Video wird geladen...

Video konnte nicht geladen werden

Beim Laden dieses Videos ist ein Problem aufgetreten. Dies könnte an einem vorübergehenden Netzwerkproblem liegen oder das Video ist möglicherweise nicht verfügbar.

We cut open the Kirin 9030 and put it under an electron microscope. The smallest metal pitch measures 32.5 nanometers. That is tighter than Intel 18A, their brand new leading edge node. A Chinese fab with no EUV is out-pitching Intel's EUV node by roughly 10 percent. What's going... on here? 🤯 "This is the HiSilicon Kirin 9030, the chip inside Huawei's newest flagship phone. A few weeks ago, we cut it open, put it under an electron microscope, and measured the smallest wires inside the chip, the metal pitch. And what we saw was unexpected. The smallest metal pitch inside the Kirin 9030 measures only 32.5 nanometers. That’s smaller than the metal pitch in Panther Lake, which is based on Intel's brand-new 18A node. A Chinese fab, cut off from the most advanced tools, without EUV, is packing its wires about ten percent tighter than Intel's leading edge EUV node. What’s going on here?"show more

SemiAnalysis

173,680 subscribers

1,017,918 Aufrufe • vor 2 Monaten •via X (Twitter)

Wissenschaft & Technologie

Anya Rossi• Live Now

Private livecam show

37 Kommentare

Profilbild von Black Metal Trading

Black Metal Tradingvor 2 Monaten

AI-experts on X: THEY DISTILLED INTEL CODE COPY THEFT BAN COMMUNISM !!!!

Profilbild von SemiAnalysis

SemiAnalysisvor 2 Monaten

Full breakdown here:

Profilbild von Barrett

Barrettvor 2 Monaten

I’ve said so many times “do not underestimate China”

Profilbild von Jeff Teza

Jeff Tezavor 2 Monaten

Semi companies built small features for years using multi-patterning. The issue is mask and processing costs (throughput) vs. EUV and CAD tool support. We built nm dimension devices to characterize at TI in the 1980's using tricks. There are many ways to skin a cat ;)

Profilbild von Rufus Rutilius

Rufus Rutiliusvor 2 Monaten

cc @CommunityNotes SemiAnalysis' teardown confirms the 32.5 nm M0 metal pitch on SMIC N+3 (Kirin 9030), tighter than Intel 18A's shipped 36 nm pitch. However, their full analysis notes this is a cherry-picked metric. Overall transistor density on N+3 (~113 MTr/mm²) trails Intel 18A by ~38% and is similar to TSMC N6. Intel's looser M0 pitch is a deliberate choice enabled by PowerVia backside power.

Profilbild von (✿◡‿◡)

(✿◡‿◡)vor 2 Monaten

it's built on SMIC's 5nm class N+3 process using flat planar design. Kirin 2026 will use logical folding.

Profilbild von Thomas Broadfoot 🍁👻

Thomas Broadfoot 🍁👻vor 2 Monaten

12nm, a ten year old tech has 32nm pitch. What’s up with that?

Profilbild von Michael Koeris

Michael Koerisvor 2 Monaten

Are they really cut off from EUV?

Profilbild von SemiAnalysis

SemiAnalysisvor 2 Monaten

Yes, SMIC uses DUV

Profilbild von Michael Koeris

Michael Koerisvor 2 Monaten

Very interesting …. Hm hm hm

Profilbild von von

vonvor 2 Monaten

When a problem cannot be explained, there is another answer: China copied American technology.

Profilbild von C Y

C Yvor 2 Monaten

Small metal pitch is relatively easy to do if you pair it with old generation transistor technology.

Profilbild von carbonat🇦🇩

carbonat🇦🇩vor 2 Monaten

Before getting any wrong conclussions bassed on the headline alone, please watch the video. It's not that rosy.

Profilbild von Noonien Soong

Noonien Soongvor 2 Monaten

They must have distilled it.

Profilbild von FrontierNotes

FrontierNotesvor 2 Monaten

as most in foundry industry knows, you don’t necessarily need EUV to achieve smaller pitch. you just need the right process schemes to achieve this

Profilbild von Ju_Jason

Ju_Jasonvor 2 Monaten

两个月后华为会发布麒麟9050，那个才是正式采用他们最新的技术-韬定律的第一款芯片。

Profilbild von energyguy123456

energyguy123456vor 2 Monaten

@highyieldYT This is a clickbait title. You guys are better than that.

Profilbild von Venkat Dukkipati

Venkat Dukkipativor 2 Monaten

Need semiconductor experience to answer this. EUV is not required for metal lines and is only used for transistor layers which is made of different layer/material. Also metal lines do not define technology node.

Profilbild von SemiVision 🇹🇼 👁️👁️

SemiVision 🇹🇼 👁️👁️vor 2 Monaten

No EUV is fine

Profilbild von ActionDan

ActionDanvor 2 Monaten

What's going on here? @grok did semianalysis provide the answer somewhere?

Profilbild von 𝒘𝒐𝒓𝒍𝒅 𝒃𝒆𝒈𝒂𝒏 𝒕𝒐 𝒃𝒖𝒓𝒏

𝒘𝒐𝒓𝒍𝒅 𝒃𝒆𝒈𝒂𝒏 𝒕𝒐 𝒃𝒖𝒓𝒏vor 2 Monaten

At the end of the day, this is still just pushing DUV multi-patterning to its absolute limit — not a real leap into the EUV era

Profilbild von Scrooge McDuck

Scrooge McDuckvor 2 Monaten

So maybe they have EUV

Profilbild von Akane

Akanevor 2 Monaten

18Aは1.8nmです積層ピッチ32nmと言いますが敗戦幅と積層ピッチは関係ありません EUVで小さくなるのは配線幅であって積層ピッチではありませんナンセンス

Profilbild von Cool Master

Cool Mastervor 2 Monaten

Maybe China got their EUV machines via SGP or Thailand?

Profilbild von Venkman Capital

Venkman Capitalvor 2 Monaten

Well 18A can go down to 32nm in the future (maybe now) and if you're doing 32.5 with DUV tools that could take quad patterning (or more) with horrible yields

Profilbild von kalv cell

kalv cellvor 2 Monaten

Maybe, maybe not. Even if they packed it as tight as you observed using whatever methodology, what's their yield?

