Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

$ASML EUV tools print features as small as ~13nm making them essential for manufacturing AI chips. Every chip like $NVDA Blackwell, $AMD MI300, $AAPL M4, $GOOGL TPU & $AMZN Trainium requires EUV lithography. You either use ASML’s machines or you don’t make advanced chips.

Shay Boloor

446,450 subscribers

57,766 views • 6 months ago •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

$ASML has become one of the clearest ways to invest in the global AI chip buildout because it controls one of the most important bottlenecks in semiconductor manufacturing. Its lithography systems act like ultra-precise printers that create the tiny circuit patterns used in advanced chips and ASML remains the only company producing EUV machines at scale. $TSM cannot manufacture $NVDA Blackwell or Rubin GPUs without those systems which means nearly every dollar flowing into advanced AI compute ultimately passes through ASML’s equipment.

$ASML has become one of the clearest ways to invest in the global AI chip buildout because it controls one of the most important bottlenecks in semiconductor manufacturing. Its lithography systems act like ultra-precise printers that create the tiny circuit patterns used in advanced chips and ASML remains the only company producing EUV machines at scale. $TSM cannot manufacture $NVDA Blackwell or Rubin GPUs without those systems which means nearly every dollar flowing into advanced AI compute ultimately passes through ASML’s equipment.

Shay Boloor

62,457 views • 1 month ago

These machines don’t look like they build the future of AI. But they do. What you are seeing is not an industrial robot just move. It is part of the process that makes the world’s most advanced chips possible. TRUMPF laser systems are used to manufacture key components for ASML’s EUV lithography machines. ❗️Without this step, modern AI hardware would simply not exist. • They machine parts with nanometer-level precision • They enable the optics and components needed for EUV lithography • Their stability directly affects chip yield and quality • No TRUMPF lasers, no ASML EUV. No EUV, no advanced AI chips. It is easy to focus on GPUs and models. But the real bottleneck sits much deeper in the industrial stack. ——- Weekly robotics and AI insights. Subscribe free:

These machines don’t look like they build the future of AI. But they do. What you are seeing is not an industrial robot just move. It is part of the process that makes the world’s most advanced chips possible. TRUMPF laser systems are used to manufacture key components for ASML’s EUV lithography machines. ❗️Without this step, modern AI hardware would simply not exist. • They machine parts with nanometer-level precision • They enable the optics and components needed for EUV lithography • Their stability directly affects chip yield and quality • No TRUMPF lasers, no ASML EUV. No EUV, no advanced AI chips. It is easy to focus on GPUs and models. But the real bottleneck sits much deeper in the industrial stack. ——- Weekly robotics and AI insights. Subscribe free:

Ilir Aliu

43,443 views • 23 days ago

These machines don’t look like they build the future of AI. But they do. What you are seeing is not an industrial robot just move. It is part of the process that makes the world’s most advanced chips possible. TRUMPF laser systems are used to manufacture key components for ASML’s EUV lithography machines. ❗️Without this step, modern AI hardware would simply not exist. • They machine parts with nanometer-level precision • They enable the optics and components needed for EUV lithography • Their stability directly affects chip yield and quality • No TRUMPF lasers, no ASML EUV. No EUV, no advanced AI chips. It is easy to focus on GPUs and models. But the real bottleneck sits much deeper in the industrial stack. —- Weekly robotics and AI insights. Subscribe free:

These machines don’t look like they build the future of AI. But they do. What you are seeing is not an industrial robot just move. It is part of the process that makes the world’s most advanced chips possible. TRUMPF laser systems are used to manufacture key components for ASML’s EUV lithography machines. ❗️Without this step, modern AI hardware would simply not exist. • They machine parts with nanometer-level precision • They enable the optics and components needed for EUV lithography • Their stability directly affects chip yield and quality • No TRUMPF lasers, no ASML EUV. No EUV, no advanced AI chips. It is easy to focus on GPUs and models. But the real bottleneck sits much deeper in the industrial stack. —- Weekly robotics and AI insights. Subscribe free:

Ilir Aliu

99,899 views • 8 months ago

Here are 20 companies that will dominate the AI age: 1/ ASML $ASML is revolutionizing the semiconductor industry by enabling chipmakers to produce the world’s most advanced processors with atomic-scale precision. Shrinking transistors to fit more power into smaller chips faces fundamental physical limits. The industry struggled with escalating costs, slower innovation cycles, and diminishing returns from older lithography methods. With $ASML, the process is radically different, its extreme ultraviolet (EUV) lithography machines use cutting-edge optics and physics to etch circuitry at the nanometer scale, enabling faster, more energy-efficient, and more powerful chips. This breakthrough is the backbone of progress in AI, cloud computing, smartphones, and virtually every advanced technology. As AI spreads to all aspects of our economies, an ever-growing number of chips is needed, driving demand for ASML machines and maintenance services.

Here are 20 companies that will dominate the AI age: 1/ ASML $ASML is revolutionizing the semiconductor industry by enabling chipmakers to produce the world’s most advanced processors with atomic-scale precision. Shrinking transistors to fit more power into smaller chips faces fundamental physical limits. The industry struggled with escalating costs, slower innovation cycles, and diminishing returns from older lithography methods. With $ASML, the process is radically different, its extreme ultraviolet (EUV) lithography machines use cutting-edge optics and physics to etch circuitry at the nanometer scale, enabling faster, more energy-efficient, and more powerful chips. This breakthrough is the backbone of progress in AI, cloud computing, smartphones, and virtually every advanced technology. As AI spreads to all aspects of our economies, an ever-growing number of chips is needed, driving demand for ASML machines and maintenance services.

Ray Myers

307,231 views • 1 year ago

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 views • 1 month ago

🚨 JUST IN: President Trump recommends chip manufacturers IMMEDIATELY move their manufacturing facilities to AMERICA, as conversations with Xi confirmed how eager China is to take over Taiwan Chips are ESSENTIAL for AI and technological advances, and Taiwan is the world's chip manufacturing hub at the moment. That's one of the main reasons China wants to take Taiwan. TRUMP: "I would like to see everybody making chips in Taiwan come into America. To be honest with you, I think it's the greatest thing can you do. It's a heated situation. There is no question about it. As you know, we have massive amounts of chip companies now from Taiwan already coming in. "We expect to have 40% to 50% of the world chip business by the end of my term."

