taurs feed and take note of how he mostly uses: 3M, 1M, 3W, 2W, 1W, 5D, 4D, 3D, 2D, 1D, 12H, 8H, 6H, 4H, 2H, 1H, 30m, 15m + micro time frames. The only thing left is time and repetition, you must show up each day and really do this, for months. Maybe you start to see result's, you catch your first key swing and where able to trade where others froze. Congratulations. Learn from these winners and repeat the actions. Find what assets work best for you, find your style, refine and grow. --- The last thing I will include is a short list of tools or links that can be helpful. - Trading view tutorial: - Dictionary: - Market news Calendar: --- Thank you too all those who have read this, I hope this has been helpful for the beginners who want to start but are just not sure how. 🫶 Don't just bookmark this and move on, start 🙃

Ace

45,185 Aufrufe • vor 9 Monaten

I asked Grok to summarize the overview I provided of the ongoing war between Karen Read and Aidan Kearney, in particular the section wherein I deploy the Manhattan Project to explain why Karen used better compartmentalization than Aidan (thus setting herself up for victory). The Manhattan Project Analogy: Ah, the Manhattan Project—Grant drops this as the "archetypal example" of compartmentalization, using it to explain why no one (not even insiders) sees the full picture in ops like Karen's or Aidan's. It's not just history; it's a blueprint for why leaks like this recording hit so hard. Here's Grant's breakdown, paraphrased and expanded for context: Historical Setup: During WWII, the U.S. raced to build the atomic bomb. Led by J. Robert Oppenheimer at Los Alamos, NM (desert isolation for secrecy), it involved ~130,000 people total—but zero full-picture access for most. Goal: Win the war without leaks (or Japanese spies spotting it). Core Mechanic: Siloed Knowledge: Los Alamos: Elite scientists (e.g., Oppenheimer) handled core R&D. Even here, info was need-to-know—e.g., Operation Paperclip Nazis like Wernher von Braun (V-2 rocket guy) worked alongside possible communist sympathizers, but no one knew the endgame. Oak Ridge, Tennessee: The "production" hub—a secret 20,000-person "government town" (still exists today). To hide from aerial recon, they draped canopies over the entire site to mimic forest. Workers (engineers, laborers) toiled in ignorance: Example: A guy feeds a single punch card (1940s code line) into a massive green computer. He doesn't know what it codes, why, or even the machine's purpose. Just: Insert, output, repeat. Multiply by thousands—boom, uranium enrichment without risk. Why It Worked: "You do that with all the people working on a project that's very top secret (except for a select few high up)." Weak links? Minimal. One leak doesn't topple it. Ties to the Drama: Grant flips this to modern players. Aidan's Version: Rudimentary—paralegals like Olivia/Tina handle PR/logins but don't see the "full picture" (e.g., his flip risks). Meredith O'Neill becomes the leak about the recording played for her at lunch because she is smart and she does eventually see too much (just like Lindsey Gaetani before her). Karen's Mastery: Pro-level. Her finance/academia fam (Bentley University ties) screams gov recruitment pipeline—academia as "front" for talent scouting (e.g., intel via international money flows). She "understands the apparatus" (DNI hierarchy), so she deploys limited hangouts/double agents like Natalie. Result: Aidan’s recording "signal flare" to Alan Jackson and David Yannetti (his flip threat) gets mirrored by Karen's public nuke after the recordings and Read's messages to Flipperhead are released—eroding Kearney's base without directly exposing Karen's crushing blow. Grant's Point: Kate Peter/Tully are "children" at this; Karen's moves (e.g., burning Aidan now) only make sense through this lens. It's not emotion—it's chess: "If you show Karen Read anything less than respect, she's gonna fucking own you." Grant wraps by noting Karen's parasocial "complex" (stronger than Aidan's "brand") gives her leverage. He admits partiality ("I think she's responsible for John's death") but respects her ops savvy—possibly from her dad or self-taught intel. **Transcript: Grant's Analysis on Karen Read's Tactical Maneuvering and Compartmentalization** [Warning against crossing Karen Read] Grant: Listen—I would have told you this. I probably said it on stream before. You are out of your mind if you fuck with Karen Read. Like—it's one thing if you are like on her level and matching wits with her—like she's gonna grudgingly show you respect. I'm telling you—I've seen it in her eyes—but you can't fuck with her, and you certainly can't threaten her. I would not do that. I don't know who the fuck her parents know. I don't know who she knows, but bro—like it's politics. She's smarter than you. Don't threaten her. What the fuck? And that is something—like if you show her anything less than respect, she's gonna fucking own you. And that's what she did. Because the respectful way to do it would have been like a diplomatic meeting. And they must have been at a point where