Profilbild von Makoto | .....

Makoto | .....vor 2 Monaten

But, but, Chinese can only copy, right? Right?!

Profilbild von whoooooo

whoooooovor 2 Monaten

可能是我过拟合了，但毫无疑问答案是他们蒸馏了Intel

Profilbild von Kokolomoto

Kokolomotovor 2 Monaten

Chinese have CUV

Profilbild von Shane

Shanevor 2 Monaten

I don't believe this story that is portrayed that FABs are hard to build. I'm in a trade where it's hands on with electronics and machines. I don't see how making chips is that hard, yea it is hard but most of it is theory work, once the system is in place normal employees work in these FABs. It's just light burning channels in silicon then vapour despostion of metals in a chamber. Then acid etch and onto the next layer.

Profilbild von Sean

Seanvor 2 Monaten

We gotta get you a lesson in pronouncing Chinese brand names. Huawei is a hard one but I know you've got this.

Profilbild von Atomist_No.2💣

Atomist_No.2💣vor 2 Monaten

@lithos_graphein 👀!

Profilbild von Vortex

Vortexvor 2 Monaten

Out pitching an EUV node without EUV is impressive engineering.

Profilbild von beefcube (execute/commies)⚔️

beefcube (execute/commies)⚔️vor 2 Monaten

the question is what are the yields? you take the one that works well, but how many defects were made in order to make a working one?

Profilbild von 情報収集っぽい ^_^

情報収集っぽい ^_^vor 2 Monaten

a 32nm metal pitch doesn't mean anything

Profilbild von 41kingway

41kingwayvor 2 Monaten

华为蒸馏了 EUV

Profilbild von Bondan Rufen

Bondan Rufenvor 2 Monaten

It may be the yield issue. We don't how good the yield of Kirin. You can make very tight pitch, but if yield is very low, it doesn't make sense.

Ähnliche Videos

ASML's EUV machine is one of the most complex machines ever built by humanity. Some *insane* stats about it: → Without EUV machines, it's impossible to manufacture cutting-edge chips for things like $AAPL iPhones, $NVDA GPUs, and $AMZN data centers. → $ASML is the *only* producers of EUV machines in the world. → Shipping a single EUV machine requires 40 freight containers, three cargo planes, and 20 trucks. → Each EUV machine costs around $200M, consists of more than 100,000 parts and weighs 200 tonnes. → The mirrors inside (supplied by Zeiss) are said to be the flattest objects ever produced. If scaled up to the size of Germany, the largest unevenness would be a tenth of a millimeter. → The next generation High-NA EUV systems have been shipped to $INTC and $TSM and costs closer to $400M. This is human ingenuity at its best. Thanks Intel for the fab tour!

ASML's EUV machine is one of the most complex machines ever built by humanity. Some insane stats about it: → Without EUV machines, it's impossible to manufacture cutting-edge chips for things like $AAPL iPhones, $NVDA GPUs, and $AMZN data centers. → $ASML is the only producers of EUV machines in the world. → Shipping a single EUV machine requires 40 freight containers, three cargo planes, and 20 trucks. → Each EUV machine costs around $200M, consists of more than 100,000 parts and weighs 200 tonnes. → The mirrors inside (supplied by Zeiss) are said to be the flattest objects ever produced. If scaled up to the size of Germany, the largest unevenness would be a tenth of a millimeter. → The next generation High-NA EUV systems have been shipped to $INTC and $TSM and costs closer to $400M. This is human ingenuity at its best. Thanks Intel for the fab tour!

Quartr

76,853 Aufrufe • vor 1 Jahr

EUV machines are the most complicated tools humans make. Their supply chain has over 10,000 individual suppliers, and any one of them not scaling fast enough can bottleneck the entire AI industry. An EUV tool fires lasers at a tiny tin droplet three times in precise sequence, blasting it hard enough to emit EUV light. That light bounces off 18 multilayer mirrors onto the wafer. Meanwhile, the two platforms inside the machine - one holding the stencil, one holding the chip - are flying back and forth at 9Gs in opposite directions. The successive passes have to land on top of each other to within 3 nanometers. If any part of this is off, yield goes to zero. Take just one component. The mirrors are mostly supplied by Carl Zeiss, who have probably fewer than a thousand people working on them. In turn, Carl Zeiss rely on machines from Switzerland to deposit each of the layers, and use a coating process co-developed with a different German company. None of these companies have woken up. They’re gradually increasing production, but nowhere near the levels necessary for what the labs want by the end of the decade. Dylan Patel predicts production can't scale beyond about 100 EUV machines per year by 2030, no matter how much money gets thrown at the problem. In the medium term this is the key bottleneck on scaling.

EUV machines are the most complicated tools humans make. Their supply chain has over 10,000 individual suppliers, and any one of them not scaling fast enough can bottleneck the entire AI industry. An EUV tool fires lasers at a tiny tin droplet three times in precise sequence, blasting it hard enough to emit EUV light. That light bounces off 18 multilayer mirrors onto the wafer. Meanwhile, the two platforms inside the machine - one holding the stencil, one holding the chip - are flying back and forth at 9Gs in opposite directions. The successive passes have to land on top of each other to within 3 nanometers. If any part of this is off, yield goes to zero. Take just one component. The mirrors are mostly supplied by Carl Zeiss, who have probably fewer than a thousand people working on them. In turn, Carl Zeiss rely on machines from Switzerland to deposit each of the layers, and use a coating process co-developed with a different German company. None of these companies have woken up. They’re gradually increasing production, but nowhere near the levels necessary for what the labs want by the end of the decade. Dylan Patel predicts production can't scale beyond about 100 EUV machines per year by 2030, no matter how much money gets thrown at the problem. In the medium term this is the key bottleneck on scaling.

