Who is on this subjectordered by the date of their first quote here
2 of 6 lanes rest on fewer than 5 quotes and are marked thin. Offsets are days from the middle first-quote date, 7 Jun 2023 — a date, and nothing else. It is not a claim about who reached a view first.
-
Gavin Baker
48
26 Nov 2019 – 4 Aug 2026
-1289 d
- Dylan Patel 62 2 Feb 2023 – 22 Jul 2026 -125 d
- Stanley Druckenmiller 1 thin 7 Jun 2023 – 7 Jun 2023 +0 d
-
Brad Gerstner
99
16 Jun 2023 – 29 May 2026
+9 d
-
Bill Gurley
23
22 Feb 2024 – 10 Apr 2026
+260 d
-
David Tepper
3
thin
26 Sep 2024 – 18 Sep 2025
+477 d
The chronologysourced and dated, oldest first
-
-
Gavin Baker
Baker quantifies Intel's manufacturing lead as 15-20% advantage that Taiwan Semi overtook in 2018.
“And in 2018, and it's important that the Taiwan Semi seven nanometer node is equivalent to the Intel 10 nanometer node.”
-
-
-
Gavin Baker
Baker says Nasdaq returns in 2021 were dominated by Google, Microsoft, Nvidia, and Tesla, requiring 60% concentration.
“The market return, particularly for the Nasdaq, was really dominated by a few stocks, Google, Microsoft, Nvidia, Tesla.”
-
-
-
Dylan Patel
Patel says memory now represents 30-50% of GPU manufacturing cost and NVIDIA applies 4x markup.
“And and now today, if you look at it, the cost of memory is 30 to 50% of the GPU cost. Great cost. And then obviously, NVIDIA does a four x markup on their”
-
Stanley Druckenmiller
Druckenmiller argues AI secular moves last years not months, expecting NVIDIA ownership for two to three years minimum.
“If this is a secular move, if this thing is real, you just don't have ten month moves. That's not how it works. Even the .com bubble lasted two, two and a half years.”
-
Brad Gerstner
Gerstner says AI stocks including Nvidia are now at or within 10% of his year-end price targets.
“a lot of the AI related stocks are now at or within 10% of our end of year price target. And when you when you see that, take something like Nvidia,”
-
Brad Gerstner
Gerstner says AI stocks including Nvidia are now at or within 10% of his year-end price targets.
“a lot of the AI related stocks are now at or within 10% of our end of year price target. And when you when you see that, take something like Nvidia,”
-
Brad Gerstner
Gerstner notes Nvidia revised Q2 datacenter guidance from $7 billion to $11 billion, defying negative expectations.
“People thought data center revenues this year were gonna be negative, and they just revised their guide from 7,000,000,000 in q two to to 11,000,000,000 in q two.”
-
Dylan Patel
Patel says NVIDIA is shipping more GPU flops this year than in its entire data center history combined.
“There's more GPU flops shipping this year that NVIDIA shipped their entire history for the data center.”
-
Dylan Patel
Patel forecasts NVIDIA will ship over one million H100 and A100 GPUs this year.
“NVIDIA is gonna ship over a million, h one hundreds plus a a one hundreds this year.”
-
Dylan Patel
Patel states NVIDIA will build and sell approximately 400,000 H100s in Q3.
“they're gonna build at about 400,000 GPUs and and sell about 400,000 h one hundreds in q three.”
-
Dylan Patel
Patel explains NVIDIA strategically allocates GPUs to new entrants like Inflection to maintain multiple customers and competition.
“it was like, oh, well, who's this random startup? Like, do I wanna get them 22,000 GPUs this year? Well, actually, yeah.”
-
Brad Gerstner
Gerstner notes consensus forecast predicted negative 6% NVIDIA data center growth and hard landing in early 2024.
“if we look at the start of the year, the consensus forecast for NVIDIA was that data center growth would be negative 6% this year.”
-
Brad Gerstner
Gerstner notes consensus forecast predicted negative 6% NVIDIA data center growth and hard landing in early 2024.
“if we look at the start of the year, the consensus forecast for NVIDIA was that data center growth would be negative 6% this year.”
-
Brad Gerstner
Gerstner argues NVIDIA at $400 from $100 looks expensive but trades only 20x next year's earnings.
“But the fact of the matter is on consensus forecast for next year, it's trading just over $20 or 20 times earnings.”
-
Brad Gerstner
Gerstner says professional analysts predicted NVIDIA data center revenue would decline 6 percent, highlighting how wrong forecasts were.
“The consensus estimate for NVIDIA this year was that data center growth revenue growth was gonna be negative 6%. Right? That's the pros. They got it that wrong.”
-
Brad Gerstner
Gerstner says professional analysts predicted NVIDIA data center revenue would decline 6 percent, highlighting how wrong forecasts were.
“The consensus estimate for NVIDIA this year was that data center growth revenue growth was gonna be negative 6%. Right? That's the pros. They got it that wrong.”
-
Dylan Patel
Patel says NVIDIA will sell over 3 million total GPUs next year and over a million H100s this year.
“NVIDIA is gonna sell well over 3,000,000 total GPUs next year, over a million H100s this year alone. There's a lot of GPU capacity coming online. It's an incredible amount.”
-
Dylan Patel
Patel states NVIDIA will sell over one million H100s this year and over three million total GPUs next year.
“NVIDIA is gonna sell well over 3,000,000 total GPUs next year, over a million H100s this year alone. There's a lot of GPU capacity coming online.”
-
Dylan Patel
Patel predicts NVIDIA will announce a new chip in March that is three to four times better than current generation.
“NVIDIA is releasing a new chip, you know, in you know, they're gonna announce it in March and they're gonna release it, you know, and ship it, you know, q two, q three next year anyways. Right? And that chip will probably be three or four times as good.”
-
-
-
Brad Gerstner
Gerstner says MANG (Microsoft, Amazon, Nvidia, Google) went from near-zero VC investing to $25 billion last year.
“we have this explosion in venture capital investing coming out of four companies, right? He called them MANG. Microsoft, Amazon, know, Nvidia and Google.”
-
Brad Gerstner
Gerstner argues MANG participation is distorting market prices compared to arm's length transactions with financial investors.
“I do think that it is a really important thing that you're pointing out, which is at a very minimum, I think we can say that the participation of MANG, Microsoft, Amazon, Nvidia, and Google, is distorting the price in the market in a way that wouldn't occur if it was all arm's length transaction with financial investors.”
-
Brad Gerstner
Gerstner recalls NVIDIA data center revenue forecasts shifted from -6% to +25% last year.
“Think about NVIDIA at the start of last year. Data center revenues were expected to shrink by 6%. Right?”
-
Brad Gerstner
Gerstner recalls NVIDIA data center revenue forecasts shifted from -6% to +25% last year.
“Think about NVIDIA at the start of last year. Data center revenues were expected to shrink by 6%. Right?”
-
Brad Gerstner
Gerstner notes NVIDIA traded at 20x earnings last year, its lowest multiple ever.
“So NVIDIA ended last year at 20 times earnings. Its lowest multiple of earnings it's ever traded at, despite the fact that it's the the the purest play AI name in in the space.”
-
Brad Gerstner
Gerstner notes NVIDIA traded at 20x earnings last year, its lowest multiple ever.
“So NVIDIA ended last year at 20 times earnings. Its lowest multiple of earnings it's ever traded at, despite the fact that it's the the the purest play AI name in in the space.”
-
Brad Gerstner
Gerstner says Altimeter's NVIDIA numbers remain well above Street consensus for this year.
“Our numbers are still well above the street for this year. I think there's a lot of doubt. You know, the the common, discussion at the end of last year”
-
Brad Gerstner
Gerstner says NVIDIA holds 90% plus market share and trades at historical average multiples.
“So if you wanna express that bet, right, which is AI leadership at 25 to 30 times earnings, which is consistent with historical averages for NVIDIA.”
-
Brad Gerstner
Gerstner says NVIDIA holds 90% plus market share and trades at historical average multiples.