🚨 JUST IN: President Trump recommends chip manufacturers IMMEDIATELY move their manufacturing facilities to AMERICA, as conversations with Xi confirmed how eager China is to take over Taiwan Chips are ESSENTIAL for AI and technological advances, and Taiwan is the world's chip manufacturing hub at the moment. That's one of the main reasons China wants to take Taiwan. TRUMP: "I would like to see everybody making chips in Taiwan come into America. To be honest with you, I think it's the greatest thing can you do. It's a heated situation. There is no question about it. As you know, we have massive amounts of chip companies now from Taiwan already coming in. "We expect to have 40% to 50% of the world chip business by the end of my term."

Nick Sortor

1,983,771 views • 3 months ago

WHAT IF CHINA INVADES TAIWAN + Strait of Hormuz STAYS CLOSED🛑 Nobody is connecting these dots yet Taiwan makes 90% of the world's most advanced chips. TSMC manufactures for EVERY major tech company on the planet. If those fabs go dark, this isn't a dip. This is a STRUCTURAL COLLAPSE. Stocks that get DESTROYED 👇 Semiconductors. $TSM, $NVDA, $AMD, $AAPL, $QCOM, $AVGO, $ASML. Every single one depends on TSMC to make their chips. There is NO backup. Nvidia can't make GPUs. Apple can't make iPhones. AMD can't make processors. Full stop. The AI boom DIES overnight. $MSFT, $AMZN, $GOOG, $META. Every data center buildout, every model training run, every hyperscaler capex plan depends on chips that ONLY TSMC can produce at scale. Their entire growth thesis is GONE. Consumer and shipping. $DELL, $HPQ, Sony all lose their chip supply. The Taiwan Strait carries 50% of global container shipping. $ZIM, $FDX, $UPS all get crushed by the blockade. NOW here's where it gets interesting Stocks that BENEFIT 👇 Defense. $LMT, $RTX, $NOC, $GD, $LHX, $PLTR. Spending goes VERTICAL overnight. Domestic chips. $INTC becomes the most important company in America as the ONLY Western advanced fab. $GFS, $AMAT, $LRCX, $KLAC all surge as we scramble to build domestic capacity. Safe havens. $XOM, $CVX on the oil spike. $GLD, $SLV on the flight to safety. $AA on the aluminum supply chain chaos. THE PART NOBODY IS TALKING ABOUT Taiwan doesn't just make chips for tech. They produce 35% of ALL chips globally. Cars. Medical devices. Military equipment. Appliances. A Taiwan invasion doesn't crash the stock market. It crashes the GLOBAL ECONOMY. This is not a prediction. But if you're not thinking about this risk while the Strait of Hormuz is ALREADY shut down and semiconductors are ALREADY under pressure from helium shortages, you're not paying attention.

WHAT IF CHINA INVADES TAIWAN + Strait of Hormuz STAYS CLOSED🛑 Nobody is connecting these dots yet Taiwan makes 90% of the world's most advanced chips. TSMC manufactures for EVERY major tech company on the planet. If those fabs go dark, this isn't a dip. This is a STRUCTURAL COLLAPSE. Stocks that get DESTROYED 👇 Semiconductors. $TSM, $NVDA, $AMD, $AAPL, $QCOM, $AVGO, $ASML. Every single one depends on TSMC to make their chips. There is NO backup. Nvidia can't make GPUs. Apple can't make iPhones. AMD can't make processors. Full stop. The AI boom DIES overnight. $MSFT, $AMZN, $GOOG, $META. Every data center buildout, every model training run, every hyperscaler capex plan depends on chips that ONLY TSMC can produce at scale. Their entire growth thesis is GONE. Consumer and shipping. $DELL, $HPQ, Sony all lose their chip supply. The Taiwan Strait carries 50% of global container shipping. $ZIM, $FDX, $UPS all get crushed by the blockade. NOW here's where it gets interesting Stocks that BENEFIT 👇 Defense. $LMT, $RTX, $NOC, $GD, $LHX, $PLTR. Spending goes VERTICAL overnight. Domestic chips. $INTC becomes the most important company in America as the ONLY Western advanced fab. $GFS, $AMAT, $LRCX, $KLAC all surge as we scramble to build domestic capacity. Safe havens. $XOM, $CVX on the oil spike. $GLD, $SLV on the flight to safety. $AA on the aluminum supply chain chaos. THE PART NOBODY IS TALKING ABOUT Taiwan doesn't just make chips for tech. They produce 35% of ALL chips globally. Cars. Medical devices. Military equipment. Appliances. A Taiwan invasion doesn't crash the stock market. It crashes the GLOBAL ECONOMY. This is not a prediction. But if you're not thinking about this risk while the Strait of Hormuz is ALREADY shut down and semiconductors are ALREADY under pressure from helium shortages, you're not paying attention.

JEFE TRADES 🔪

222,056 views • 5 months ago

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

12,579 views • 24 days ago

Many still don’t understand why Elon is building Terafab Terafab is an extension to all the chip makers in the world It’s not about replacement, not a rivalry and absolutely not competing It’s being built to fulfill the massive chip orders that Tesla, SpaceX and xAI actually need TSMC’s most advanced 2nm capacity is totally booked through 2028 Tesla signed a massive $16.5 billion deal with Samsung back in July 2025 to produce AI6 chips at their Taylor, Texas factory and Samsung is building a Tesla Exclusive chip manufacturing plant to full fill this orders When Elon announced Terafab on March 21, 2026...he made it clear: “That rate is much less than we’d like. We either build the Terafab or we don’t have the chips, and we need the chips, so we build the Terafab” He basically told the chip makers: “Produce as much as you comfortably can. We will take them all. Actually we want even more” Even today Elon said: "SpaceX/Tesla will be always be major customers of TSMC and not competitors in the normal sense of the word" Current production rates are much less than they need....That’s why Terafab exists Terafab is an extension to every chip maker… not competition, not rivalry, absolutely not Even Intel has joined as a partner Even if chip supply improves, massive bottlenecks still exist with memory and advanced packaging You simply can't risk those supply chain breaks at this scale That’s why Terafab is being built to vertically integrate everything - chips, memory, advanced packaging all under one roof, targeting 1 terawatt of AI compute capacity per year This is a ludicrous amount of chips that no chipmaker currently produces at this scale. I don't think even TSMC and Samsung truly understand these numbers yet It’s about building the capacity the future actually demands