Dwarkesh Patel

154,462 Aufrufe • vor 6 Monaten

ACCELERATOR-BASED LITHOGRAPHY AND THE INDUCTION STORAGE RING LIGHT SOURCE The future of chipmaking may look less like a factory and more like a power grid. The proposed “Terafab” paradigm reimagines semiconductor manufacturing at utility scale: chip design, wafer fabrication, EUV lithography, memory, advanced packaging, and testing all under one roof, with the ambition of producing more than one terawatt of AI compute capacity per year. Its most radical innovation is treating light as a utility. Today, each EUV scanner relies on its own laser-produced plasma source, firing lasers at molten tin to generate 13.5 nm light. The process is inefficient, intensely hot, debris-heavy, and difficult to scale. Terafab replaces those individual sources with centralized, accelerator-driven free-electron lasers capable of distributing multi-kilowatt EUV light across an entire network of scanners. The advantages could be transformative: • No tin contamination or destructive plasma debris • Higher efficiency through electron-beam energy recovery • Greater photon flux to suppress stochastic defects at sub-3 nm nodes • Redundant accelerators that keep scanners operating during maintenance • Tunable wavelengths, potentially enabling 6.x nm “Beyond EUV” lithography THAT LAST POINT MATTERS ENORMOUSLY Conventional EUV is locked to the atomic emission of tin. Free-electron lasers are not. Their wavelength can be tuned through the energy of the electron beam creating a possible path beyond today’s 13.5 nm limit. If realized, Terafab would represent more than a larger semiconductor plant. It would transform lithography from a collection of isolated tools into shared industrial infrastructure. The next era of chipmaking may not be defined by a better machine. It may be defined by an entirely new architecture for manufacturing intelligence at civilization scale.

ACCELERATOR-BASED LITHOGRAPHY AND THE INDUCTION STORAGE RING LIGHT SOURCE The future of chipmaking may look less like a factory and more like a power grid. The proposed “Terafab” paradigm reimagines semiconductor manufacturing at utility scale: chip design, wafer fabrication, EUV lithography, memory, advanced packaging, and testing all under one roof, with the ambition of producing more than one terawatt of AI compute capacity per year. Its most radical innovation is treating light as a utility. Today, each EUV scanner relies on its own laser-produced plasma source, firing lasers at molten tin to generate 13.5 nm light. The process is inefficient, intensely hot, debris-heavy, and difficult to scale. Terafab replaces those individual sources with centralized, accelerator-driven free-electron lasers capable of distributing multi-kilowatt EUV light across an entire network of scanners. The advantages could be transformative: • No tin contamination or destructive plasma debris • Higher efficiency through electron-beam energy recovery • Greater photon flux to suppress stochastic defects at sub-3 nm nodes • Redundant accelerators that keep scanners operating during maintenance • Tunable wavelengths, potentially enabling 6.x nm “Beyond EUV” lithography THAT LAST POINT MATTERS ENORMOUSLY Conventional EUV is locked to the atomic emission of tin. Free-electron lasers are not. Their wavelength can be tuned through the energy of the electron beam creating a possible path beyond today’s 13.5 nm limit. If realized, Terafab would represent more than a larger semiconductor plant. It would transform lithography from a collection of isolated tools into shared industrial infrastructure. The next era of chipmaking may not be defined by a better machine. It may be defined by an entirely new architecture for manufacturing intelligence at civilization scale.

Lacey

13,405 Aufrufe • vor 2 Monaten

OK SO FINALLY I have a means to show you what I have been saying for a while.......Watch this clip, share it, so people understand what we are saying here!!! In this Clip, JLR© is having a discussion with Watching the World , and we get to see that front door lock!! Here is what I have been saying.....The Front door is a material called EXPANDED METAL, YOU CAN SEE THROUGH IT. The Door IS a Security Door, that you would have to cut through with two locks. A Lock that can be unlocked via a key on the handle, and a DEADbolt lock that would be locked on the inside of the door. This PREVENTS someone from being able to attack you when you OPEN the HOUSE door that is inside it, and behind this screen door. That door only opens when that deadbolt is unlocked from inside! ALso notice we see NO damage to the door or frame in this video!!

OK SO FINALLY I have a means to show you what I have been saying for a while.......Watch this clip, share it, so people understand what we are saying here!!! In this Clip, JLR© is having a discussion with Watching the World , and we get to see that front door lock!! Here is what I have been saying.....The Front door is a material called EXPANDED METAL, YOU CAN SEE THROUGH IT. The Door IS a Security Door, that you would have to cut through with two locks. A Lock that can be unlocked via a key on the handle, and a DEADbolt lock that would be locked on the inside of the door. This PREVENTS someone from being able to attack you when you OPEN the HOUSE door that is inside it, and behind this screen door. That door only opens when that deadbolt is unlocked from inside! ALso notice we see NO damage to the door or frame in this video!!

Capt KEVman!!!

46,092 Aufrufe • vor 7 Monaten

SITUATION EXPLAINED: A Chinese state-backed company started mass-producing DUV lithography tools, a step below EUV but still significant. • ASML shares fell as much as 6.5%, their lowest since early June, after The Information reported a Shanghai-based, state-backed company began mass-producing immersion DUV lithography machines • The company is Shanghai Yuliangsheng, with ties to Huawei and the SiCarrier equipment group, it brought together immersion DUV development teams from other Chinese firms • Plans call for 5 tools this year and 20 next year, with confirmed customers SMIC, CXMT, and Hua Hong Semiconductor • Real caveat: SMIC has actually been trialing this tool since September 2025, and active mass production isn't targeted until 2027 at the earliest • Independent analysis from the AI Futures Project puts commercial-scale Chinese immersion DUV in the mid-2030s, with ASML still holding 98.7% of the immersion lithography market • Chinese chipmakers are currently only allowed to buy ASML's older DUV tools, not its cutting-edge EUV machines • The MATCH Act, moving through Congress, would widen restrictions specifically on immersion DUV equipment • China is separately developing its own domestic EUV machine, but that project remains at the prototype stage and is likely years away from producing working chips • Nikon and Canon are the only other established DUV/lithography vendors outside ASML, alongside SMEE as another Chinese domestic challenger Theo Jaffee: "Someone has to own the bottom of the market. Chinese DUV would be the same thing for ASML. The frontier leading-edge chip fabs will still be using ASML EUV machines, but less leading-edge fabs would use DUV machines, like, for example, it would be really helpful for Huawei."