“So if you wanna express that bet, right, which is AI leadership at 25 to 30 times earnings, which is consistent with historical averages for NVIDIA.”
-
Brad Gerstner
Gerstner says Altimeter's best AI bets over the past two years were NVIDIA, Microsoft, and Snowflake.
“the best bets on AI over the course of the last eighteen months or two years, we were fortunate to be in, were NVIDIA and Microsoft, Snowflake.”
-
Brad Gerstner
Gerstner says Altimeter's best AI bets over the past two years were NVIDIA, Microsoft, and Snowflake.
“the best bets on AI over the course of the last eighteen months or two years, we were fortunate to be in, were NVIDIA and Microsoft, Snowflake.”
-
Brad Gerstner
Gerstner claims Meta is the largest AI profit beneficiary in public markets outside NVIDIA.
“They are the single largest beneficiary of AI from a profit perspective in the public markets today outside of NVIDIA.”
-
Brad Gerstner
Gerstner claims Meta is the largest AI profit beneficiary in public markets outside NVIDIA.
“They are the single largest beneficiary of AI from a profit perspective in the public markets today outside of NVIDIA.”
-
Brad Gerstner
Gerstner notes critics compared AI training spending to 2001 dark fiber, claiming pullforward had occurred.
“Everybody said at the last end of last year, we had pulled forward all of the training. You know, this was dark fiber from Internet in 2001.”
-
Brad Gerstner
Gerstner says NVIDIA traded at 20x earnings, its lowest multiple ever, and is now up 25-30 percent.
“It was overvalued despite the fact it was trading at 20 times earnings, its lowest multiple in history. Now it's up 25 or 30% after opening, the year down.”
-
Brad Gerstner
Gerstner cites Microsoft, Amazon, and Meta all confirming massive AI inference and future infrastructure investments in NVIDIA.
“It's up because Susan Lee from Meta just told us yesterday that they're going to make huge future investments in infrastructure, read NVIDIA, in order to support all of these initiatives.”
-
Brad Gerstner
Gerstner cites Microsoft, Amazon, and Meta all confirming massive AI inference and future infrastructure investments in NVIDIA.
“It's up because Susan Lee from Meta just told us yesterday that they're going to make huge future investments in infrastructure, read NVIDIA, in order to support all of these initiatives.”
-
Bill Gurley
Gurley calls the Altman-Masa plan to compete with Nvidia radical given AMD and others have decades head start.
“There are already people competing with Nvidia. AMD's competing with Like, there are other people that have somewhat of a head start. Like decades.”
-
Brad Gerstner
Gerstner quotes Jensen predicting data center infrastructure will double from $1T to $2T in four to five years.
“And over the course of the next four to five years, we'll have 2,000,000,000,000 of data centers powering software around the world and it will all be accelerated compute.”
-
Brad Gerstner
Gerstner argues consensus Nvidia forecast assumes market share drops from 55% to 26% by 2028.
“Okay, so the consensus forecast that has the stock at $700 a share assumes, if you believe this TAM to be accurate, assumes that their share will go from 55% today to 26% in 2028.”
-
Brad Gerstner
Gerstner says NVIDIA EPS was forecast at $5.70 but ended up at $25, far exceeding expectations.
“The EPS at the beginning of last year, the earnings per share was expected to be $5.70, and now it looks like it's gonna be $25.”
-
Brad Gerstner
Gerstner says Microsoft and NVIDIA remain relatively safe investments for AI exposure amid uncertainty.
“I think that if you're an investor, owning some of the companies that you think you're really confident are going to continue to compound and benefit from AI, the Microsofts of the world, the Nvidias of the world, seems to me like a relatively safe place.”
-
Brad Gerstner
Gerstner states consensus Nvidia estimate for next year is around $3.50 per share.
“the consensus estimate for next year for Nvidia, I think is around $3.50 per share.”
-
Brad Gerstner
Gerstner states consensus Nvidia estimate for next year is around $3.50 per share.
“the consensus estimate for next year for Nvidia, I think is around $3.50 per share.”
-
Brad Gerstner
Gerstner's Nvidia estimate is closer to $5 per share, ahead of consensus since fall 2022.
“we're we're closer to $5 a share. We've been ahead of the consensus estimate since the fall of twenty two.”
-
Brad Gerstner
Gerstner says AI customers must place orders five years in advance and secure power and data centers.
“they have to place orders five years in advance for these products. They have to go find power. They have to find data centers”
-
Bill Gurley
Gurley reports Nvidia alone accounts for 5% of S&P 500 returns while 493 other companies contribute only 4%.
“The other mag six are responsible for 5%. And then the other 493 companies in the S and P 500 are responsible for 4% of the 14% year to date returns.”
-
Gavin Baker
Baker explains xAI achieved coherence in a 100,000 GPU cluster with Hopper without next-generation networking technologies.
“And because of that, they're able to get enough density that they could effectively make a 100,000 GPU cluster coherent even with Hopper without next generation networking technologies from NVIDIA, Broadcom, and others.”
-
Gavin Baker
Baker explains xAI achieved coherence in a 100,000 GPU cluster with Hopper without next-generation networking technologies.
“And because of that, they're able to get enough density that they could effectively make a 100,000 GPU cluster coherent even with Hopper without next generation networking technologies from NVIDIA, Broadcom, and others.”
-
Brad Gerstner
Gerstner proposes dedicating individual nuclear reactors at Diablo to major tech companies for their data centers.
“I could imagine if you had four more reactors sitting here, right, if you could land them at the right price, you could have one reactor for Meta, one for Amazon, one for Microsoft, one for Nvidia, you could have the data center sitting right next to them.”
-
Brad Gerstner
Gerstner frames NVIDIA debate as 6 million GPUs versus bearish 4.5 million estimate for next year.
“We're kind of at 6,000,000 GPUs for next year. The bearish people are at like 4,500,000. The numbers will ultimately tell.”
-
David Tepper
Tepper says he has no confidence in NVIDIA earnings forecasts for 2026 and 2027.
“I have no idea in '26 and '27. I am no idea. Okay? I mean, as if I was but I don't believe I don't believe my analysts.”
-
David Tepper
Tepper says he likes NVIDIA at current price levels despite earlier concerns.
“I I I kinda like NVIDIA, the price. Okay? So don't get me wrong. I I I could just love it.”
-
Dylan Patel
Patel says China still receives over a million NVIDIA H20 GPUs annually despite October sanctions.
“Right? Because you have you have sanctions on how many NVIDIA GPUs you can get in. Now, they're still north of a million a year.”
-
Dylan Patel
Patel says China could centralize over a million NVIDIA H20 chips into one data center if scale-pilled.
“they can centralize the chips like crazy. Right now, oh, oh, million chips that NVIDIA shipping in q three and q four, the h twenty, let's just put them all in this one data center.”
-
Bill Gurley
Gurley argues NVIDIA's moat is misunderstood because it spans the full stack, not just chips.
“one of the most misunderstood things about NVIDIA is how deep the true NVIDIA moat is, right?”
-
Brad Gerstner
Gerstner notes NVIDIA's 2023 revenue was $60B versus analyst forecasts of $26B in January 2023.
“The forecast for Nvidia at that dinner in January 2023 was that you would do 26,000,000,000 of revenue for the year 2023. You did 60,000,000,000, right?”
-
Brad Gerstner
Gerstner notes NVIDIA's 2023 revenue was $60B versus analyst forecasts of $26B in January 2023.
“The forecast for Nvidia at that dinner in January 2023 was that you would do 26,000,000,000 of revenue for the year 2023. You did 60,000,000,000, right?”
-
Bill Gurley
Gurley says analysts miss NVIDIA's ecosystem thinking and multi-year planning beyond current architecture.
“I think analysts always focus on the current architectural bet. But I think one of the biggest takeaways from this conversation is that we're thinking about the entire ecosystem and many years out.”
-
Brad Gerstner
Gerstner says NVIDIA generates $4M revenue and $2M profit per employee, unprecedented efficiency.
“You know NVIDIA is in a league of its own really, know, at about 4,000,000 of revenue per employee, about 2,000,000 of profits or free cash flow per employee.”