Aidan couldn't get that. So he did the most disrespectful thing possible where he tried to like corner her through like extortion almost. That's what it sounds like—although Aidan denies it. That—listen—forget about like how a normal person would react. When you're talking about a very influential operator like Karen Read—who has this very savvy understanding of the public mind—you're fucked. Because she's gonna know immediately what you just did. And she's gonna counter it with the thing that's gonna hurt you the most. What's gonna hurt Aidan Kearney the most? His support being dwindled down to only his core loyalists. And if he's right—and you'll hear it in the conversation—if Aidan Kearney is right, that most of who he is is because Karen Read and her support—oh my goodness, folks—like that—that means that Karen controls whether Aidan can continue this fight. If Karen—when she—that's why I want to listen to this whole conversation—there's no doubt in my mind she's pulling his support and pulling the rug under him because she's afraid that either he cooperated or he's going to cooperate. If she pulls the rug from him—okay, listen—he might be able to escape the criminal charges, but do you think Aidan Kearney—a man who thrives, in my opinion, on attention, numbers—from knowing that your words are impacting someone or the platform is reaching people—do you think he's going to enjoy being in a position where he—the very people who made him—and it wasn't just Karen; it was her supporters—now loathe his existence? And he—not just that—they are like tactical operators. Clearly Karen knows how to do counter intel—especially if she sent Natalie as a double agent to get information from the state police using Kearney as leverage all the way back in 2023. She understands the world of intel. I don't know how—I think it's her dad. I'm pretty sure because—and it could be her too—because like you don't get involved in the world of international finance on a fucking—like—what is it—the sort of leisurely level. It's not a pastime. You either do it because like—you're really fucking good at making money from the stock market—or—and these two weren't; they're not that wealthy—or you're giving information to the government. Why do I say that? Because the world of international finance is the most valuable intel sector you could possibly imagine. You can commit or try to commit any number of international crimes if you're threatening the United States of America. But I guarantee you're moving money around to do it. So who's the best possible sources for that? High-level financial people. So I don't know if either they were a Jason—and they were also academics. Okay. And a lot—what folks have to understand is when I—when people say like academia—it does not mean that you are just smart. Anyone who—who's good at studying could become a professor and be in academia. What a lot of folks should understand is that academia is a front for the government. It has always been a front for the government. Where do you think they headhunt from? Academia—well like—at the higher you get up the academic ladder—all you're really doing is getting more and more involved in the government. I'm not saying anything that anyone involved with this does not know. Like high-level academics are involved with the government. That's like the backbone of our system. Now a lot of the actual education—I think it's gotten a little out of hand with some of these majors, some of these colleges and universities who are offering [them]. That's not the point. The point is to create a—curate a talent pool to make the United States stronger. And a lot of it is government recruitment. Okay. And so Karen Read being all the way up at the top at Bentley—which is a very interconnected university with the government, trust me—that just makes me think she understands this—whether she was a Jason. Listen—you can understand what the intelligence community does without being in it. I'm not in the intelligence community—I just report on the government. So I kind of see how it all works. You can understand it without being in it. But if you're in it—let me just tell you right now—if anyone Karen Read knew professionally—through family or otherwise—is in the government—and I'm not talking about a special agent like in the FBI or, you know, a case officer—I'm talking about in the apparatus of control. Okay. In the directorate of national intelligence somewhere—there's a hierarchy. All right. If she knows anyone who understands all that—that's why she was able to pull this off. Because it's not—that's why I'm not fawning or being gratuitous. I don't necessarily—I'm not partial to Karen Read. I think she has liability for John's death. What I am is cognizant of what she's capable of—so I can understand what's going on. A mind like that, okay—doesn't just do PR. PR was not going to help Karen Read here. Natalie and her PR and all that stuff—none of that was going to work. What Karen Read needed was counter intel and intel knowledge. [Explanation of compartmentalization via the Manhattan