Many still don’t understand why Elon is building Terafab Terafab is an extension to all the chip makers in the world It’s not about replacement, not a rivalry and absolutely not competing It’s being built to fulfill the massive chip orders that Tesla, SpaceX and xAI actually need TSMC’s most advanced 2nm capacity is totally booked through 2028 Tesla signed a massive $16.5 billion deal with Samsung back in July 2025 to produce AI6 chips at their Taylor, Texas factory and Samsung is building a Tesla Exclusive chip manufacturing plant to full fill this orders When Elon announced Terafab on March 21, 2026...he made it clear: “That rate is much less than we’d like. We either build the Terafab or we don’t have the chips, and we need the chips, so we build the Terafab” He basically told the chip makers: “Produce as much as you comfortably can. We will take them all. Actually we want even more” Even today Elon said: "SpaceX/Tesla will be always be major customers of TSMC and not competitors in the normal sense of the word" Current production rates are much less than they need....That’s why Terafab exists Terafab is an extension to every chip maker… not competition, not rivalry, absolutely not Even Intel has joined as a partner Even if chip supply improves, massive bottlenecks still exist with memory and advanced packaging You simply can't risk those supply chain breaks at this scale That’s why Terafab is being built to vertically integrate everything - chips, memory, advanced packaging all under one roof, targeting 1 terawatt of AI compute capacity per year This is a ludicrous amount of chips that no chipmaker currently produces at this scale. I don't think even TSMC and Samsung truly understand these numbers yet It’s about building the capacity the future actually demands

X Freeze

53,291 views • 4 months ago

The CEO of ASML said Europe is "quite behind" in the AI race. This is not a politician. This is the man who controls the only machines on earth that can print advanced chips. Every semiconductor fab from TSMC to Samsung needs his approval to exist. And he did not stop there. Europe has been having the sovereignty conversation for three years. Build fabs. Reduce dependence. Catch up. It sounds right. He said it is the wrong order entirely. His reasoning is blunt. Semiconductor manufacturing is only needed if you have people buying the chips. If Europe built a two nanometer fab today, most of those wafers would ship straight to the United States. Because the US buys 80% of the world's advanced chips. Europe buys almost none. You cannot manufacture your way to relevance. You have to create demand first. His actual sequence: start with the market. Build AI applications. Then AI products. Then chip design. Then and only then does chip manufacturing make sense. Europe skipped four steps and went straight to the last one. The sovereignty debate is loud in Brussels right now because of US export controls on AI models. Politicians are scared. The instinct is to build. He is saying the instinct is wrong. You do not get sovereignty by building factories. You get it by having something the world wants to buy. (Watch the full 11 minutes Interview at Bloomberg TV on YouTube)

The CEO of ASML said Europe is "quite behind" in the AI race. This is not a politician. This is the man who controls the only machines on earth that can print advanced chips. Every semiconductor fab from TSMC to Samsung needs his approval to exist. And he did not stop there. Europe has been having the sovereignty conversation for three years. Build fabs. Reduce dependence. Catch up. It sounds right. He said it is the wrong order entirely. His reasoning is blunt. Semiconductor manufacturing is only needed if you have people buying the chips. If Europe built a two nanometer fab today, most of those wafers would ship straight to the United States. Because the US buys 80% of the world's advanced chips. Europe buys almost none. You cannot manufacture your way to relevance. You have to create demand first. His actual sequence: start with the market. Build AI applications. Then AI products. Then chip design. Then and only then does chip manufacturing make sense. Europe skipped four steps and went straight to the last one. The sovereignty debate is loud in Brussels right now because of US export controls on AI models. Politicians are scared. The instinct is to build. He is saying the instinct is wrong. You do not get sovereignty by building factories. You get it by having something the world wants to buy. (Watch the full 11 minutes Interview at Bloomberg TV on YouTube)

Ihtesham Ali

131,550 views • 2 months ago

.Dylan Patel lays out how we know the hard upper bound on how much compute can be produced annually by 2030: around 200 GW/year. That’s a crazy number (there’s about 20 GW of AI deployed in the world right now), but it’s nowhere near enough to satisfy Sam/Elon/Dario/Demis’s ambitions. Lots of things in the supply chain can be scaled up over 4 years, including things that other people think are bottlenecks, like datacenter power or fab clean room space. But the thing that’s inflexible over that timeline is the number of EUV tools. Dylan forecasts that production of ASML’s EUV tools will scale from 60 per year now to about 100 per year by the end of the decade - which means something like 700 total machines running in 2030. For a fab to make a GW worth of the Rubin chips that NVIDIA is deploying later this year, it needs to make 55,000 3nm wafers, 6,000 5nm wafers, and 170,000 memory wafers. Each 3nm wafers needs about 20 EUV passes, so about 1.1 million passes per GW. Adding on 5nm and memory, you need two million passes. Each tool can do 75 passes per hour, so with 90% uptime that’s around 600k passes per year - so a single machine can make less than a third of a GW in a year. So in 2030, we have 700 total machines, each making 0.3ish GW a year, which means we can produce 200 GW of compute a year. That’s a lot. But Sam Altman wants a gigawatt a week by the end of the decade. Anthropic and Google will be wanting about the same. And Elon wants to be putting 100 GW in space every year. Any one of these players could maybe get what they need, but not all of them.

.Dylan Patel lays out how we know the hard upper bound on how much compute can be produced annually by 2030: around 200 GW/year. That’s a crazy number (there’s about 20 GW of AI deployed in the world right now), but it’s nowhere near enough to satisfy Sam/Elon/Dario/Demis’s ambitions. Lots of things in the supply chain can be scaled up over 4 years, including things that other people think are bottlenecks, like datacenter power or fab clean room space. But the thing that’s inflexible over that timeline is the number of EUV tools. Dylan forecasts that production of ASML’s EUV tools will scale from 60 per year now to about 100 per year by the end of the decade - which means something like 700 total machines running in 2030. For a fab to make a GW worth of the Rubin chips that NVIDIA is deploying later this year, it needs to make 55,000 3nm wafers, 6,000 5nm wafers, and 170,000 memory wafers. Each 3nm wafers needs about 20 EUV passes, so about 1.1 million passes per GW. Adding on 5nm and memory, you need two million passes. Each tool can do 75 passes per hour, so with 90% uptime that’s around 600k passes per year - so a single machine can make less than a third of a GW in a year. So in 2030, we have 700 total machines, each making 0.3ish GW a year, which means we can produce 200 GW of compute a year. That’s a lot. But Sam Altman wants a gigawatt a week by the end of the decade. Anthropic and Google will be wanting about the same. And Elon wants to be putting 100 GW in space every year. Any one of these players could maybe get what they need, but not all of them.