-
Brad Gerstner
Gerstner says NVIDIA generates $4M revenue and $2M profit per employee, unprecedented efficiency.
“You know NVIDIA is in a league of its own really, know, at about 4,000,000 of revenue per employee, about 2,000,000 of profits or free cash flow per employee.”
-
Brad Gerstner
Gerstner says CUDA library now has over 300 industry-specific acceleration algorithms across diverse fields
“he talked about, you know, the the the CUDA library now has over 300 industry specific acceleration algorithms. Yep. Right? Where they deeply learn the industry.”
-
Bill Gurley
Gurley argues NVIDIA's competitive advantage is strongest in the largest systems, not at the edge.
“it appears to me that NVIDIA's competitive advantage is strongest where the size of the system is largest, which is another way of saying what Renee said. It's flipping it on its head.”
-
Bill Gurley
Gurley explains NVIDIA's networking, NVLink, and CUDA advantages emerge specifically in the largest system deployments.
“That's when the networking piece thrives. That's where NVLink thrives. That's where CUDA really comes alive in the biggest systems that are out there.”
-
Bill Gurley
Gurley realizes NVIDIA needs multiple large system companies like CoreWeave to maximize its competitive advantage.
“Like like like if if the biggest systems are where the biggest competitive advantage is, you need as many of these big system companies as you can possibly have.”
-
Brad Gerstner
Gerstner reports Jensen predicts inference will scale 100x to billion-x with 40% of NVIDIA revenue already from inference
“he said as a consequence of that, inference is going to a 100 x, thousand x, a million x, maybe even a billion x.”
-
Bill Gurley
Gurley observes Elon Musk obtained roughly 10% of NVIDIA's quarterly availability for x.ai despite massive demand.
“And he walks in and takes what equate sounds looks like about 10% of the quarter's availability.”
-
Brad Gerstner
Gerstner says Jensen confirmed GPU clusters are already at 200-300k scale and will reach 500k to 1 million
“I said, are we already at the phase of two and three hundred thousand cluster scale? And he said, yes. And then I said, and will we go to 500,000 a million?”
-
Brad Gerstner
Gerstner says Jensen claims NVIDIA can scale business 2-3x while increasing headcount only 20-25% using AI agents
“This idea that Jensen can scale the business two or three times with, you know, increasing the headcount by, you know, 20 or 25%. Right?”
-
Dylan Patel
Patel claims four to five percent of NVIDIA GPUs fail during initial burn-in testing within two weeks.
“Buy a good chip from NVIDIA, and then it ends up failing like constantly, right?”
-
Bill Gurley
Gurley says NVIDIA differentiation is greatest at largest cluster size for LLM pre-training.
“One, the NVIDIA differentiation as we've talked about is greatest at the largest cluster size. So if pre training, once again, I only equate this to LLMs, you know, I don't think of FSD problem.”
-
Brad Gerstner
Gerstner notes NVIDIA's Hopper demand remains exceptional and will continue into next year, defying expectations it would end.
“We may have people thought hopper demand would be over by now, frankly, and he said it's going to continue well into next year.”
-
Bill Gurley
Gurley states NVIDIA holds 98% of purchased AI workloads when captive workloads are excluded.
“If you just look at, I guess, workloads people are purchasing to do work on their own, so you take the captives out, you're at 98, right?”
-
Brad Gerstner
Gerstner says Jensen confirmed a trillion dollars of data center CPU replacement workloads over four years beyond AI training.
“he said, but we're also going to have a trillion dollars of CPU replacement, of data center replacement workloads over the course of the next four years.”
-
Brad Gerstner
Gerstner argues NVIDIA at 30 PE versus Cisco's 120 PE means comparison to 2000 bubble is uninformed.
“So you would have to think that there would be 70% PE compression from here or that their revenue was going to fall by 70% or that their earnings were going to fall by 70%.”
-
Dylan Patel
Patel says Google deployed water cooling years before NVIDIA and achieves higher reliability than NVIDIA GPUs.
“Google's brought in water cooling for years. Right? NVIDIA only just realized they needed water cooling on this generation. And Google's brought in a level of reliability that NVIDIA GPUs don't have.”
-
“with Blackwell, not only is it way, way, way faster, anywhere from 10 to 15 times on really large models for inference because they've optimized it for very large language models.”
-
Bill Gurley
Patel says Blackwell delivers five times performance TCO improvement in single year, accelerating LLM cost decline.
“At least that's what Blackwell is, and we'll see what Ruben does. But, you know, five x plus in a single year for performance TCO is an insane pace.”
-
Bill Gurley
Patel distinguishes NVIDIA's revenue from positive cash flows versus Cisco's dot-com era credit-funded telecom buildout
“Cisco's revenue, a lot of it was funded through private slash credit investments into building out telecom infrastructure. Right? When we look at NVIDIA's revenue sources, very little of it is private slash credit.”
-
Bill Gurley
Gurley notes NVIDIA's highest cost of goods sold is HBM memory, not TSMC silicon, which people don't realize.
“Their highest cost of goods sold is not TSMC, which is a thing that people don't realize. It's actually HBM memory, primarily SK for now also.”
-
Bill Gurley
Gurley warns 2026 is when the reckoning comes on whether AI infrastructure spending continues at current pace.
“It's gonna bring so many people with them. But 2026 is like where the reckoning comes. Right? But, you know, will will people keep spending like this?”
-
-
-
Gavin Baker
Baker reveals high bandwidth memory is a bigger cost component for NVIDIA than Taiwan Semi, with only two capable manufacturers.
“High bandwidth memory is a bigger part of NVIDIA's cogs on GPUs than Taiwan Semi is. And today, there's two companies that can make it, Hynix and Micron.”
-
Gavin Baker
Baker reveals high bandwidth memory is a bigger cost component for NVIDIA than Taiwan Semi, with only two capable manufacturers.
“High bandwidth memory is a bigger part of NVIDIA's cogs on GPUs than Taiwan Semi is. And today, there's two companies that can make it, Hynix and Micron.”
-
Dylan Patel
Patel reveals each country is capped at 50,000 GPUs for four years while NVIDIA makes 6 million annually.
“There is there is one obvious loophole, which is well, there's, like, strict caps. Right? Like, each country can only buy 50,000 GPUs for the next four years.”
-
Dylan Patel
Patel estimates Blackwell delivers 10-15x cost improvement for inference despite NVIDIA claiming 30x at GTC.
“But now, like, Blackwell, NVIDIA's pitching 10 to 15 x improvement in cost. It's like, well, you know, they're massaging the numbers marketing.”
-
Brad Gerstner
Gerstner contrasts buying NVIDIA in fall 2022 based purely on tech fundamentals versus today's layered policy risks.
“What do I do as an investor? You know, what did I do in the fall of twenty two that led me into NVIDIA in the first place?”
-
Brad Gerstner
Gerstner contrasts buying NVIDIA in fall 2022 based purely on tech fundamentals versus today's layered policy risks.
“What do I do as an investor? You know, what did I do in the fall of twenty two that led me into NVIDIA in the first place?”
-
Dylan Patel
Patel says NVIDIA made 4 million GPUs last year and will produce 7 million this year.
“Nvidia made over 4,000,000 GPUs last year, they're making over 7,000,000 this year, right? High end data center GPUs.”
-
Brad Gerstner
Gerstner says 3.7 million kids born yearly would get private accounts, with Uber, Dell, Oracle, and Nvidia contributing.
“And we have companies like Uber and Dell and many, many others, Oracle, Nvidia, etcetera, all say they'll contribute to the accounts of the kids of their employees.”
-
Brad Gerstner
Gerstner says 3.7 million kids born yearly would get private accounts, with Uber, Dell, Oracle, and Nvidia contributing.
“And we have companies like Uber and Dell and many, many others, Oracle, Nvidia, etcetera, all say they'll contribute to the accounts of the kids of their employees.”
-
Brad Gerstner
Gerstner says the UAE-US AI campus announced is five gigawatts, equivalent to 2.5 million GPUs.