Project] When I say compartmentalization—what you all have to realize is I'm talking about how the Manhattan Project—that's like the archetypal example of compartmentalization—how the Manhattan Project to develop the bomb that won the war for the United States in World War II—how that worked. The way that that worked is you had Los Alamos, okay, in New Mexico with Oppenheimer and whatever the hell—some of the Operation Paperclip people—which I'm not very happy with. We took Otto von Braun—who developed the V2 rocket for the Nazis. We brought him over via Operation Paperclip. We implanted him at Los Alamos with fucking Oppenheimer. I'm pretty sure it was like a communist sympathizer. Anyway—we sent them down to Los Alamos—the actual research scientists working on the core of the bomb. But to develop a nuclear bomb—you need 20,000 people at the time working simultaneously on production. You're not going to do that at Los Alamos. One: why would you ever expose them to the inner workings of the tech? It's nuclear material. You are not going to have 20,000 people around it. That's why it was in the middle of a desert. Third of all—they would know too much. So what did they do? Okay—look up Oak Ridge, Tennessee. Oak Ridge, Tennessee is a town—it's a government town still to this day. It's one of the most—it's not as top secret as it used to be. But back in the day—like during World War II—they put fucking canopies over the whole thing—20,000-person town—canopies over all of it. So it would just look like trees from the air in case the Japanese managed to come and bomb us. They never did—thank God. But anyway—at least on the mainland—obviously they got Pearl Harbor, and we're still upset about that. But the point is Oak Ridge, Tennessee, okay—it had people employed across a number of disciplines, all right—and they would go into—I'm giving you an example—one guy would go into a room, all right, and he would walk up to a giant computer. It was an old computer—we're talking the '40s here—it was a big computer, like a big green box. He would take a punch card. Okay—this is how you used to code—write computer code—he would put—take one punch card with one line of code—put it in the machine—take it out—put it down. He had no idea what he was doing. He didn't know what the punch card had on it. He didn't even understand what the machine did. That's compartmentalization. He's like—you do that with all of the people working on a project that's very top secret. So if you're thinking as Karen Read—Aidan Kearney does like a rudimentary version of it—even Tully does a rudimentary version of it—and Kate Peter—compared to Karen Read—Karen Read, Alan Jackson—whatever—understand the intelligence community. I don't fucking know how, but they do. So they compartmentalize. That's how they have pulled this whole thing off. They compartmentalize—no one ever really saw the full picture. When you—if you are a schematic mind like that—when you do something like reveal that Aidan Kearney has sent you a recording of conversations between you because you want the public to know that Aidan is doing this to you—you are tactically sabotaging him. Why you do that at this moment—when you are an expert in counter intel—thus requires that level of understanding. You cannot just say, "Oh, I don't like Karen Read; she must be a moron." No—if you want to understand why she's acting—you have to think about her tactical intelligence—because then you can reconstruct what the goal of this move is. It can only be designed to kneecap his support. I mean—when I say kneecap—I'm not talking Tonya Harding beating the woman at the ice rink. I'm talking—you can make it so this man's numbers are lower than Kate Peter talking about Cyraxx—like 2,000 views of video, if that, all right. That's what it will come down to. And she wants him to feel that. I think that it's a little bit like—it's 97% tactical. It seems to be that this is the moment where she feels he needs to lose all his support—like right now. Second—it feels a little like a little personal—like it's not just that she's causing him to lose all his support. There are ways to do that without doing this. I believe what Karen's really done here is she's taken the one thing from Aidan that gives him the strength to keep going day to day—which is his public persona and image—his support—her support. And so he can go around saying all he wants—"I owe it to Karen; she made me"—do you think that's how he really feels? Or do you think he feels that he's the only reason she's where she is? Now—if that's how they each feel—you're at a stalemate. Aidan thinks he's the reason Karen got to where she is. Karen thinks that she's the reason Aidan has support and is known in the region. Who's right? Who's right? That's what this conversation is going to be about. And I'm telling you—Karen's right. Karen has more support than Aidan Kearney. Okay—it's just a basic—you can look at the numbers. Karen has more support than Aidan Kearney. Karen has more loyalists. Karen—I don't even understand the complex, okay—but her parasocial complex that