Dwarkesh Patel

114,538 views • 5 months ago

$IREN is by far my favorite buy among the neo cloud companies right now. While $NBIS is up more than 180% YTD, $IREN is sitting at just +55%. The market is still massively underestimating one thing: Power. $IREN 5GW of contracted power is a real moat, and in the AI era, that matters more than almost anything else. As Jensen $NVDA said: “None of the bottlenecks like memory or chip production last longer than a couple of years. What’s different is energy. You can’t create an industry without energy. You can’t build anything without energy, and those things take a long time.” This is the real AI bottleneck. Not chips. Not memory. Energy. $IREN already has it. Those 5GW will be fully utilized. Just like $AMD eventually massively outperformed $NVDA after lagging for a long time, I believe the same setup is happening here. Right now, $IREN is lagging $NBIS. That won’t last forever. Buy the laggard.

$IREN is by far my favorite buy among the neo cloud companies right now. While $NBIS is up more than 180% YTD, $IREN is sitting at just +55%. The market is still massively underestimating one thing: Power. $IREN 5GW of contracted power is a real moat, and in the AI era, that matters more than almost anything else. As Jensen $NVDA said: “None of the bottlenecks like memory or chip production last longer than a couple of years. What’s different is energy. You can’t create an industry without energy. You can’t build anything without energy, and those things take a long time.” This is the real AI bottleneck. Not chips. Not memory. Energy. $IREN already has it. Those 5GW will be fully utilized. Just like $AMD eventually massively outperformed $NVDA after lagging for a long time, I believe the same setup is happening here. Right now, $IREN is lagging $NBIS. That won’t last forever. Buy the laggard.

Daniel

123,728 views • 3 months ago

Neil thinks 90% of AI workloads will run in the background versus real time. In that future, latency matters less and cost matters more, and he's configured his company around a unique way of serving tokens at the cheapest possible price. "There's no bad chips. There's only bad pricing, and I will make any chip work at the right price. Let's talk about AMD, great chips overall. People don't understand how to program them very well. That's music to my ears. I'm happy for them to sleep on this chip and for me to buy as much as I can." "One of the ways I describe what we do is, we will buy any chip anywhere in the world for any duration of time. That is a level of flexibility and liquidity that no one else has right now. You'd have basically zero buyers for a data center that is ninety-five percent uptime. I'm that first buyer. First we scavenge chips, and then we scavenge power for those chips. The idea is, in both cases, I do not want to be bidding against Anthropic or OpenAI for compute capacity. I'm not gonna win against them, and I don't want to. I want to be more creative and use the supply that they don't find legible today. Over time I amass enough aggregate supply...and build my aggregate factory that is unbeatable in economics." "My whole goal is to so dramatically expand the supply of power across the United States that I have a home for a lot of chips that otherwise would not have earned their place in a data center." "We've really pushed AI to be an interactive chatbot tool, and everyone has chosen latency optimization because the shape of usage was chatbot oriented. That's the most profound change we're going to see in the next year, we're gonna move away from chatbots to more proactive or background agents, and in that world, it makes a lot more sense to build a stack around throughput." "I love this market because it's unbounded. There's no human in the loop, so you can consume as many tokens as you like in the background versus human attention span. So long term, we're going to end this year at maybe fifty-fifty background and real time workloads, but I see this going to ninety ten in favor of background." "The best latency is no latency at all. When you wake up in the morning, the work's already been done overnight, you didn't even have to ask for it. What we'd like is the agent to operate on more human time scales. You don't manage your colleagues every five minutes. You come back and check in maybe once a week." "My job is to make the tokens as cheap as humanly possible, and I will do it through every layer in the stack available to me."

Neil thinks 90% of AI workloads will run in the background versus real time. In that future, latency matters less and cost matters more, and he's configured his company around a unique way of serving tokens at the cheapest possible price. "There's no bad chips. There's only bad pricing, and I will make any chip work at the right price. Let's talk about AMD, great chips overall. People don't understand how to program them very well. That's music to my ears. I'm happy for them to sleep on this chip and for me to buy as much as I can." "One of the ways I describe what we do is, we will buy any chip anywhere in the world for any duration of time. That is a level of flexibility and liquidity that no one else has right now. You'd have basically zero buyers for a data center that is ninety-five percent uptime. I'm that first buyer. First we scavenge chips, and then we scavenge power for those chips. The idea is, in both cases, I do not want to be bidding against Anthropic or OpenAI for compute capacity. I'm not gonna win against them, and I don't want to. I want to be more creative and use the supply that they don't find legible today. Over time I amass enough aggregate supply...and build my aggregate factory that is unbeatable in economics." "My whole goal is to so dramatically expand the supply of power across the United States that I have a home for a lot of chips that otherwise would not have earned their place in a data center." "We've really pushed AI to be an interactive chatbot tool, and everyone has chosen latency optimization because the shape of usage was chatbot oriented. That's the most profound change we're going to see in the next year, we're gonna move away from chatbots to more proactive or background agents, and in that world, it makes a lot more sense to build a stack around throughput." "I love this market because it's unbounded. There's no human in the loop, so you can consume as many tokens as you like in the background versus human attention span. So long term, we're going to end this year at maybe fifty-fifty background and real time workloads, but I see this going to ninety ten in favor of background." "The best latency is no latency at all. When you wake up in the morning, the work's already been done overnight, you didn't even have to ask for it. What we'd like is the agent to operate on more human time scales. You don't manage your colleagues every five minutes. You come back and check in maybe once a week." "My job is to make the tokens as cheap as humanly possible, and I will do it through every layer in the stack available to me."