“So in in Abu Dhabi, they announced a five gigawatt US AI you know, it's a UAE, US AI campus. It's an incredible architectural campus.”
-
Brad Gerstner
Gerstner estimates NVIDIA lost track to $50 billion China business from chip export ban.
“I think if you forecast out two or three years, they were on track to be a $50,000,000,000 business, I think in China.”
-
Brad Gerstner
Gerstner argues NVIDIA's $40 billion China business keeps them on CUDA and export bans risk forcing Taiwan conflict.
“It's $40,000,000,000 business NVIDIA in China to keep them on the CUDA ecosystem rather than allowing them to have a monopoly on the Huawei ecosystem.”
-
Brad Gerstner
Gerstner argues NVIDIA's $40 billion China business keeps them on CUDA and export bans risk forcing Taiwan conflict.
“It's $40,000,000,000 business NVIDIA in China to keep them on the CUDA ecosystem rather than allowing them to have a monopoly on the Huawei ecosystem.”
-
Brad Gerstner
Gerstner says NVIDIA had 80% China market share with developers in CUDA ecosystem, generating billions in US taxes.
“NVIDIA has you know, it went from 80% market share in China where all the developers, second largest developer market in the world, were were developing in CUDA.”
-
Brad Gerstner
Gerstner says NVIDIA had 80% China market share with developers in CUDA ecosystem, generating billions in US taxes.
“NVIDIA has you know, it went from 80% market share in China where all the developers, second largest developer market in the world, were were developing in CUDA.”
-
Dylan Patel
Patel describes the Bumpgate incident where NVIDIA laptop GPUs had defective solder balls.
“There's a generation of NVIDIA GPUs for laptops. Right? And and chips have solder balls on the bottom. Right?”
-
Dylan Patel
Patel explains Bumpgate issue where thermal expansion differences caused GPU solder ball failures in Apple laptops.
“And what ended up happening is because of that different rate of expansion, the solder balls connecting the chip and the board would crack.”
-
Dylan Patel
Patel explains AI hardware companies made architectural bets that failed when model architectures evolved in unexpected directions.
“But still the model's way too big to fit on it. This is, like, very simple. Right? You know, the same thing's happening in the other direction.”
-
“But, like, you know, can you give them no GPUs? No. They're gonna retaliate. Like, there is a middle ground, and, like, Huawei is eventually going to have a lot of production capacity,”
-
Dylan Patel
Patel lists NVIDIA's structural advantages across networking, HBM, process nodes, speed to market, and supplier negotiations.
“NVIDIA's gonna have better networking than you. They're gonna have better, HBM. They're gonna have better process node. They're gonna come to market faster.”
-
Dylan Patel
Patel argues competing with NVIDIA requires leaping forward beyond supply chain advantages across networking, memory, manufacturing and components.
“They're gonna have better negotiations with whether it's TSMC or SK Hynix and the memory and silicon side or all the rack people or, like, copper cables, everything, they're gonna have better cost efficiency.”
-
Dylan Patel
Patel explains competitors need 5x hardware advantage over NVIDIA but risk failure if AI workloads shift before shipping.
“So now you need to do something, you know, that will give you five x advantage, right, in hardware efficiency for a certain type of workload, and then pray the workload doesn't shift.”
-
Dylan Patel
Patel forecasts NVIDIA revenue exceeding $300 billion next year as infrastructure spending reaches nation-state scale.
“NVIDIA's revenue this year is gonna be, like, over $200,000,000,000, and next year expects over 300,000,000,000 plus Google's gonna spend, like, $50,000,000,000 on TPU data centers.”
-
Dylan Patel
Patel suggests NVIDIA should use its massive tax bill to invest directly in data center infrastructure despite customer conflicts.
“Now this is obviously gonna be, like, crazy because, like, now they're buying g p their own GPUs and putting them in data centers and doing stuff, and they're competing with their own customers,”
-
Brad Gerstner
Gerstner notes Jensen raised compute buildout forecast from $2 trillion to $3-4 trillion through decade end.
“Jensen said, you know, when he was on our pod last year, he said, we think the build out between now and the end of the decade's $2,000,000,000,000 of total compute.”
-
Brad Gerstner
Gerstner calculates 50% growth would bring NVIDIA to $300 billion datacenter revenue, about $8 per share.
“So the consensus for next year is 250,000,000,000 of data center revenue. If they grow up 50%, that's gonna be closer to 300,000,000,000 or closer to $8 a share.”
-
Brad Gerstner
Gerstner believes there's over 50% probability NVIDIA can sell Blackwell chips in China, a $50 billion market.
“China is a $50,000,000,000 market for NVIDIA unto itself. So I believe that the probability they're gonna be able to sell chips China is north of 50%.”
-
Brad Gerstner
Gerstner believes there's over 50% probability NVIDIA can sell Blackwell chips in China, a $50 billion market.
“China is a $50,000,000,000 market for NVIDIA unto itself. So I believe that the probability they're gonna be able to sell chips China is north of 50%.”
-
Dylan Patel
Patel states GB 200 servers from Nvidia cost well over $3 million, preventing time-slicing users on individual servers.
“These servers cost hundreds of thousands of dollars if nothing new. Know GB 200 from Nvidia cost well over $3,000,000 so the cost of these things is huge”
-
Dylan Patel
Patel notes new NVIDIA servers require 140 kilowatts per rack versus prior CPU data centers at kilowatts.
“Data centers for CPUs, 10 megawatts of rack, 12 megawatts of rack. Kilowatts. The new ones, kilowatts, right? Kilowatts of rack, sorry. The new NVIDIA servers, 140 kilowatts of rack.”
-
David Tepper
Tepper says he owns NVIDIA but trades in and out, not holding a constant position.
“I do own NVIDIA, but I go back and forth and back and forth a little bit, because I do, you know, I will trade a little bit.”
-
Dylan Patel
Patel estimates NVIDIA had north of $20 billion in China H20 revenue that had to be written off after ban.
“Our our revenue estimate for NVIDIA in China for just h 20 was north of 20,000,000,000 because that's what they were booking in capacity slash had to write off.”
-
Dylan Patel
Patel estimates NVIDIA had over $20 billion in H20 China revenue booked before ban and write-off.
“Our our revenue estimate for NVIDIA in China for just h 20 was north of 20,000,000,000 because that's what they were booking in capacity slash had to write off. Yeah. And then it got banned.”
-
Dylan Patel
Patel recounts that NVIDIA ordered Xbox production volume before receiving Microsoft's official order.
“No. No. No. Like, NVIDIA ordered the volume for the Xbox before Microsoft gave them the order.”
-
Dylan Patel
Patel explains Blackwell introduces third memory tier within Tensor Cores requiring complex programming model for full performance.
“Now with Blackwell, there's actually even a third tier of memory, which is memory within the Tensor Core.”
-
Dylan Patel
Patel reveals NVIDIA doubled system-level testing time for Blackwell compared to Hopper generation.
“NVIDIA learning from the issues on Hopper, with Blackwell, they actually doubled the time for system level test.”
-
Dylan Patel
Patel reveals NVIDIA's next generation will split inference into separate context processing and decode workloads, not training versus inference.
“NVIDIA's next generation actually has something very different. They're not saying, Hey, there's a training GPU and an inference GPU, right? Because either is fine.”
-
Dylan Patel
Patel explains NVIDIA is splitting inference into decode and prefill workloads, but provisioning for unknown future ratios is challenging.
“what sort of NVIDIA's pitching is like a split of inference into two workloads, and we'll see if they're successful. There's a lot of challenges on the infrastructure side.”
-
Brad Gerstner
Gerstner warns red flags arise when chip makers fund single customers who then buy their chips with that capital.
“So there's no other potential customers, and the buyer would not have had the ability to buy it but for that capital. That to me raises big red flags.”
-
Bill Gurley
Gurley reveals NVIDIA promised to buy any CoreWeave service availability they cannot sell to others.