SITUATION EXPLAINED: A Chinese state-backed company started mass-producing DUV lithography tools, a step below EUV but still significant. • ASML shares fell as much as 6.5%, their lowest since early June, after The Information reported a Shanghai-based, state-backed company began mass-producing immersion DUV lithography machines • The company is Shanghai Yuliangsheng, with ties to Huawei and the SiCarrier equipment group, it brought together immersion DUV development teams from other Chinese firms • Plans call for 5 tools this year and 20 next year, with confirmed customers SMIC, CXMT, and Hua Hong Semiconductor • Real caveat: SMIC has actually been trialing this tool since September 2025, and active mass production isn't targeted until 2027 at the earliest • Independent analysis from the AI Futures Project puts commercial-scale Chinese immersion DUV in the mid-2030s, with ASML still holding 98.7% of the immersion lithography market • Chinese chipmakers are currently only allowed to buy ASML's older DUV tools, not its cutting-edge EUV machines • The MATCH Act, moving through Congress, would widen restrictions specifically on immersion DUV equipment • China is separately developing its own domestic EUV machine, but that project remains at the prototype stage and is likely years away from producing working chips • Nikon and Canon are the only other established DUV/lithography vendors outside ASML, alongside SMEE as another Chinese domestic challenger Theo Jaffee: "Someone has to own the bottom of the market. Chinese DUV would be the same thing for ASML. The frontier leading-edge chip fabs will still be using ASML EUV machines, but less leading-edge fabs would use DUV machines, like, for example, it would be really helpful for Huawei."

MTS

16,220 Aufrufe • vor 2 Monaten

$Elon Musk just put a number on the flaw at the center of Nvidia’s empire. Wall Street has not done the math yet. Nvidia’s Blackwell is the most sought-after silicon on Earth. Every AI lab wants it. Every sovereign nation is bidding for it. Blackwell runs every model, for every company, in every data center on the planet. That universality built the empire. It is also the fracture point. Musk: “We believe the AI5 chip will be about a third of the power of an Nvidia Blackwell for roughly comparable performance. And much less than 10% of the cost.” One-third the power. Comparable performance. Less than ten percent of the cost. Musk: “This is a chip that is very much optimized for the Tesla AI software stack. It’s not meant to be a general purpose chip.” Nvidia builds silicon that serves a million different customers. Every transistor spent on universal compatibility is a transistor not dedicated to one task. Tesla is building silicon for exactly one customer. Itself. When you strip away every function you will never call, you do not get a lesser chip. You get a weapon. Here is what the market refuses to see. Data centers drink unlimited power from the grid. Robots run on batteries. Musk: “In order to have a functional robot, you have to have a great AI chip. And it needs to be an inexpensive chip and it needs to be very power efficient.” You cannot put a Blackwell inside a walking machine. It would drain the battery before it crossed the room. The entire AI revolution lives inside air-conditioned buildings bolted to the electrical grid. Musk is not competing for that market. He is engineering the silicon that survives outside of it. One-third the power is not a spec sheet footnote. It is the physics threshold that severs intelligence from the wall socket. Without that number, every robot on Earth stays tethered. With it, the algorithm walks. Less than ten percent of the cost is not a pricing strategy. It is the line where a machine brain stops being a capital expenditure and becomes a commodity component. When the chip inside a humanoid costs less than the motors in its legs, you do not manufacture hundreds of robots. You manufacture millions. Wall Street is valuing the AI revolution by who dominates the data center. Musk is building the only silicon designed to leave one. Nvidia built the brain of the cloud. Musk is building the brain of the physical world. No one has priced that in yet.$

Elon Musk just put a number on the flaw at the center of Nvidia’s empire. Wall Street has not done the math yet. Nvidia’s Blackwell is the most sought-after silicon on Earth. Every AI lab wants it. Every sovereign nation is bidding for it. Blackwell runs every model, for every company, in every data center on the planet. That universality built the empire. It is also the fracture point. Musk: “We believe the AI5 chip will be about a third of the power of an Nvidia Blackwell for roughly comparable performance. And much less than 10% of the cost.” One-third the power. Comparable performance. Less than ten percent of the cost. Musk: “This is a chip that is very much optimized for the Tesla AI software stack. It’s not meant to be a general purpose chip.” Nvidia builds silicon that serves a million different customers. Every transistor spent on universal compatibility is a transistor not dedicated to one task. Tesla is building silicon for exactly one customer. Itself. When you strip away every function you will never call, you do not get a lesser chip. You get a weapon. Here is what the market refuses to see. Data centers drink unlimited power from the grid. Robots run on batteries. Musk: “In order to have a functional robot, you have to have a great AI chip. And it needs to be an inexpensive chip and it needs to be very power efficient.” You cannot put a Blackwell inside a walking machine. It would drain the battery before it crossed the room. The entire AI revolution lives inside air-conditioned buildings bolted to the electrical grid. Musk is not competing for that market. He is engineering the silicon that survives outside of it. One-third the power is not a spec sheet footnote. It is the physics threshold that severs intelligence from the wall socket. Without that number, every robot on Earth stays tethered. With it, the algorithm walks. Less than ten percent of the cost is not a pricing strategy. It is the line where a machine brain stops being a capital expenditure and becomes a commodity component. When the chip inside a humanoid costs less than the motors in its legs, you do not manufacture hundreds of robots. You manufacture millions. Wall Street is valuing the AI revolution by who dominates the data center. Musk is building the only silicon designed to leave one. Nvidia built the brain of the cloud. Musk is building the brain of the physical world. No one has priced that in yet.

Dustin

160,633 Aufrufe • vor 6 Monaten

A copper saw can't cut granite on its own. The metal is far softer than the stone. The trick is that the blade was never really doing the cutting. Stonemason Roger Hopkins tried copper chisels on granite for NOVA and found he was losing a lot of metal while very little stone came off. Tool specialist Denys Stocks, who spent about two decades recreating Egyptian tools, came at it differently. Pour sand into the groove, set a copper blade on top, and drag it back and forth. Quartz sand is about as hard as granite, so the grains grind the groove deeper with every stroke. It works, slowly. In Stocks' tests, wetting the sand didn't make it cut any faster, it just made the saw harder to pull. Rough quarrying was even blunter. At the Unfinished Obelisk in Aswan, archaeologists believe quarrymen pounded trenches with lumps of dolerite and hauled the dust out in baskets. Before seeing this, what did you assume the Egyptians used to cut granite?