Patrick OShaughnessy

81,360 views • 7 days ago

The Three Tailwinds Nobody's Pricing Into Nvidia 🔊 On Monday's edition of The Macro Show, Director of Research Daryl Jones Jonesy and $HGRO Portfolio Manager Sam Rahman Sam Rahman explained how a PM actually positions into a print like Nvidia's, $NVDA. As Rahman noted, he owns both Nvidia and Marvell, $MRVL — the only two semis currently on the signal strength list — but keeps his active Nvidia position small, with far bigger active weights in software names like Amazon, $AMZN and Microsoft, $MSFT. His approach into big benchmark events is not to overthink the positioning, because get on the wrong side of the news flow and you can get hurt badly. The number that frames the quarter, Rahman explained, is that Nvidia's data-center business has the potential to nearly double year-over-year, an extraordinary feat for a company its size: "Think about the size of this company doubling its core sales line year over year." Beyond the print, he flagged three tailwinds: 1. The transition from Blackwell to the faster, more efficient Vera Rubin chip, with hyperscaler orders already landing. 2. SpaceX emerging as a large, genuinely incremental new buyer of data-center infrastructure post-IPO. 3. A roughly 15% price increase on chips and servers next year, driven by memory-chip inflation, that flows straight to revenue. Subscribe to The Macro Show for daily market analysis you won't find anywhere else:

The Three Tailwinds Nobody's Pricing Into Nvidia 🔊 On Monday's edition of The Macro Show, Director of Research Daryl Jones Jonesy and $HGRO Portfolio Manager Sam Rahman Sam Rahman explained how a PM actually positions into a print like Nvidia's, $NVDA. As Rahman noted, he owns both Nvidia and Marvell, $MRVL — the only two semis currently on the signal strength list — but keeps his active Nvidia position small, with far bigger active weights in software names like Amazon, $AMZN and Microsoft, $MSFT. His approach into big benchmark events is not to overthink the positioning, because get on the wrong side of the news flow and you can get hurt badly. The number that frames the quarter, Rahman explained, is that Nvidia's data-center business has the potential to nearly double year-over-year, an extraordinary feat for a company its size: "Think about the size of this company doubling its core sales line year over year." Beyond the print, he flagged three tailwinds: 1. The transition from Blackwell to the faster, more efficient Vera Rubin chip, with hyperscaler orders already landing. 2. SpaceX emerging as a large, genuinely incremental new buyer of data-center infrastructure post-IPO. 3. A roughly 15% price increase on chips and servers next year, driven by memory-chip inflation, that flows straight to revenue. Subscribe to The Macro Show for daily market analysis you won't find anywhere else:

Hedgeye

13,455 views • 8 days ago

Google just launched a direct attack on Nvidia's most valuable asset. Not their chips. Their SOFTWARE. And if this works, Nvidia's $4 trillion empire collapses. Here's what just leaked: Google is building "TorchTPU" - a secret project that makes PyTorch seamlessly run on Google's TPU chips instead of Nvidia GPUs. Why does this matter? PyTorch is the MOST USED AI framework on Earth. Every AI developer uses it. And PyTorch was built around Nvidia's CUDA software. Wall Street analysts call CUDA "Nvidia's strongest defensive wall." It's the reason companies can't easily switch away from Nvidia even when alternatives exist. You don't just buy Nvidia chips. You buy into their entire ecosystem. Switching costs MILLIONS in engineering work. Months of rewrites. Performance drops. So companies stay locked in. Even when Nvidia raises prices. Even when supply runs short. That's not a hardware moat. That's a SOFTWARE prison. And Google just found the escape route. Here's the problem Nvidia created for itself: Google's TPU chips are actually GOOD. Competitive performance. Better availability. Lower cost. But developers won't use them because Google's chips run JAX (Google's internal framework), not PyTorch. That means if you want to use Google TPUs, you have to rewrite your entire codebase. Nobody wants to do that. So Google TPUs sit unused while developers fight over Nvidia chips. Until now. TorchTPU makes PyTorch run natively on Google hardware. No rewrites. No performance loss. No months of engineering. You just... switch. And Google is partnering with META (who built PyTorch) to make it happen. They're even considering OPEN-SOURCING parts of it to speed adoption. Translation: Google is willing to give this away for free just to break Nvidia's lock. The implications are insane: Every company currently paying Nvidia's premium prices suddenly has a way out. Oracle, Microsoft, OpenAI - all locked into Nvidia's ecosystem - can switch to Google. Nvidia's pricing power evaporates overnight. And the timing is perfect: Nvidia is already facing heat. Semiconductor index dropped 3% today. Oracle just lost their biggest investor over AI spending concerns. Companies are realizing AI infrastructure costs are unsustainable. Now Google hands them an alternative. Same performance. Lower cost. Better availability. Jensen Huang knows exactly what this means. CUDA has been Nvidia's untouchable advantage for YEARS. It's why Nvidia trades at 50x earnings while AMD trades at 25x. The software moat justified the premium. But if Google removes that switching cost? Nvidia becomes just another chip company. And chip companies compete on price, not ecosystem lock-in. Here's what happens next: Google needs 12-18 months to make TorchTPU production-ready. If it works, cloud providers will adopt it instantly. They WANT an alternative to Nvidia's monopoly pricing. Amazon already building their own Trainium chips. Microsoft making Maia. They're all trying to escape Nvidia. Google just gave them the software bridge. Nvidia's response options are limited: They can't buy Google. Can't kill PyTorch (Meta owns it). Can't stop open source. Their only play is to keep improving CUDA faster than Google can catch up. But that's a race, not a moat. The market isn't pricing this in yet. Nvidia down 2% today. Google down 2%. Investors think this is just "another competitor." They don't understand this is an attack on the FOUNDATION of Nvidia's valuation. Hardware is replaceable. Software lock-in is what made Nvidia worth $4 trillion. Google is attacking the lock-in. Watch what happens in 2026 when TorchTPU goes live and companies realize they can actually leave Nvidia. The "Nvidia is unstoppable" narrative dies. And a $4 trillion valuation built on software moats gets repriced.