“One of the more peculiar of all the deals is, and this was disclosed in a CoreWeave filing, was NVIDIA has promised to buy any of CoreWeave's service availability that they can't sell to anyone else. That is very unusual. That's not the same as making an investment.”
-
Brad Gerstner
Gerstner cites NVIDIA forecast of $200B revenue this year growing to $350B in five years, equating to specific gigawatts.
“This is the NVIDIA sell side forecast. Okay? Forecast this year is for about 200,000,000,000 in revenues, growing to about 350,000,000,000 in revenues over the next five years.”
-
Brad Gerstner
Gerstner cites NVIDIA forecast of $200B revenue this year growing to $350B in five years, equating to specific gigawatts.
“This is the NVIDIA sell side forecast. Okay? Forecast this year is for about 200,000,000,000 in revenues, growing to about 350,000,000,000 in revenues over the next five years.”
-
Bill Gurley
Gurley flags NVIDIA's unusual promise to buy unsold CoreWeave capacity as revealed in regulatory filings.
“One of the more peculiar of all the deals is, and this was disclosed in a CoreWeave filing, was NVIDIA has promised to buy any of CoreWeave's service availability that they can't sell to anyone else. That is very unusual. That's not the same as making an investment.”
-
Bill Gurley
Gurley flags NVIDIA's promise to buy unsold CoreWeave capacity as highly unusual arrangement.
“One of the more peculiar of all the deals is, and this was disclosed in a CoreWeave filing, was NVIDIA has promised to buy any of CoreWeave's service availability that they can't sell to anyone else.”
-
Bill Gurley
Gurley says NVIDIA promised to buy any unsold CoreWeave service availability, disclosed in CoreWeave filing, calling it very unusual.
“That is very unusual. That's not the same as making an investment. That could easily help CoreWeave with their debtors and and getting more debt financing.”
-
Brad Gerstner
Gerstner notes NVIDIA dropped to $92 on tariffs but has recovered to $180.
“Remember the deep sea moment? Stock stock was down 25%. And then on the tariff moment, the stock got down to $92 a share. It's at a 180 today.”
-
Dylan Patel
Patel reports AMD MI355 beats NVIDIA B200 in some document processing scenarios with open source software.
“in some cases, AMD actually does have a publicly usable open source implementation that beats NVIDIA even, right, with the MI355 versus V200, which is a surprise, right?”
-
Dylan Patel
Patel reports AMD MI355 beats NVIDIA B200 on performance-TCO with publicly usable open source software in certain use cases.
“AMD actually does have a publicly usable open source implementation that beats NVIDIA even, right, with the MI355 versus V200, which is a surprise, right?”
-
Dylan Patel
Patel reports AMD performance improved dramatically over two months, similar to NVIDIA Blackwell's software optimization trajectory.
“Over two months, the performance of AMD dramatically improved because they went from, hey, software problems are always a thing, to over a two month period, they've dramatically improved.”
-
Dylan Patel
Patel quantifies that 20% performance per watt advantage translates to only 4% TCO difference on NVIDIA deployments.
“But capital is far more of a limiting factor. Because if you have enough power, a 20% difference in performance per watt only ends up being a 4% difference in TCO.”
-
Dylan Patel
Patel claims NVIDIA takes a 5x markup on manufacturing cost, making power efficiency less significant for TCO.
“So, in most cases, don't on an NVIDIA deployment, right? That's where NVIDIA takes 5x markup on their manufacturing cost, right?”
-
Dylan Patel
Patel states NVIDIA takes 5x markup on manufacturing cost, making power differences more significant for AMD deployments.
“If it's another deployment, if it's AMD, then that that that 20% power difference might translate to eight or 9% TCO difference when you when you talk about power cost and data center capacity cost.”
-
Brad Gerstner
Gerstner breaks down OpenAI's compute commitments: $500 million to Nvidia, $300 million to AMD and Oracle, $250 billion to Azure.
“with, you know, big commitments, 500,000,000 to Nvidia, 300,000,000 to AMD and Oracle, $250,000,000,000 to Azure.”
-
Brad Gerstner
Gerstner breaks down OpenAI's compute commitments: $500 million to Nvidia, $300 million to AMD and Oracle, $250 billion to Azure.
“with, you know, big commitments, 500,000,000 to Nvidia, 300,000,000 to AMD and Oracle, $250,000,000,000 to Azure.”
-
Brad Gerstner
Gerstner warns that if AI revenues don't materialize, the public market AI trade will have a problem beyond NVIDIA.
“The market the public market does not have a way to play pure AI outside of NVIDIA. It is going to be watching.”
-
Brad Gerstner
Gerstner warns that if AI revenues don't materialize, the public market AI trade will have a problem beyond NVIDIA.
“The market the public market does not have a way to play pure AI outside of NVIDIA. It is going to be watching.”
-
Brad Gerstner
Gerstner notes NVIDIA added almost $200 billion of revenue in the last three years.
“You know, you've seen NVIDIA add almost $200,000,000,000 of revenue in the last three years. Yeah. Truly extraordinary.”
-
Brad Gerstner
Gerstner notes NVIDIA added almost $200 billion of revenue in the last three years.
“You know, you've seen NVIDIA add almost $200,000,000,000 of revenue in the last three years. Yeah. Truly extraordinary.”
-
Brad Gerstner
Gerstner argues NVIDIA at 23-24x forward earnings and Mag Seven at 25-30x are not bubble valuations.
“That is not the stuff that bubbles are made of. Right? I think the rest of the mag seven all trades somewhere between twenty five and thirty times, maybe met is at 21 times earnings.”
-
Brad Gerstner
Gerstner reveals Altimeter invested in NVIDIA in December 2022, up 10x as revenue grew from $30B to over $200B.
“The stock's up over ten ten times. 10 x over that period of time as the revenue went from roughly $30,000,000,000 a year to over $200,000,000,000 a year.”
-
Brad Gerstner
Gerstner says NVIDIA indicated a $100 billion run rate expected by end of next year.
“And they just told us that we should expect a run rate by the end of next year of a $100,000,000,000.”
-
“And the reason for that is it was by far the most complex product transition we've ever gone through in technology. Going from hopper to Blackwell, first you go from air cooled to liquid cooled.”
-
Gavin Baker
Baker details Blackwell's complexity: racks tripled in weight and quadrupled in power consumption versus Hopper.
“Going from hopper to Blackwell, first you go from air cooled to liquid cooled. The rack goes from weighing round numbers, a thousand pounds to 3,000 pounds.”
-
Gavin Baker
Baker explains Blackwell racks weigh 3,000 pounds and consume 130 kilowatts, equivalent to 130 American homes.
“The rack goes from weighing round numbers, a thousand pounds to 3,000 pounds. Goes from round numbers, 30 kilowatts, which is 30 American homes to 130 kilowatts, which is 130 American homes.”
-
Gavin Baker
Baker details Blackwell transition: racks triple in weight and quadruple in power from 30 to 130 kilowatts.
“Goes from round numbers, 30 kilowatts, which is 30 American homes to 130 kilowatts, which is 130 American homes.”
-
Gavin Baker
Baker explains Blackwell's complexity created deployment challenges that reasoning models had to bridge.
“Had reasoning not come along, there would have been no AI progress from mid twenty twenty four through essentially”
-
Gavin Baker
Baker argues Google has a temporary pre-training advantage as the lowest cost producer of tokens using advanced TPUs.
“So I think Google for sure has this temporary advantage right now from a pre training perspective. I think it's also important that they've been the lowest cost producer of tokens.”
-
Gavin Baker
Baker predicts xAI will release the first Blackwell-trained model in early 2026 because Elon builds fastest.
“One, we will see the first models trained on Blackwell in early twenty twenty six. I think the first Blackwell model will come from xAI.”
-
Gavin Baker
Baker predicts Blackwell models will be amazing because pre-training scaling laws remain intact.
“We know that scaling laws for pre training are intact, And this means the Blackwell models are going to be amazing.”
-
Gavin Baker
Baker explains Google pays Broadcom a 50 to 55 percent gross margin for back-end chip design and Taiwan Semi management.