she's created is stronger. You might call it a brand—I think that degrades the insidious nature of it. I call it a parasocial complex. That is stronger than Aidan Kearney's. [Transition to the conversation with greetings] So what we're going to listen to right now is—oh, hello, Francesca Towel. Oh—a lot of folks are coming in. Hello, Rose Water. Hello, Maureen. It's great to see you all. Hello, J.I.5. We're going to listen to this conversation. I'm going to explicate it for you. I think you have enough background to get it now—but just be aware—without this background—that would have made no sense whatsoever. I promise you. **Aidan:** Am I on? **Host/Other:** You're on. **Aidan:** So who are you? Who is this? **Chris:** Don't worry about it. It doesn't matter who I am. **Aidan:** Well, it does. You're some fucking kangaroo court motherfucker talking about her. What the fuck do you know about anything? **Chris:** Well, I know exactly what you've been doing. **Aidan:** So—well, what the fuck are the sites you're talking about? No—no—recites—what you talk about? No—receipts. I've got to shit up. Let's see him. Let's fucking see your receipts. **Lily:** Hang on, Aidan—you know me. I'm the host. I'm Lily. **Aidan:** Yes, Lily. Hi, Lily. How are you? I'm sorry. I'm just wondering... **Lily:** I know you may not know Chris, but you know me. And so I just wanted to say hello. **Aidan:** Yeah, no, but... Grant: Oh, I also want to let you know this comes with a warning. They use very vulgar language. Some of them are from Commonwealth realms countries. So the language they're using is not as offensive as it would be if you used it in America. Aidan uses some very offensive language. This is for the purposes of analysis and commentary. I do not condone, endorse, promote, or otherwise suggest anyone engage in the use of this language. I personally don't like it. I use the F word from time to time, okay? And maybe like the S word—but I do not say some of the terms they're going to use—especially because one of them is very offensive, okay, to women. And I'm sorry ahead of time that he uses it—but you should hear Aidan's true colors. **Aidan:** This koala motherfucker is up here making shit up, running his... [Recap of text messages and setup for listening] Grant: So what you have to realize is these texts you're seeing on the screen got released because of this conversation. You're going to hear Joe Flipperhead talking about them. Now in the text messages, you can see Joe reaches out to Karen—Joe Flipperhead. And Karen's going to say she's trying to bounce back, but life is not quite happening. "Had a falling out with Aidan as everyone eventually does. Found out Aidan's been taping our phone conversations and sharing them with people and then telling everyone he doesn't understand how I blew him off for Howie Carr. Some anonymous person sent David and Alan a 33-minute phone call I had with Aidan that was all recorded without my knowledge. That was my final straw. He's done a lot of sneaky stuff with me, but this is above and beyond." And then Joe Flipperhead's like, "Do you want your side out there? If yes, I'm with you. If not, all good—just let me know. Have a good weekend." Karen says, "Sure. I told many people my side. This is the last straw. Would never and have never betrayed him. Meanwhile, he has put me in harm's way in a huge way multiple times." Okay, so we're going to listen. "Okay, okay, all right, all right—no trolling. We should—we should be banning people like that. You have been banned. You have been banned. No trolling. Absolutely no trolling. Now gaslighting and manipulative subversion is the hallmark of a lot of the forces in the orbit of this case. So none of that. We have a lot of Blue Wall of Towel friends here. Don't stand for that. Hello, Christy Mack. Great to see you. Hello. Stay tuned, Wendy. Hello, Bunny. Hello, F.B.I.—my friend, F.B.I. DOJ corruption survivor. And hello, Meredith—which is not Meredith O'Neill. This is Meredith the Towel friend. It's great to see you all. And as I said, if you see anybody trolling in the chat—now is not the day for it. Towel's health is not well. And I think there are a lot of people who want to undermine the agency of the unheard and the vulnerable in this situation. There are a lot of people who want to gaslight right now because where this is going is explosive. And furthermore, we're about to listen to the conversation. So what you're going to hear in this conversation is it's going to be Aidan Kearney and Joe Flipperhead—who's named Nick—and a guy named Chris who Aidan Kearney calls a koala. They're going to be talking about what we just talked about. But remember—these text messages haven't been released. So Aidan doesn't actually know they're coming. He's being told of this and going on an X Space and reacting in real time. Now I'm going to pause from time to time, and I'll try to flesh out some of the less clear parts. But as we read through the transcript and as you see this all, I think it will be clear to you—clear to you—the implications. So let's listen."