Google just launched a direct attack on Nvidia's most valuable asset. Not their chips. Their SOFTWARE. And if this works, Nvidia's $4 trillion empire collapses. Here's what just leaked: Google is building "TorchTPU" - a secret project that makes PyTorch seamlessly run on Google's TPU chips instead of Nvidia GPUs. Why does this matter? PyTorch is the MOST USED AI framework on Earth. Every AI developer uses it. And PyTorch was built around Nvidia's CUDA software. Wall Street analysts call CUDA "Nvidia's strongest defensive wall." It's the reason companies can't easily switch away from Nvidia even when alternatives exist. You don't just buy Nvidia chips. You buy into their entire ecosystem. Switching costs MILLIONS in engineering work. Months of rewrites. Performance drops. So companies stay locked in. Even when Nvidia raises prices. Even when supply runs short. That's not a hardware moat. That's a SOFTWARE prison. And Google just found the escape route. Here's the problem Nvidia created for itself: Google's TPU chips are actually GOOD. Competitive performance. Better availability. Lower cost. But developers won't use them because Google's chips run JAX (Google's internal framework), not PyTorch. That means if you want to use Google TPUs, you have to rewrite your entire codebase. Nobody wants to do that. So Google TPUs sit unused while developers fight over Nvidia chips. Until now. TorchTPU makes PyTorch run natively on Google hardware. No rewrites. No performance loss. No months of engineering. You just... switch. And Google is partnering with META (who built PyTorch) to make it happen. They're even considering OPEN-SOURCING parts of it to speed adoption. Translation: Google is willing to give this away for free just to break Nvidia's lock. The implications are insane: Every company currently paying Nvidia's premium prices suddenly has a way out. Oracle, Microsoft, OpenAI - all locked into Nvidia's ecosystem - can switch to Google. Nvidia's pricing power evaporates overnight. And the timing is perfect: Nvidia is already facing heat. Semiconductor index dropped 3% today. Oracle just lost their biggest investor over AI spending concerns. Companies are realizing AI infrastructure costs are unsustainable. Now Google hands them an alternative. Same performance. Lower cost. Better availability. Jensen Huang knows exactly what this means. CUDA has been Nvidia's untouchable advantage for YEARS. It's why Nvidia trades at 50x earnings while AMD trades at 25x. The software moat justified the premium. But if Google removes that switching cost? Nvidia becomes just another chip company. And chip companies compete on price, not ecosystem lock-in. Here's what happens next: Google needs 12-18 months to make TorchTPU production-ready. If it works, cloud providers will adopt it instantly. They WANT an alternative to Nvidia's monopoly pricing. Amazon already building their own Trainium chips. Microsoft making Maia. They're all trying to escape Nvidia. Google just gave them the software bridge. Nvidia's response options are limited: They can't buy Google. Can't kill PyTorch (Meta owns it). Can't stop open source. Their only play is to keep improving CUDA faster than Google can catch up. But that's a race, not a moat. The market isn't pricing this in yet. Nvidia down 2% today. Google down 2%. Investors think this is just "another competitor." They don't understand this is an attack on the FOUNDATION of Nvidia's valuation. Hardware is replaceable. Software lock-in is what made Nvidia worth $4 trillion. Google is attacking the lock-in. Watch what happens in 2026 when TorchTPU goes live and companies realize they can actually leave Nvidia. The "Nvidia is unstoppable" narrative dies. And a $4 trillion valuation built on software moats gets repriced.

Ricardo

1,617,104 views • 8 months ago

Marc Andreessen: "It's pretty likely in 5 years that AI chips will be cheap and plentiful." This is him counting the trillions going into the ground, the GPT-5 class model that now runs on two MacBooks, and why he thinks the chip shortage ends. His words: "There's just hundreds of billions or at this point trillions of dollars maybe going into the ground in all these things. And so the per unit cost of the AI companies are going to drop like a rock over the course of the next decade." "If you track the capability of the leading edge models over time what you find is after 6 or 12 months there's a small model that's just as capable." "The new version of Kimi is a reasoning model that is at least according to the benchmark so far is basically a replication of the reasoning capabilities of GPT5." "Shrunk down to be able to run on either it's like one MacBook or two MacBooks" "They're now so valuable, generating so many profits that it's the bat signal of all time to the rest of the chip industry to figure out how to advance the state-of-the-art AI chips." "Extremely positive for the economics of the kinds of companies that we invest in." Bookmark & watch today ↓

Marc Andreessen: "It's pretty likely in 5 years that AI chips will be cheap and plentiful." This is him counting the trillions going into the ground, the GPT-5 class model that now runs on two MacBooks, and why he thinks the chip shortage ends. His words: "There's just hundreds of billions or at this point trillions of dollars maybe going into the ground in all these things. And so the per unit cost of the AI companies are going to drop like a rock over the course of the next decade." "If you track the capability of the leading edge models over time what you find is after 6 or 12 months there's a small model that's just as capable." "The new version of Kimi is a reasoning model that is at least according to the benchmark so far is basically a replication of the reasoning capabilities of GPT5." "Shrunk down to be able to run on either it's like one MacBook or two MacBooks" "They're now so valuable, generating so many profits that it's the bat signal of all time to the rest of the chip industry to figure out how to advance the state-of-the-art AI chips." "Extremely positive for the economics of the kinds of companies that we invest in." Bookmark & watch today ↓