“And the managing Taiwan Semi is like stamping out that house like Lennar or, you know, D. R. Horton. And for doing those two latter parts, Broadcom earns a 50 to 55% gross margin.”
-
Brad Gerstner
Gerstner says $250 billion on NVIDIA implies $500 billion total AI CapEx spending.
“one of the things that keeps investors like me up in the nights is a 250,000,000,000 being spent on NVIDIA on on the semi side of things.”
-
Brad Gerstner
Gerstner says $250 billion on NVIDIA implies $500 billion total AI CapEx spending.
“one of the things that keeps investors like me up in the nights is a 250,000,000,000 being spent on NVIDIA on on the semi side of things.”
-
-
-
Brad Gerstner
Gerstner notes NVIDIA consensus is 65% earnings growth in 2025 and 30% in 2027 at 25x multiple.
“The consensus estimate this year, I think, is that they're gonna grow earnings about 65%, maybe 30% in 2027. You said it's an earnings driven market, so they don't need to expand multiple.”
-
Brad Gerstner
Gerstner says CoreWeave stands alone among Neo Clouds due to its performance and strategic importance to NVIDIA's Rubin deployment.
“We think CoreWeave stands alone, in terms of its performance, the software that is built on top of the cloud, its execution capability, its strategic importance to NVIDIA in terms of deploying Rubin, etcetera.”
-
Brad Gerstner
Gerstner says CoreWeave stands alone among Neo Clouds due to its performance and strategic importance to NVIDIA's Rubin deployment.
“We think CoreWeave stands alone, in terms of its performance, the software that is built on top of the cloud, its execution capability, its strategic importance to NVIDIA in terms of deploying Rubin, etcetera.”
-
Brad Gerstner
Gerstner says AI capex jumped from $150B in 2023 to over $500B this year in committed purchase orders.
“In 2023, these companies spent $150,000,000,000 on capex building out data centers. This year, over 500,000,000,000. That's not speculative. That is purchase orders. Those are buildings.”
-
Brad Gerstner
Gerstner predicts Google's TPU will not match NVIDIA Rubin's token efficiency despite advantages.
“First, Google is a massive consumer of NVIDIA GPUs, Right? Not only for its own workloads, but also the workloads, you know, in Google Cloud.”
-
Brad Gerstner
Gerstner predicts Google's TPU will not match NVIDIA Rubin's token efficiency despite advantages.
“First, Google is a massive consumer of NVIDIA GPUs, Right? Not only for its own workloads, but also the workloads, you know, in Google Cloud.”
-
Brad Gerstner
Gerstner cites 5x inference and 3.5x training performance improvements on NVIDIA Rubin chips.
“He just improved his inference performance by five x, not even including Grok, which I think will give him, you know, another huge boost.”
-
Dylan Patel
Patel reports 10-15% of NVIDIA GPUs fail and require RMA within first two weeks of cluster deployment.
“When you first turn on the cluster, about ten to fifteen percent of them fail RMA in the first two weeks. Wow. And then that's fine. Like you have to receipt them, whatever.”
-
Dylan Patel
Patel says 10-15% of NVIDIA GPUs fail and need RMA in the first two weeks after cluster deployment.
“When you first turn on the cluster, about ten to fifteen percent of them fail RMA in the first two weeks. Wow. And then that's fine.”
-
Dylan Patel
Patel says Hopper GPU failure rates improved to 5% while Blackwell remains at 10-15%, expecting higher rates for next generation.
“hopper's now at 5%, but black belt's still 10 to 15%. Wow. Right? Actually started out higher than that. Sure. And when a new generation comes out, it's gonna be higher than 15%.”
-
Dylan Patel
Patel says NVIDIA must be 2x better than competitors to justify their 75% plus margins.
“NVIDIA recognizes they're they're the leader, they're the tent pole. Hey, in one respect, they can just run faster than everyone, but it's kind of hard to be two x better than Google or or OpenAI or whoever else's internal chip, right, to justify their, you know, 75% plus margins.”
-
Dylan Patel
Patel says NVIDIA must be 2-4x better than competitors to justify 75% margins and 4x pricing above costs.
“And then they have to be two x to four x better to justify four x better to justify their margins because that's what they're charging above cogs.”
-
Dylan Patel
Patel explains Jensen fears specialized chips could undercut NVIDIA's margins if they only made general-purpose GPUs.
“Jensen is very paranoid about losing. Right? These specializations if he just kept making his mainline chip would mean people could you know point point solutions for specific parts of the market would crush him on cost and performance, then he can't justify his margin.”
-
Dylan Patel
Patel explains NVIDIA acquired Grok to get engineering resources for multiple chip architectures.
“acquiring Rock is like how you get those resources to make more solutions for different parts of the market. And as far as like, are they threatened?”
-
Brad Gerstner
Gerstner says the big five will spend $650-700 billion on AI and are still token constrained with unfulfilled demand.
“650,000,000,000 to 700,000,000,000 of spend out of the big five companies in The US, and they say they're still token constrained.”
-
Brad Gerstner
Gerstner says he wants his personal and fund net worth levered to AI, with NVIDIA as largest public position.
“I want my personal net worth. I want my funds net worth levered against AI because all human progress is gonna be derived from machines helping humans think and augment human thinking.”
-
Dylan Patel
Patel says NVIDIA has locked up over 60% of supply chain capacity this year in long-term contracts.
“The NVIDIA has locked up over 60% of the capacity this year in long term contracts alone, and they're buying more on top of that.”
-
Dylan Patel
Patel says NVIDIA has locked up over 60% of market capacity in long-term contracts this year.
“NVIDIA has locked up over 60% of the capacity this year in long term contracts alone, and they're buying more on top of that.”
-
Dylan Patel
Patel reports NVIDIA is negotiating over $250 billion in supply contracts across components this year.
“And as you look at what they're negotiating in the market today, there's over 250,000,000,000 of wafers, of memory, of substrates, of PCBs, of networking equipment that they're going to sign this year,”
-
Dylan Patel
Patel reports NVIDIA will sign over $250 billion in supply contracts this year across components.
“as you look at what they're negotiating in the market today, there's over 250,000,000,000 of wafers, of memory, of substrates, of PCBs, of networking equipment that they're going to sign this year,”
-
Dylan Patel
Patel contrasts competitors' limited scale with NVIDIA's supply chain designed for tens of millions of chips.
“I can buy 10,000. I could buy a 100,000. I can't buy millions, tens of millions, which is what Jensen's, setting his supply chain up for.”
-
Dylan Patel
Patel argues NVIDIA is setting up supply chain to manufacture tens of millions of AI chips.
“I can't buy millions, tens of millions, which is what Jensen's, setting his supply chain up for.”
-
Bill Gurley
Gurley argues circular deals explain why Nvidia's multiple won't go higher despite strong growth.
“People wanna understand why the Nvidia multiple won't go higher. I think it's the circular deals. Like if they're not material, don't do them.”
-
Bill Gurley
Gurley questions Nvidia's contract guaranteeing to buy CoreWeave's excess capacity as abnormal business practice.
“But then they wrote a contract with CoreWeave, which said if CoreWeave ever has extra capacity, we'll buy that capacity, which presumably helps it get more debt financing, that kind of thing.”
-
Dylan Patel
Patel says AI chips consuming all TSMC 3nm and 2nm capacity is forcing companies to alternative fabs.
“More importantly is that AI is buying all the capacity on three nanometer and two nanometer in a couple years that, people are having to turn to other directions.”
-
Dylan Patel
Patel argues NVIDIA acquired Grok partly because it's manufactured on Samsung, avoiding TSMC capacity constraints.
“Part of it was because they want to have really fast inference, but like part of it is that Grok is manufactured on Samsung.”
-
Dylan Patel
Patel explains NVIDIA acquired Grok partly because it manufactures on Samsung due to no TSMC 3nm capacity available.
“Because there's no three nanometer capacity for them at TSMC, they need to chip somewhere else, and if you just if AI is as crazy as like we believe it is, and demand is as crazy as we believe it is, it's gonna be even crazier next year,”
-
Bill Gurley
Gurley cites rumors of thirty billion dollar Nvidia investment into OpenAI, calls modern venture capital sport of kings.