Grant Smith Ellis

19,048 Aufrufe • vor 10 Monaten

The July 4th weekend All-In The All-In Podcast turned into a long argument about who owns the intelligence layer. The besties think enterprises just woke up to a trap they had been walking into, here's how the conversation went (save this): ◽️ The Palantir-Nvidia deal is a bet against the model-layer duopoly. Palantir will use Nvidia's Nemotron open models to build a custom frontier-quality model for US government agencies, and the agencies own the hardware, the data, and the weights. Sacks framed it as structural: an application company and a chip company both want a competitive model layer, so they are natural partners against a two-provider middle. ◽️ Alex Karp's CNBC "crashout" was actually the thesis. Karp argued enterprises have lost trust in the frontier labs and want to own their compute, models, data, and alpha. Sacks translated it as a new definition of enterprise AI safety: safety means the model provider cannot hoover up your proprietary knowledge and turn it into its next product. ◽️ Figma is the cautionary tale that made it real. Anthropic launched Claude Design into Figma's category, its chief product officer sat on Figma's board and resigned only 3 days before launch, and Figma's stock is down about 50% this year while Anthropic's valuation surged. Sacks listed Claude Science, Security, Legal, Financial, and Code as the same move: dominate the model layer, then take the lucrative verticals. ◽️ The playbook has a name, and it is Microsoft and Google. Sacks argued Anthropic is running the operating-system strategy: own the layer everyone builds on, then walk up the stack. His Google receipt is that fewer than half of searches now send you off-site, versus an early Google that prided itself on how fast it kicked you away. ◽️ The BCG number is what raises the stakes. Chamath cited a BCG return-on-capital-employed study: the cost of capital is back to its long-run 8 to 11%, and half of large US companies cannot earn returns above it. If you are already teetering on your cost of capital, handing your alpha to a provider that may compete with you is not a luxury risk, it is fatal. ◽️ The 16.4x number is the whole argument in one data point. Chamath ran a code-migration task through 8090's harness. Wrapping Claude was 1.4x cheaper and 1.5x faster than Claude Opus alone. Wrapping the best open-source model was 16.4x cheaper, at about 3x slower. For a background task, three extra hours to cut cost by 16x is not a close call. ◽️ Even at 100x cheaper, enterprises were saying no for the wrong reason. Chamath relayed an ex-Meta PM's point that companies reject open models over China and safety fears, when they could host those same open weights on their own GPUs in US data centers with nothing flowing back. The safety objection, she argued, is backwards: the leak is the data you hand the frontier labs. ◽️ Friedberg says the frontier labs are trying to commoditize their own customers. Anthropic has been signing up life-sciences companies to feed a new life-focused model in exchange for early access, and nearly everyone he has talked to now refuses, recognizing that data they spent billions generating becomes worthless once it is pooled with everyone else's. ◽️ The deployment topology is shifting from big hubs to distributed spokes. Friedberg's map: the old assumption was a few capital-advantaged mega-clusters plus inference clouds. The new one is large hubs, medium hubs (enterprise training clusters), and distributed spokes, including on-prem inference in your own building. Owning your weights is the point. ◽️ Chamath's endgame is running GLM himself. An industry contact told him that with harness post-training and telemetry, an open Chinese model like GLM could get as good as Anthropic's Mythos. His conclusion: take GLM, control it soup-to-nuts on US hardware with only US citizens touching it, and pay a fraction. ◽️ The Apple analogy sharpens why renting intelligence is different from renting distribution. Chamath argued Apple is the only platform that respected developers, deliberately keeping its stock apps basic to protect the ecosystem and collect its 30% tax. There is no 30% tax on open models, and worse, you cannot rent intelligence from the same place that rents it to your competitor without ending up identical to them. ◽️ Nvidia's open model is now good enough to matter. Calacanis claimed you cannot tell Jensen Huang's Nemotron from Claude on 95% of searches, and that Nvidia downplayed the model until now to avoid alarming its top customers. The gloves came off once OpenAI, Anthropic, and Elon all signaled their own silicon ambitions. ◽️ Sacks sized the duopoly: roughly $60B and $40B in ARR. Anthropic is around ~$60 billion of ARR, OpenAI at ~$40 billion, and no one else generates meaningful model-layer revenue. Sacks's policy line: the US does not ban monopolies, only anti-competitive tactics, but the government should do nothing to make the duopoly more likely. ◽️ The token deflation call: 90% a year for three years. Calacanis predicted token costs fall 90% annually for three years, putting the price of intelligence near free and making it rational to waste tokens on hardware you already own. Friedberg's version is a 70/20/10 split between big cloud, local, and other clouds. ◽️ A wave of platform lock-in spending is already landing. Calacanis flagged Microsoft standing up a roughly $2.5 billion forward-deployed-engineer effort and Amazon spending about $1 billion on the same, plus OpenAI's version. His read: enterprises will slam the door, because letting a provider's engineers study your business is how it ends up in their model. ◽️ The server-per-employee prediction. Calacanis expects every employee to get $10,000 to $20,000 of local compute, a Mac Studio or a high-RAM Dell, running a personal local model that syncs to a thin laptop. A server per person, so nothing leaks. ◽️ On jobs, the data does not show present-tense loss. Sacks cited a RAMP and Revelio Labs study of over 21,000 US firms: the heaviest AI spenders grew headcount about 10% over two years, and entry-level headcount grew even faster at 12%. Friedberg's harder claim: there is no AI job loss yet, only clunky, gradual value creation, and the media will not reverse its narrative because that destroys its credibility. ◽️ The displacement case is real but forward-dated. The counterpoint on the show was that customer support, entry-level data entry and BPO, and driving are the near-term displacements, with Waymo cited as present-tense evidence: in markets where it hits critical mass, Uber and Lyft stop recruiting drivers. Sacks noted most US entry-level support was already offshored, so the acute risk sits in those countries first. ◽️ The human-premium counternarrative. Friedberg argued that as automation spreads, human interaction gets a premium: the skilled bartender, the real driver, the human-in-the-loop tier. He cited the company (referenced as Klarna) that hyped replacing its whole support team with AI, then reversed a year later on brand grounds. ◽️ The export-control episode needed three conditions, and Sacks says do not over-read it. Commerce lifted controls on Anthropic's Fable 5 after two weeks, with Mythos 5 restored to US customers around June 26 once co-founder Tom Brown replaced Dario as lead negotiator. Sacks's three conditions: Dario boasting for months about a cyber weapon, Amazon reporting failed guardrails in testing, and Dario refusing to roll Fable back. His message to allies: this was a particular set of circumstances rather than the debut of a standing lever. ◽️ The import question nobody answered cleanly. Calacanis pressed on why the US blocks Chinese cars and drones but not Chinese open models like DeepSeek and Kimi. Sacks's answer: a forked open model run on US hardware stops being Chinese, and banning open source would isolate the US and impose a token tax on American enterprises, so let the market decide if American open models win. ◽️ The California fiscal story is a business-climate story. Friedberg walked through the numbers behind Newsom's "balanced" $351B budget: expenses exceed revenue and $20-40B is borrowed to close the gap, the budget grew 65% in six years ($215B to $355B), personal income tax is $142B of ~$211B revenue with the top 1% (150,000 people) paying $70B of it, and the corporate rate of 8.9% sits far above Texas at zero. ◽️ The tax base is leaving, and the state is now taxing everyone else. Friedberg cited 1 to 1.5% of adjusted gross income leaving each year (about 15% over a decade), at least 15 Fortune 500 HQs and ~2,100 firms gone since 2019, and a new 8% software sales tax hitting Word, Gmail, and ChatGPT subscriptions plus a health-insurance tax, on top of a now-permanent 14.4% top bracket. The liabilities behind it run $1.4T in debt, up to $1.5T in unfunded pensions senior to state bonds, and ~$40B/year in out-year deficits. Lastly, the line that framed the whole show: "You can't rent intelligence from the same place that rents it to your competitor." That is the sovereignty thesis in one sentence, and every number in this episode is an argument for it. ____ Follow Fireside Alpha for more summaries on key business and technology conversations.

Fireside Alpha

55,816 Aufrufe • vor 1 Monat

$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 Aufrufe • vor 8 Monaten