Fireside Alpha

79,776 views • 1 month ago

$AMD's heading to $5T MC LT| Lowest $/M tokens 🧵 The real reason why Institutions are FOMOing into AMD while other Semi stocks are underperforming ($NVDA $AVGO) Not Financial Advice! DYOR! Under Dr. Lisa Su’s leadership, AMD has transformed from a distant challenger into a formidable force in AI infrastructure, delivering the industry’s most compelling TCO story for high-volume inference. Her clear vision open ecosystems, aggressive annual roadmaps, rack-scale innovation, and relentless focus on tokens-per-dollar has positioned AMD’s Helios racks as the go-to solution for hyperscalers and AI natives struggling with exploding token costs, collapsing the cost down to $0.0003-$0.0005/M tokens. I will link various threads on this analysis to supply chain and wafer ratio if you are interested in understanding the full picture. In the last 3-4 months, explosive Agentic AI demand significantly increased Inference demand for Agentic AI models with 5-10 agents. If you are a listener of CNBC or Bloomberg, u should know enterprises and companies are complaining abt cost of token, and how it starts to spike up way too much to make sense. The fact that most data center today are run by $NVDA Chips, where the cost is way too high for Training or Inference. 1. Token cost Here are some quick comp, so u understand why $META OpenAI Anthropic $MSFT $AMZN Softbank $GOOGL and many more small to medium AI Natives are buying AMD CPUs and GPUs as much as they want, or pretty much AMD chips are sold out for the next 3-5 years. Inference (Cost per Million Tokens) ~$NVDA B200 / HGX: ~$0.02–$0.08 on optimized workloads (FP4/MXFP4, speculative decoding). Significant improvement over Hopper but still premium-priced. GB200 NVL72 rack-scale: $0.05–$0.25+ ~$AMD Helios Racks: $0.0003-$0.0005 per M tokens, dramatically lower than NVIDIA equivalents in owned infra. MI355X node-level: Up to 40% more tokens per dollar vs. competing solutions ( B200), driven by higher memory capacity (up to 288GB+ HBM), strong bandwidth, and lower acquisition costs. Training ~$NVDA Rubin Rack is estimated $0.7-$1.2/M Tokens ~$AMD Helios Rack is estimated $0.65-$1.0/M Tokens 2. Why Hyperscalers and AI Natives Are Choosing AMD Token consumption (especially Agentic) is outpacing even NVIDIA’s efficiency gains, making diversification mandatory for economic viability. Massive deals reflect this reality like $META, OpenAI, $MSFT, Softbank, $AMZN, Oracle, LumaAI, G42... Dr. Lisa Su’s Vision in Action: Since taking the helm, Su has driven AMD’s turnaround with disciplined execution, annual GPU cadence (MI300 → MI350 → MI400), full-stack software (ROCm 7), open ecosystems (UALink, OCP designs), and customer-centric rack-scale solutions like Helios. Her emphasis on “tokens per dollar” and TCO has turned AMD into the pragmatic choice for sustainable AI scaling. Power/Energy Efficiency: ~Helios Rack-level is estimated at 120kW-140kW with 50% more HBM4 where Inference and Training cost matter ~Rubin Rack-Level is estimated at 160kW-230kw AMD Helios shines in owned TCO, memory density, and energy flexibility at hyperscale. Cost to build 1GW data center 1GW Helios Rack full build is estimated $30-$35B 1GW Rubin Rack full build is estimated $45-$55B 3. Superior CPUs to pair with GPUs on massive scale 5-10-20GW Agentic AI. autonomous, multi-step workflows with orchestration, tool use, parallel agents, data movement, and enterprise integration has dramatically increased the importance of strong host CPUs alongside GPUs. This shifts the CPU-to-GPU ratio higher and makes balanced systems critical toward 1:1 to 5:1 as enterprises testing more than 5-10 agents. AMD EPYC Venice excels ~Leadership core density (up to 256 Zen 6 cores per socket) for running many agents in parallel, orchestration layers, and high-throughput control-plane tasks. ~Superior performance-per-core and power efficiency ( up to 2.1x higher perf/core and 2.26x better SPECpower vs. NVIDIA Grace in benchmarks). ~Tight integration in Helios: One Venice CPU + multiple MI450 GPUs per node, enabling efficient data feeding to GPUs ("zero-copy"), parallel execution, and full rack utilization for complex agentic loops. Hyperscalers (Meta, Microsoft, Amazon, Google, Softbank) and AI natives (OpenAI, Anthropic...) are adopting high-core EPYC at scale specifically for these agentic demands, as CPUs now handle a larger share of non-model work (orchestration, policy enforcement, tool calls). This complements AMD’s lower-cost GPUs for overall TCO wins. Conclusion: NVIDIA’s Vera Rubin cannot compete with a 2 years old EPYC Turin, but AMD under Dr. Lisa Su has engineered the lowest cost-per-million-tokens, highly competitive energy-efficient solutions, and superior CPU orchestration for agentic AI at scale with Helios. Dr. Su has championed this shift since at least 2023, foreseeing the rise of agentic workflows that demand far more orchestration, parallel agents, and balanced compute well before the industry fully embraced it. Her long-term vision of AI moving from simple prompts to always-on, multi-agent systems has driven AMD’s investments in high-core EPYC CPUs and integrated rack-scale solutions, perfectly positioning the company for today’s realities. Hyperscalers and AI natives effectively have no choice but to buy more AMD system for Agentic AI as leadership in economical, power-aware, high-volume internal + agentic use. However, due to supply constraints where Supply is far behind Demand, this makes multi-vendor reality along with in-house chips drive faster industry progress, lower overall costs, and better sustainability. Not Financial Advice! DYOR! Video source: Microsoft Build 2026

$AMD's heading to $5T MC LT| Lowest $/M tokens 🧵 The real reason why Institutions are FOMOing into AMD while other Semi stocks are underperforming ($NVDA $AVGO) Not Financial Advice! DYOR! Under Dr. Lisa Su’s leadership, AMD has transformed from a distant challenger into a formidable force in AI infrastructure, delivering the industry’s most compelling TCO story for high-volume inference. Her clear vision open ecosystems, aggressive annual roadmaps, rack-scale innovation, and relentless focus on tokens-per-dollar has positioned AMD’s Helios racks as the go-to solution for hyperscalers and AI natives struggling with exploding token costs, collapsing the cost down to $0.0003-$0.0005/M tokens. I will link various threads on this analysis to supply chain and wafer ratio if you are interested in understanding the full picture. In the last 3-4 months, explosive Agentic AI demand significantly increased Inference demand for Agentic AI models with 5-10 agents. If you are a listener of CNBC or Bloomberg, u should know enterprises and companies are complaining abt cost of token, and how it starts to spike up way too much to make sense. The fact that most data center today are run by $NVDA Chips, where the cost is way too high for Training or Inference. 1. Token cost Here are some quick comp, so u understand why $META OpenAI Anthropic $MSFT $AMZN Softbank $GOOGL and many more small to medium AI Natives are buying AMD CPUs and GPUs as much as they want, or pretty much AMD chips are sold out for the next 3-5 years. Inference (Cost per Million Tokens) ~$NVDA B200 / HGX: ~$0.02–$0.08 on optimized workloads (FP4/MXFP4, speculative decoding). Significant improvement over Hopper but still premium-priced. GB200 NVL72 rack-scale: $0.05–$0.25+ ~$AMD Helios Racks: $0.0003-$0.0005 per M tokens, dramatically lower than NVIDIA equivalents in owned infra. MI355X node-level: Up to 40% more tokens per dollar vs. competing solutions ( B200), driven by higher memory capacity (up to 288GB+ HBM), strong bandwidth, and lower acquisition costs. Training ~$NVDA Rubin Rack is estimated $0.7-$1.2/M Tokens ~$AMD Helios Rack is estimated $0.65-$1.0/M Tokens 2. Why Hyperscalers and AI Natives Are Choosing AMD Token consumption (especially Agentic) is outpacing even NVIDIA’s efficiency gains, making diversification mandatory for economic viability. Massive deals reflect this reality like $META, OpenAI, $MSFT, Softbank, $AMZN, Oracle, LumaAI, G42... Dr. Lisa Su’s Vision in Action: Since taking the helm, Su has driven AMD’s turnaround with disciplined execution, annual GPU cadence (MI300 → MI350 → MI400), full-stack software (ROCm 7), open ecosystems (UALink, OCP designs), and customer-centric rack-scale solutions like Helios. Her emphasis on “tokens per dollar” and TCO has turned AMD into the pragmatic choice for sustainable AI scaling. Power/Energy Efficiency: ~Helios Rack-level is estimated at 120kW-140kW with 50% more HBM4 where Inference and Training cost matter ~Rubin Rack-Level is estimated at 160kW-230kw AMD Helios shines in owned TCO, memory density, and energy flexibility at hyperscale. Cost to build 1GW data center 1GW Helios Rack full build is estimated $30-$35B 1GW Rubin Rack full build is estimated $45-$55B 3. Superior CPUs to pair with GPUs on massive scale 5-10-20GW Agentic AI. autonomous, multi-step workflows with orchestration, tool use, parallel agents, data movement, and enterprise integration has dramatically increased the importance of strong host CPUs alongside GPUs. This shifts the CPU-to-GPU ratio higher and makes balanced systems critical toward 1:1 to 5:1 as enterprises testing more than 5-10 agents. AMD EPYC Venice excels ~Leadership core density (up to 256 Zen 6 cores per socket) for running many agents in parallel, orchestration layers, and high-throughput control-plane tasks. ~Superior performance-per-core and power efficiency ( up to 2.1x higher perf/core and 2.26x better SPECpower vs. NVIDIA Grace in benchmarks). ~Tight integration in Helios: One Venice CPU + multiple MI450 GPUs per node, enabling efficient data feeding to GPUs ("zero-copy"), parallel execution, and full rack utilization for complex agentic loops. Hyperscalers (Meta, Microsoft, Amazon, Google, Softbank) and AI natives (OpenAI, Anthropic...) are adopting high-core EPYC at scale specifically for these agentic demands, as CPUs now handle a larger share of non-model work (orchestration, policy enforcement, tool calls). This complements AMD’s lower-cost GPUs for overall TCO wins. Conclusion: NVIDIA’s Vera Rubin cannot compete with a 2 years old EPYC Turin, but AMD under Dr. Lisa Su has engineered the lowest cost-per-million-tokens, highly competitive energy-efficient solutions, and superior CPU orchestration for agentic AI at scale with Helios. Dr. Su has championed this shift since at least 2023, foreseeing the rise of agentic workflows that demand far more orchestration, parallel agents, and balanced compute well before the industry fully embraced it. Her long-term vision of AI moving from simple prompts to always-on, multi-agent systems has driven AMD’s investments in high-core EPYC CPUs and integrated rack-scale solutions, perfectly positioning the company for today’s realities. Hyperscalers and AI natives effectively have no choice but to buy more AMD system for Agentic AI as leadership in economical, power-aware, high-volume internal + agentic use. However, due to supply constraints where Supply is far behind Demand, this makes multi-vendor reality along with in-house chips drive faster industry progress, lower overall costs, and better sustainability. Not Financial Advice! DYOR! Video source: Microsoft Build 2026