A copper saw can't cut granite on its own. The metal is far softer than the stone. The trick is that the blade was never really doing the cutting. Stonemason Roger Hopkins tried copper chisels on granite for NOVA and found he was losing a lot of metal while very little stone came off. Tool specialist Denys Stocks, who spent about two decades recreating Egyptian tools, came at it differently. Pour sand into the groove, set a copper blade on top, and drag it back and forth. Quartz sand is about as hard as granite, so the grains grind the groove deeper with every stroke. It works, slowly. In Stocks' tests, wetting the sand didn't make it cut any faster, it just made the saw harder to pull. Rough quarrying was even blunter. At the Unfinished Obelisk in Aswan, archaeologists believe quarrymen pounded trenches with lumps of dolerite and hauled the dust out in baskets. Before seeing this, what did you assume the Egyptians used to cut granite?

Epic Vault

81,487 Aufrufe • vor 15 Tagen

Luxon butchering Te Reo before he launches into one of his woke progressive speeches is like hearing the pop and hiss of cracking open a can of Bud Light. Cam Slater has written today that the Nats are helicoptering in Paula Bennet to stab him from the front while his MP’s pat him on the back - in order to feel where to stick their knives. Luxon’s pathetic political career is coming to an end, so how about a retrospective ? “Chris Luxon this is your life !” Everyone with half a brain knew what he was going to be like. Did those who voted for him watch his maiden speech? Here it is if you missed it May 24th 2021. It’s truly a speech that perhaps only Ardern herself would be so cynical to deliver with a straight face. Here’s the break down of the 4 minute speech. * Tay Rayo introduction * Diversity pitch * Immigration pitch * Diversity pitch * Climate change pitch * Globalism pitch * Air NZ career pitch * Globalism Pitch * Diversity pitch * Critical race theory pitch * Gender ideology pitch * Diversity pitch * Rainbow tick pitch * Getting things done pitch * Getting things done pitch * Diversity pitch * Getting diversity done pitch * voters will get what they deserve pitch So who remembers this speech? And yes it was THIS BAD, no joke. Maybe everyone who voted National was just so keen to dump Labour they fooled themselves that Luxon was the panacea? Also Trump rising from the ashes has put the spot light on a diminutive man with an even smaller appreciation of what the average sane kiwi wants. They can change him out, for a cosmetic attempt at pivoting but that party will never change - at the core of its fundamental philosophy is - woke nothingisim.

Luxon butchering Te Reo before he launches into one of his woke progressive speeches is like hearing the pop and hiss of cracking open a can of Bud Light. Cam Slater has written today that the Nats are helicoptering in Paula Bennet to stab him from the front while his MP’s pat him on the back - in order to feel where to stick their knives. Luxon’s pathetic political career is coming to an end, so how about a retrospective ? “Chris Luxon this is your life !” Everyone with half a brain knew what he was going to be like. Did those who voted for him watch his maiden speech? Here it is if you missed it May 24th 2021. It’s truly a speech that perhaps only Ardern herself would be so cynical to deliver with a straight face. Here’s the break down of the 4 minute speech. * Tay Rayo introduction * Diversity pitch * Immigration pitch * Diversity pitch * Climate change pitch * Globalism pitch * Air NZ career pitch * Globalism Pitch * Diversity pitch * Critical race theory pitch * Gender ideology pitch * Diversity pitch * Rainbow tick pitch * Getting things done pitch * Getting things done pitch * Diversity pitch * Getting diversity done pitch * voters will get what they deserve pitch So who remembers this speech? And yes it was THIS BAD, no joke. Maybe everyone who voted National was just so keen to dump Labour they fooled themselves that Luxon was the panacea? Also Trump rising from the ashes has put the spot light on a diminutive man with an even smaller appreciation of what the average sane kiwi wants. They can change him out, for a cosmetic attempt at pivoting but that party will never change - at the core of its fundamental philosophy is - woke nothingisim.

Holyhekatuiteka

11,074 Aufrufe • vor 1 Jahr

UNUSUAL A new Chinese chip is pushing the West out of the AI competition. Researchers in China have developed an AI chip that outperforms the Nvidia A100 by 478 times. A team from Peking University created a new neuromorphic chip capable of reconstructing complex brain surfaces in less than half a second, achieving performance up to 478 times greater than the Nvidia A100 in this specialized task. This advantage comes from a different architectural design: the chip integrates storage and processing within the same memory array, reducing the data‑transfer bottleneck seen in traditional processors, lowering energy consumption, and boosting efficiency in targeted applications. The achievement is also notable because the chip was manufactured using a 40‑nanometer process — a much older technology node — demonstrating that high performance can be reached through architectural innovation rather than relying on cutting‑edge fabrication. Observers say this development reflects China’s growing strength in creating innovative domestic solutions in semiconductors and artificial intelligence, and its progress toward advanced next‑generation chips.

UNUSUAL A new Chinese chip is pushing the West out of the AI competition. Researchers in China have developed an AI chip that outperforms the Nvidia A100 by 478 times. A team from Peking University created a new neuromorphic chip capable of reconstructing complex brain surfaces in less than half a second, achieving performance up to 478 times greater than the Nvidia A100 in this specialized task. This advantage comes from a different architectural design: the chip integrates storage and processing within the same memory array, reducing the data‑transfer bottleneck seen in traditional processors, lowering energy consumption, and boosting efficiency in targeted applications. The achievement is also notable because the chip was manufactured using a 40‑nanometer process — a much older technology node — demonstrating that high performance can be reached through architectural innovation rather than relying on cutting‑edge fabrication. Observers say this development reflects China’s growing strength in creating innovative domestic solutions in semiconductors and artificial intelligence, and its progress toward advanced next‑generation chips.