“And so, there's rumors today of a $30,000,000,000 investment from Nvidia into OpenAi on top of everything they've already already raised and spent.”
-
Bill Gurley
Gurley cites rumors of thirty billion dollar Nvidia investment into OpenAI, calling it sport of kings venture capital.
“there's rumors today of a $30,000,000,000 investment from Nvidia into OpenAi on top of everything they've already already raised and spent.”
-
Dylan Patel
Patel says Nvidia and AMD both lie about peak flops specs which are impossible to achieve.
“All their quoted specs are lies impossible to achieve Whether it's Nvidia or AMD, neither of them you can ever hit their peak flops.”
-
Dylan Patel
Patel reports AMD and Nvidia engineers treat Inference X as a competitive leaderboard.
“AMD and Nvidia. They love to compete with each other. And AMD engineers and Nvidia engineers look at Inference X as a leader board.”
-
Brad Gerstner
Gerstner says 80% of Altimeter's capital is in memory, logic, and compute companies like Nvidia, SK Hynix, and CoreWeave.
“80% of our capital has been in memory and logic and compute, you know, companies like Nvidia, companies like SK Hynix, companies like CoreWeave, etcetera. And so that's consumed a lot of capital.”
-
Brad Gerstner
Gerstner says 80% of Altimeter's capital is in memory, logic, and compute companies like Nvidia, SK Hynix, and CoreWeave.
“80% of our capital has been in memory and logic and compute, you know, companies like Nvidia, companies like SK Hynix, companies like CoreWeave, etcetera. And so that's consumed a lot of capital.”
-
Brad Gerstner
Gerstner notes Nvidia at $195 trades lower than six months ago at 13-14x earnings.
“At a $195, Nvidia is trading lower than it was six months ago. So it's hard to say at 13 or 14 times”
-
Brad Gerstner
Gerstner notes Nvidia at $195 trades lower than six months ago at 13-14x earnings.
“At a $195, Nvidia is trading lower than it was six months ago. So it's hard to say at 13 or 14 times”
-
Brad Gerstner
Gerstner asserts Nvidia is terribly under-owned and undervalued, and Altimeter is happy to hold it.
“I think NVIDIA is terribly, under owned. I think it's terribly, undervalued today, and so we're happy to sit in that and to own it.”
-
Brad Gerstner
Gerstner asserts Nvidia is terribly under-owned and undervalued, and Altimeter is happy to hold it.
“I think NVIDIA is terribly, under owned. I think it's terribly, undervalued today, and so we're happy to sit in that and to own it.”
-
Brad Gerstner
Gerstner mentions ChatGPT 5.5 trained on Blackwell chips and Vera Rubin rolling out later this year.
“On new Blackwell chips. We got Vera Rubin rolling out later this year. Notwithstanding all of the noise in the market,”
-
Brad Gerstner
Gerstner cites a trillion dollars in Nvidia orders over the next six to eight quarters.
“But I think if you just look at the numbers, a trillion dollars over the course of the next, you know, six to eight quarters, I think speaks for itself.”
-
Brad Gerstner
Gerstner says memory trades at 5x earnings and NVIDIA at 13-14x fully taxed GAAP earnings.
“Think about memory as an industry is only trading at five times earning. In the case of NVIDIA, trading at 13 or 14, fully taxed gap earnings.”
-
Brad Gerstner
Gerstner says memory trades at 5x earnings and NVIDIA at 13-14x fully taxed GAAP earnings.
“Think about memory as an industry is only trading at five times earning. In the case of NVIDIA, trading at 13 or 14, fully taxed gap earnings.”
-
Brad Gerstner
Gerstner predicts NVIDIA will be the first $10 trillion company and still believes this in May 2026.
“I've said before, I think NVIDIA will be the first $10,000,000,000,000 company. I you know, as I sit here, you know, in in May '26, I believe that to be true.”
-
Brad Gerstner
Gerstner predicts NVIDIA will be the first $10 trillion company and still believes this in May 2026.
“I've said before, I think NVIDIA will be the first $10,000,000,000,000 company. I you know, as I sit here, you know, in in May '26, I believe that to be true.”
-
Gavin Baker
Baker predicts orbital compute will solve power shortages in five to seven years but wafer shortages will persist longer.
“We're gonna address the watt shortages with orbital compute for sure in the next five to seven years. But the wafer shortage, I think, is going to persist for a long time.”
-
Gavin Baker
Baker says Jensen Huang visits Taiwan Semi quarterly wanting capacity doubled or tripled but they only expand 5%.
“And so they're just simply not expanding capacity as fast as Jensen wants. Jensen goes there every three months and, you know, maybe they expand, you know, 5%. He wants them to double or triple.”
-
Gavin Baker
Baker believes NVIDIA could sell $11.5 trillion worth of chips next year if TSMC expanded capacity sufficiently.
“And if they doubled or tripled capacity, like, NVIDIA could probably sell $11,520,000,000,000 worth of chips next year. I really believe that.”
-
Gavin Baker
Baker believes NVIDIA could sell $11.5 trillion worth of chips next year if TSMC expanded capacity sufficiently.
“And if they doubled or tripled capacity, like, NVIDIA could probably sell $11,520,000,000,000 worth of chips next year. I really believe that.”
-
Gavin Baker
Baker believes NVIDIA could sell $11.52 trillion in chips next year if TSMC tripled capacity but bubble risks prevent it.
“And if they doubled or tripled capacity, like, NVIDIA could probably sell $11,520,000,000,000 worth of chips next year. I really believe that. But the other side of that might be very painful for everyone.”
-
Gavin Baker
Baker believes NVIDIA could sell $11.52 trillion in chips next year if TSMC tripled capacity but bubble risks prevent it.
“And if they doubled or tripled capacity, like, NVIDIA could probably sell $11,520,000,000,000 worth of chips next year. I really believe that. But the other side of that might be very painful for everyone.”
-
Gavin Baker
Baker predicts Trainium will dominate 2026 like TPUs did in 2025, with Trainium 3 ramping in second half.
“Tranium is going to be to 2026, especially in the second half of this year when Tranium three really ramps, as TPUs were to twenty twenty five.”
-
Gavin Baker
Baker argues Trainium is most underestimated because frontier mixture-of-expert models require switched scale-up networks for inference.
“And so Trainium is for sure the most underestimated, not only because of those design choices, but because the all of these frontier models are what are called mixture of expert models.”
-
Gavin Baker
Baker states only NVIDIA and Amazon Trainium have functioning switched scale-up networks for inference today.
“And the only two functioning switched scale up networks in the world today are the ones that power NVIDIA GPUs and Amazon's Trainiums.”
-
Gavin Baker
Baker states only NVIDIA and Amazon Trainium have functioning switched scale-up networks for inference today.
“And the only two functioning switched scale up networks in the world today are the ones that power NVIDIA GPUs and Amazon's Trainiums.”
-
“They've announced a trillion dollars over the course of the next six to eight quarters of demand on on, you know, Blackwell's and now Vera Rubins.”
-
Gavin Baker
Baker reveals Jensen Huang has never had a contract with TSMC, operating entirely on handshakes and fairness.
“I do think it's wild that Jensen has never had a contract with Taiwan Semi. They do business on what seems fair in handshakes.”
-
Gavin Baker
Baker argues NVIDIA could sell $2-3 trillion in GPUs if TSMC expanded capacity, but TSMC's restraint prevents a bubble.
“If Taiwan Semi did what Jensen wanted, I think Nvidia could sell $2,000,000,000,000 of GPUs in '26 or '27, maybe 2,500,000,000,000, maybe 3,000,000,000,000. But there is a limit where consumers would consume so much.”
-
Gavin Baker
Baker argues NVIDIA could sell $2-3 trillion in GPUs if TSMC expanded capacity, but TSMC's restraint prevents a bubble.