Mike

145,992 views • 3 months ago

Jensen Huang just said the semiconductor industry needs to grow 5-10x over the next decade and he named the exact bottlenecks investors should be watching (Save this). His argument is that this isn't a normal cyclical boom that will bust like past chip cycles because it's driven by a structural shift, the world needs a whole new intelligence layer of infrastructure on top of energy, internet, and roads and that layer runs entirely on chips. Using WSTS's 2026 industry estimate of $1.5 trillion, a 5-10x expansion would put the semiconductor industry at roughly $7.5 trillion to $15 trillion within ten years. He specifically called out five areas that are already in shortage, memory, storage, optical interconnects, packaging, and TSMC foundry capacity, and said the whole industry is short because this is infrastructure demand, not seasonal demand. Here's who stands to benefit in each of those five bottlenecks he named. Memory and storage is the most direct beneficiary, since Micron is the only major US based DRAM and NAND producer and has been reporting surging prices tied to AI server demand while SK Hynix and Samsung, both foreign, dominate the high bandwidth memory that feeds directly into GPUs like Nvidia's. Optical interconnects benefits companies like Coherent and Lumentum which make the optical transceivers moving data between GPU clusters, alongside Credo Technology, which makes high speed interconnect chips for the same data movement problem inside AI clusters. Packaging is where Nvidia's own chips get bottlenecked, since advanced packaging capacity mostly sits with TSMC, but Amkor Technology and ASE Technology are the independent outsourced packaging players that pick up overflow demand when TSMC can't keep up. Chip equipment makers that supply the tools needed to actually build more capacity across all these bottlenecks also benefit, including ASML for lithography, and Lam Research, KLA Corporation and Applied Materials for etching, deposition, and inspection equipment, since a 5-10x industry expansion requires massive new fab buildout that all runs through these companies first. Milk Road Pro is tracking the entire supply chain bottlenecks, if you want our entire AI trades around this, you can come join us just for a dollar.

Jensen Huang just said the semiconductor industry needs to grow 5-10x over the next decade and he named the exact bottlenecks investors should be watching (Save this). His argument is that this isn't a normal cyclical boom that will bust like past chip cycles because it's driven by a structural shift, the world needs a whole new intelligence layer of infrastructure on top of energy, internet, and roads and that layer runs entirely on chips. Using WSTS's 2026 industry estimate of $1.5 trillion, a 5-10x expansion would put the semiconductor industry at roughly $7.5 trillion to $15 trillion within ten years. He specifically called out five areas that are already in shortage, memory, storage, optical interconnects, packaging, and TSMC foundry capacity, and said the whole industry is short because this is infrastructure demand, not seasonal demand. Here's who stands to benefit in each of those five bottlenecks he named. Memory and storage is the most direct beneficiary, since Micron is the only major US based DRAM and NAND producer and has been reporting surging prices tied to AI server demand while SK Hynix and Samsung, both foreign, dominate the high bandwidth memory that feeds directly into GPUs like Nvidia's. Optical interconnects benefits companies like Coherent and Lumentum which make the optical transceivers moving data between GPU clusters, alongside Credo Technology, which makes high speed interconnect chips for the same data movement problem inside AI clusters. Packaging is where Nvidia's own chips get bottlenecked, since advanced packaging capacity mostly sits with TSMC, but Amkor Technology and ASE Technology are the independent outsourced packaging players that pick up overflow demand when TSMC can't keep up. Chip equipment makers that supply the tools needed to actually build more capacity across all these bottlenecks also benefit, including ASML for lithography, and Lam Research, KLA Corporation and Applied Materials for etching, deposition, and inspection equipment, since a 5-10x industry expansion requires massive new fab buildout that all runs through these companies first. Milk Road Pro is tracking the entire supply chain bottlenecks, if you want our entire AI trades around this, you can come join us just for a dollar.

Milk Road AI

29,530 views • 1 month ago