China pulse 🇨🇳

31,271 Aufrufe • vor 3 Monaten

$AAOI SKYLIT RECAP Replay tool is one of the best features Skylit has to offer Single handedly the best way to learn, backtest, and understand how gamma works Watch below’s $AAOI incredible +17% move caught on replay. Below is my explanation as to why having the map could have helped you play this today: 1st node in the way of the stock price going up, $105 node🟢. Notice how in morning, price was struggling here, and on the third attempt, finally broke through Once $105 was broken, notice the strike’s volume from $105–> $110, almost NOTHING. This is what we call a GAP between nodes, meaning price will move QUICK 🚀 from node to node here, look at the chart and see how clean the move from $105 to $110 was.. Now add this being a negative gamma node to upside, the move is even more quick Now at $110, purple🟣negative node, notice how it takes a bit to break this, that is because despite it being negative, it can also act as a roadblock While this is trying to break $110, they slip in a nice brand new position at $120 👀 Once $110 was broken see how quick $115 came? Once again, because the low volumes on those strikes between two major nodes Towards EOD, $115 faded away, $120 gained, and we hit the final price target during after hours✅

Giul

15,309 Aufrufe • vor 6 Monaten

A guy put an XFRA box outside his house and now gets paid $2,500 a month for doing nothing The box is about the size of an AC unit and mounts right on the wall. Inside it connects your home to a distributed AI compute network. Companies pay to run their AI workloads through it around the clock. He just plugged it in and that was it. Now $2,500 lands in his account every single month. That is $30,000 a year from a box he never touches. It is not Nvidia. It is not a government program. It is XFRA and regular homes are quietly becoming AI infrastructure. Every house a node. Every node earning. The people who move first lock in the best rates. Save this before your neighborhood figures it out before you.

A guy put an XFRA box outside his house and now gets paid $2,500 a month for doing nothing The box is about the size of an AC unit and mounts right on the wall. Inside it connects your home to a distributed AI compute network. Companies pay to run their AI workloads through it around the clock. He just plugged it in and that was it. Now $2,500 lands in his account every single month. That is $30,000 a year from a box he never touches. It is not Nvidia. It is not a government program. It is XFRA and regular homes are quietly becoming AI infrastructure. Every house a node. Every node earning. The people who move first lock in the best rates. Save this before your neighborhood figures it out before you.

winkle.

16,795 Aufrufe • vor 4 Monaten

🇺🇸 ELON KILLED THE SMARTPHONE: "WE'RE NOT GOING TO HAVE A PHONE IN THE TRADITIONAL SENSE" No, Tesla isn't building a phone. They're leapfrogging the entire concept. The device in your pocket won't be a phone anymore. It'll be a super-thin AI edge node, basically a screen with radios that pings Grok/xAI servers for EVERYTHING in real time. No apps. No operating system. No local processing. Just instant photorealistic video generation, simulations, predictive actions (it orders your coffee before you know you want it), all streamed from the cloud like your brain is directly plugged into the AI. Exact quote that broke the internet: “We’re not going to have a phone in the traditional sense. What we’ll call a phone will really be an edge node for AI inference with some radios to connect.” This is already how Grok works in Tesla cars: voice in, server crunches, perfect response out. Now imagine that in your pocket, on steroids. The ultimate edge node is your brain talking directly to the AI, no screen needed. The smartphone dies in the next 5 years. The AI companion era begins. And xAI is building the brain of the entire thing. The future isn't a better iPhone. It's the death of the phone as we know it. Elon showed the blueprint. Source: Elon Musk, Joe Rogan, Sawyer Merritt

🇺🇸 ELON KILLED THE SMARTPHONE: "WE'RE NOT GOING TO HAVE A PHONE IN THE TRADITIONAL SENSE" No, Tesla isn't building a phone. They're leapfrogging the entire concept. The device in your pocket won't be a phone anymore. It'll be a super-thin AI edge node, basically a screen with radios that pings Grok/xAI servers for EVERYTHING in real time. No apps. No operating system. No local processing. Just instant photorealistic video generation, simulations, predictive actions (it orders your coffee before you know you want it), all streamed from the cloud like your brain is directly plugged into the AI. Exact quote that broke the internet: “We’re not going to have a phone in the traditional sense. What we’ll call a phone will really be an edge node for AI inference with some radios to connect.” This is already how Grok works in Tesla cars: voice in, server crunches, perfect response out. Now imagine that in your pocket, on steroids. The ultimate edge node is your brain talking directly to the AI, no screen needed. The smartphone dies in the next 5 years. The AI companion era begins. And xAI is building the brain of the entire thing. The future isn't a better iPhone. It's the death of the phone as we know it. Elon showed the blueprint. Source: Elon Musk, Joe Rogan, Sawyer Merritt

Mario Nawfal

1,861,339 Aufrufe • vor 10 Monaten

We built the fastest PDF -> markdown parser in the world 🚀⚡️ AND it’s more accurate than any other open-source, model-free parser (pymupdf4llm, opendataloader, pdf-inspector, markitdown) on 3 standardized benchmarks: olmOCR0-bench, opendataloader-bench, ParseBench Introducing LiteParse v2.1. The v2 base version was already the fastest document->text parser on the planet, and with this new release we’ve introduced markdown. It is fully open-source (Apache 2.0) and free, is usable from CLI/Rust/Node/Python/WASM, and is also installable as a one-click agent skill. Check it out: Come check out LiteParse:

We built the fastest PDF -> markdown parser in the world 🚀⚡️ AND it’s more accurate than any other open-source, model-free parser (pymupdf4llm, opendataloader, pdf-inspector, markitdown) on 3 standardized benchmarks: olmOCR0-bench, opendataloader-bench, ParseBench Introducing LiteParse v2.1. The v2 base version was already the fastest document->text parser on the planet, and with this new release we’ve introduced markdown. It is fully open-source (Apache 2.0) and free, is usable from CLI/Rust/Node/Python/WASM, and is also installable as a one-click agent skill. Check it out: Come check out LiteParse:

Jerry Liu

364,182 Aufrufe • vor 3 Monaten

The most basic test SemiAnalysis applies to a GPU cluster health check is whether it could possibly work on paper. Several could not. "We've been through a few of these clusters where there's no way it could possibly work. These are health checks so bad they're worse than no health checks, because they're actively interfering with jobs." "On Amazon HyperPod Slurm, the first time we tested, there was a health check that needed the node to be healthy before it could run. It could only run once the node was back in the fleet, and it was needed to bring the node back into the fleet." "There are probably five or ten examples of this throughout testing where the health check never had a chance. It makes you wonder whether these people are really putting it through the paces before they hand it off to customers."