“If Taiwan Semi did what Jensen wanted, I think Nvidia could sell $2,000,000,000,000 of GPUs in '26 or '27, maybe 2,500,000,000,000, maybe 3,000,000,000,000. But there is a limit where consumers would consume so much.”
-
Gavin Baker
Baker argues Taiwan Semi's capacity constraint prevents AI bubble; NVIDIA could sell $2-3T GPUs otherwise.
“If Taiwan Semi did what Jensen wanted, I think Nvidia could sell $2,000,000,000,000 of GPUs in '26 or '27, maybe 2,500,000,000,000, maybe 3,000,000,000,000.”
-
Gavin Baker
Baker argues Taiwan Semi's capacity constraint prevents AI bubble; NVIDIA could sell $2-3T GPUs otherwise.
“If Taiwan Semi did what Jensen wanted, I think Nvidia could sell $2,000,000,000,000 of GPUs in '26 or '27, maybe 2,500,000,000,000, maybe 3,000,000,000,000.”
-
Gavin Baker
Baker predicts GPUs will have 10-15 year useful lives due to prefill-inference disaggregation, extending older chips' value.
“The disaggregation of inference means that I think these GPUs are going to have ten or fifteen year lives.”
-
Gavin Baker
Baker argues GPU useful lives will extend to 10-15 years due to inference disaggregation, contradicting AI skeptics.
“The disaggregation of inference means that I think these GPUs are going to have ten or fifteen year lives. The AI skeptics are like, oh, these companies are all cooking their books.”
-
Gavin Baker
Baker argues GPU useful lives will extend to 10-15 years due to prefill/decode disaggregation, not 1-2 years.
“The useful life of GPU is only a year or two. The useful life of CPU is only four years because the rapid technological change.”
-
Brad Gerstner
Gerstner explains Invest America funds will be invested in the top 500 US companies, making children direct shareholders.
“All of the money that goes into these accounts will be invested in the best 500 companies in America. They will be direct shareholders, a little bit of Nvidia,”
-
Brad Gerstner
Gerstner explains Invest America funds will be invested in the top 500 US companies, making children direct shareholders.
“All of the money that goes into these accounts will be invested in the best 500 companies in America. They will be direct shareholders, a little bit of Nvidia,”
-
Dylan Patel
Patel explains NVLink connects only 72 GPUs while Google ICI connects 8,000 chips without switches, creating architectural trade-offs.
“NVIDIA, the NVLink can only connect 72 GPUs. For Google, their ICI can connect 8,000 chips at super high bandwidth, but you have to pass through other chips to get there because there's no switch.”
-
Gavin Baker
Baker says Jensen Huang was obviously exceptional when they first met in early 2000s.
“I was a pretty young man. It was obvious to me he was exceptional, like for sure one of the most exceptional people I'd ever met.”
-
Gavin Baker
Baker identifies Jensen as one of the two or three most exceptional people he's ever met in his life.
“But like now, for sure, Jensen is, you know, one of the two or three most exceptional people I've ever met in my life.”
-
Dylan Patel
Patel says SemiAnalysis initially doubted Jensen's 25x Blackwell claim, predicting only 15-20x improvement.
“Jensen, when he originally launched Blackwell, had claimed it would be a 25x improvement. And at the time, no one believed him. Right? It's Jensen, right?”
-
Dylan Patel
Patel emailed Jensen documenting skepticism about Blackwell's 25x claim from industry observers.
“I emailed him. I like, hey, Jensen. You know, back in 2024, you said or back back when you launched Blackwell, said '24 twenty five x.”
-
Dylan Patel
Patel predicts Nvidia Vera Rubin will have smoother deployment than Blackwell due to GB's new architecture problems.
“It's it seems like it'll be a much smoother ramp than GB. You know, GB had a lot of problems because it was brand new,”
-
Dylan Patel
Patel argues all successful non-NVIDIA AI hardware adoption has come from labs like Anthropic and OpenAI, not enterprises.
“all of the successful AI hardware adoption of non NVIDIA flavors has been Anthropic adopting TPUs. Obviously, Google doing their own with their TPUs. Or Anthropic and OpenAI now adopting Trainium, or OpenAI adopting Cerebras.”
-
Gavin Baker
Baker says no one predicted old GPU prices would go vertical, everyone expected gradual declines.
“I don't think anyone in '24 or '25 thought that the prices of old GPUs would be going vertical. Everybody thought, hey. We're gonna be smart. We're gonna sign these long term contracts.”
-
Gavin Baker
Baker reports GPU rental prices doubled from mid-$2 to nearly $4 per hour over seven months for identical clusters.
“And they had rented a cluster of several thousand black wells, and we'll just call it somewhere in the mid $2 per GPU hour.”
-
Gavin Baker
Baker reports inference cloud expects to pay 100% more for Blackwells when contracts expire, showing hyperscalers are underearning.
“They went on a podcast, and they essentially said, we are planning to pay 100% more for Blackwell's when our contract expires. And that just means that essentially all the hyperscalers are under earning.”
-
Gavin Baker
Baker notes NVIDIA trades at its lowest forward PE in ten years.
“NVIDIA is actually, as we record this, at its lowest forward PE of the last ten years.”
-
Gavin Baker
Baker notes NVIDIA trades at its lowest forward PE in ten years.
“NVIDIA is actually, as we record this, at its lowest forward PE of the last ten years.”
-
Gavin Baker
Baker identifies four companies that matter at scale in long-term supply agreements for AI chips.
“Let's just think about the game theory of breaking an LTA. So there's four companies that matter at scale. There's Amazon with their tradiums.”
-
Gavin Baker
Baker argues only four companies matter at scale for AI chips: Amazon, Google, AMD, and NVIDIA.
“So there's four companies that matter at scale. There's Amazon with their tradiums. There's Google with their TPUs. There's AMD, and then there's NVIDIA who's, like, much bigger than everybody else combined.”
-
Gavin Baker
Baker argues only four companies matter at scale for AI chips: Amazon, Google, AMD, and NVIDIA.
“So there's four companies that matter at scale. There's Amazon with their tradiums. There's Google with their TPUs. There's AMD, and then there's NVIDIA who's, like, much bigger than everybody else combined.”
-
Gavin Baker
Baker suggests memory companies should copy NVIDIA's credit wrapper model to participate in GPU revenue streams as credit markets revolt.
“Which is? I would be going to the buyers of GPUs, tradiums, and whoever and say, I'll participate in the NVIDIA credit wrapper.”
-
Gavin Baker
Baker suggests memory companies should copy NVIDIA's credit wrapper model to participate in GPU revenue streams as credit markets revolt.
“Which is? I would be going to the buyers of GPUs, tradiums, and whoever and say, I'll participate in the NVIDIA credit wrapper.”
-
Gavin Baker
Baker suggests memory companies should copy NVIDIA's credit wrapper model while they have cash and credit markets are tight.
“I'm just making this up. But, like, do something. Like, you can because you have money now and credit markets are revolting.”
-
Gavin Baker
Baker suggests memory companies should copy NVIDIA's credit wrapper model despite business instability risks.
“I would be going to the buyers of GPUs, Traniums, and whoever and say, I'll participate in the NVIDIA credit wrapper. Now their business is just inherently less stable and predictable,”
-
Gavin Baker
Baker suggests memory companies should copy NVIDIA's credit wrapper model despite business instability risks.
“I would be going to the buyers of GPUs, Traniums, and whoever and say, I'll participate in the NVIDIA credit wrapper. Now their business is just inherently less stable and predictable,”
-
Where they overlapsame programme, same subject
- Bill Gurley and Brad Gerstner — BG2 Pod, 6 appearances together
- Bill Gurley and Dylan Patel — BG2 Pod, 1 appearance together
- Brad Gerstner and Dylan Patel — BG2 Pod, 1 appearance together
- CNBC Television also carries Brad Gerstner, David Tepper and Dylan Patel on this subject, at different times.
- All-In Podcast also carries Brad Gerstner and Gavin Baker on this subject, at different times.
- Bloomberg Live also carries Brad Gerstner and Stanley Druckenmiller on this subject, at different times.