The most basic test SemiAnalysis applies to a GPU cluster health check is whether it could possibly work on paper. Several could not. "We've been through a few of these clusters where there's no way it could possibly work. These are health checks so bad they're worse than no health checks, because they're actively interfering with jobs." "On Amazon HyperPod Slurm, the first time we tested, there was a health check that needed the node to be healthy before it could run. It could only run once the node was back in the fleet, and it was needed to bring the node back into the fleet." "There are probably five or ten examples of this throughout testing where the health check never had a chance. It makes you wonder whether these people are really putting it through the paces before they hand it off to customers."

SemiAnalysis

27,640 Aufrufe • vor 14 Tagen

Alan Oppenheim, MIT professor: "Jump Trading pays quants $650K to pull one real signal out of market noise. I wrote the textbook on that exact math and taught it for free." this free MIT lecture is the math the whole "quant edge" is built on, and almost nobody selling you a course will name it. the man teaching it wrote the signal-processing textbook every engineer on earth learned from, and he put the lectures online for nothing. here is the idea. a price chart is a signal buried in noise. most of what you see is random static, and hidden inside is a faint, repeating pattern with real mathematics for pulling it out. that is not a metaphor. jim simons was a codebreaker who did exactly this to enemy transmissions, then pointed the same math at the market and made a hundred billion dollars. none of it is new. oppenheim taught this in 1987, the math is older, and every lecture is free. same story i told in the article above. the edge was never a secret formula. it was knowing a market is a noisy signal, and having the tools to read it. here is what the gurus skip. the math to separate signal from noise is free and everyone has it. the edge is cleaner data than the next desk, and the discipline to trust a faint signal across thousands of trades instead of chasing the static. the filter is free. the patience to use it is not.

Alan Oppenheim, MIT professor: "Jump Trading pays quants $650K to pull one real signal out of market noise. I wrote the textbook on that exact math and taught it for free." this free MIT lecture is the math the whole "quant edge" is built on, and almost nobody selling you a course will name it. the man teaching it wrote the signal-processing textbook every engineer on earth learned from, and he put the lectures online for nothing. here is the idea. a price chart is a signal buried in noise. most of what you see is random static, and hidden inside is a faint, repeating pattern with real mathematics for pulling it out. that is not a metaphor. jim simons was a codebreaker who did exactly this to enemy transmissions, then pointed the same math at the market and made a hundred billion dollars. none of it is new. oppenheim taught this in 1987, the math is older, and every lecture is free. same story i told in the article above. the edge was never a secret formula. it was knowing a market is a noisy signal, and having the tools to read it. here is what the gurus skip. the math to separate signal from noise is free and everyone has it. the edge is cleaner data than the next desk, and the discipline to trust a faint signal across thousands of trades instead of chasing the static. the filter is free. the patience to use it is not.

Rossst.03

281,905 Aufrufe • vor 3 Monaten

This 117-year-old Japanese food company controls whether Nvidia can build another chip. It has never made a wafer, a transistor, or a single piece of equipment. Here's what they do and what just happened: Ajinomoto commercialized monosodium glutamate in 1909. It's the seasoning in your instant ramen. And somewhere in the 1990s its chemists worked out that a byproduct of that process made an almost perfect insulating film. They called it Ajinomoto Build-up Film. The industry calls it ABF. It's the layer inside the substrate that sits directly under a high-end chip. Nvidia's accelerators need it. Intel's processors need it. The advanced packaging TSMC does needs it. And Ajinomoto holds roughly 95% of the world market for it. They've held that share for about 30 years. Now here's what makes this absurd: ABF is around 0.5% of what a high-end chip costs to make. Half of one percent. But NOTHING ships without it, because the tolerance for failure is zero and qualifying a different supplier takes 2 to 3 years of testing. So a food company ended up holding the smallest and cheapest part of the most expensive supply chain on Earth. For decades they barely priced it that way but then somebody noticed... In March a British activist fund called Palliser Capital disclosed a top 25 stake in Ajinomoto and said publicly that the company had underpriced ABF for years and should raise it by more than 30%. Two months later Ajinomoto told substrate makers the price was going up about 30%. That increase took effect this quarter. And Palliser wants more than the price. It wants the electronics business carved out of the food conglomerate entirely, because the market has been valuing a monopoly on AI infrastructure as though it were a seasoning line. Then in early August it got more interesting: Ajinomoto reportedly told its customers in mainland China it was cutting their ABF supply by 30%. China's domestic substitute covers under 5% of what the country needs. Ajinomoto hasn't confirmed that cut and the reporting comes from Chinese trade press. But nobody has denied it either. And the squeeze is already showing up in actual hardware. ABF lines have run above 95% utilization since the start of the year. Lead times went from 12 weeks in mid-2025 to 26 weeks now. One contract manufacturer in Singapore cut its AI accelerator builds by 37% in a single quarter because it couldn't get substrate. Goldman Sachs has the supply gap widening from roughly 10% today to 21% next year and 42% by 2028. But there are escape routes too: Samsung Electro-Mechanics is spending $1.2 billion building its own lines and won't have volume until late 2027. Ajinomoto's third plant doesn't produce until 2032. And in April a Chinese MSG producer called Lotus Holdings paid roughly $15 million for control of a Shenzhen ABF startup with under $1 million in annual revenue. So one seasoning company is now trying to break another seasoning company's grip on AI. The counter-case is fair though: It's half a percent of a chip, everyone is expanding, and Samsung or Sekisui probably cracks it before the decade is out. Ajinomoto going underpaid for 30 years is an argument FOR the hike. But that's a 2029 answer to a 2026 problem. Trillions of dollars of AI spending now runs through one Japanese factory that grew out of soup stock, and that factory has just learned what it's actually worth. Ajinomoto spent 30 years being too polite to charge for it. A hedge fund talked them out of that in a few weeks.

This 117-year-old Japanese food company controls whether Nvidia can build another chip. It has never made a wafer, a transistor, or a single piece of equipment. Here's what they do and what just happened: Ajinomoto commercialized monosodium glutamate in 1909. It's the seasoning in your instant ramen. And somewhere in the 1990s its chemists worked out that a byproduct of that process made an almost perfect insulating film. They called it Ajinomoto Build-up Film. The industry calls it ABF. It's the layer inside the substrate that sits directly under a high-end chip. Nvidia's accelerators need it. Intel's processors need it. The advanced packaging TSMC does needs it. And Ajinomoto holds roughly 95% of the world market for it. They've held that share for about 30 years. Now here's what makes this absurd: ABF is around 0.5% of what a high-end chip costs to make. Half of one percent. But NOTHING ships without it, because the tolerance for failure is zero and qualifying a different supplier takes 2 to 3 years of testing. So a food company ended up holding the smallest and cheapest part of the most expensive supply chain on Earth. For decades they barely priced it that way but then somebody noticed... In March a British activist fund called Palliser Capital disclosed a top 25 stake in Ajinomoto and said publicly that the company had underpriced ABF for years and should raise it by more than 30%. Two months later Ajinomoto told substrate makers the price was going up about 30%. That increase took effect this quarter. And Palliser wants more than the price. It wants the electronics business carved out of the food conglomerate entirely, because the market has been valuing a monopoly on AI infrastructure as though it were a seasoning line. Then in early August it got more interesting: Ajinomoto reportedly told its customers in mainland China it was cutting their ABF supply by 30%. China's domestic substitute covers under 5% of what the country needs. Ajinomoto hasn't confirmed that cut and the reporting comes from Chinese trade press. But nobody has denied it either. And the squeeze is already showing up in actual hardware. ABF lines have run above 95% utilization since the start of the year. Lead times went from 12 weeks in mid-2025 to 26 weeks now. One contract manufacturer in Singapore cut its AI accelerator builds by 37% in a single quarter because it couldn't get substrate. Goldman Sachs has the supply gap widening from roughly 10% today to 21% next year and 42% by 2028. But there are escape routes too: Samsung Electro-Mechanics is spending $1.2 billion building its own lines and won't have volume until late 2027. Ajinomoto's third plant doesn't produce until 2032. And in April a Chinese MSG producer called Lotus Holdings paid roughly $15 million for control of a Shenzhen ABF startup with under $1 million in annual revenue. So one seasoning company is now trying to break another seasoning company's grip on AI. The counter-case is fair though: It's half a percent of a chip, everyone is expanding, and Samsung or Sekisui probably cracks it before the decade is out. Ajinomoto going underpaid for 30 years is an argument FOR the hike. But that's a 2029 answer to a 2026 problem. Trillions of dollars of AI spending now runs through one Japanese factory that grew out of soup stock, and that factory has just learned what it's actually worth. Ajinomoto spent 30 years being too polite to charge for it. A hedge fund talked them out of that in a few weeks.

Ricardo

44,025 Aufrufe • vor 1 Monat

More GOP questions about Iran end goals. Here’s GOP Rep Pete Sessions to me tonight: “This administration is going to have to figure out: what is the goal here? Is the goal here to open back up the strait of hormuz, or is it where Israel's position would be? Or is it to do away with their desire to bomb people?…We should have a better understanding about what is left in an arsenal against not only the United States and our assets there, but also Israel and other friends that we have in the gulf. So this is an equation that they are going to have to put their heads together and make that determination.”

More GOP questions about Iran end goals. Here’s GOP Rep Pete Sessions to me tonight: “This administration is going to have to figure out: what is the goal here? Is the goal here to open back up the strait of hormuz, or is it where Israel's position would be? Or is it to do away with their desire to bomb people?…We should have a better understanding about what is left in an arsenal against not only the United States and our assets there, but also Israel and other friends that we have in the gulf. So this is an equation that they are going to have to put their heads together and make that determination.”

Jim Sciutto

17,230 Aufrufe • vor 6 Monaten

California Assemblymen David Tangipa EXPOSES Gavin Newsom and California Democrats are making them vote on redistricting WITHOUT SEEING IT “This is the most undemocratic process you could possibly imagine. We are voting on a law where none of us has seen the new amendment — This is wrong. This is wrong. This is wrong” “They just changed the law that we are supposed to vote on for the election 10 minutes ago. They are asking all of us here in the legislative body to vote on that and we haven't even seen the language that they changed. “This is the most undemocratic process you could possibly imagine. We are voting on a law where none of us has seen the new amendment. It is a new law now. They have changed it entirely, and now they are forcing it on us and on the floor. People need to see this. This is wrong. This is wrong. This is wrong. 40 million Californians deserve more than 10 minutes of deliberation, especially when your elected leaders haven't even seen it. Please share. Tell everybody. What is going on in Sacramento is the most undemocratic process you could possibly imagine.”

California Assemblymen David Tangipa EXPOSES Gavin Newsom and California Democrats are making them vote on redistricting WITHOUT SEEING IT “This is the most undemocratic process you could possibly imagine. We are voting on a law where none of us has seen the new amendment — This is wrong. This is wrong. This is wrong” “They just changed the law that we are supposed to vote on for the election 10 minutes ago. They are asking all of us here in the legislative body to vote on that and we haven't even seen the language that they changed. “This is the most undemocratic process you could possibly imagine. We are voting on a law where none of us has seen the new amendment. It is a new law now. They have changed it entirely, and now they are forcing it on us and on the floor. People need to see this. This is wrong. This is wrong. This is wrong. 40 million Californians deserve more than 10 minutes of deliberation, especially when your elected leaders haven't even seen it. Please share. Tell everybody. What is going on in Sacramento is the most undemocratic process you could possibly imagine.”

Wall Street Apes

137,249 Aufrufe • vor 1 Jahr