NVIDIA Corporation

Sacks support ×2 ▲ on 🤖 AI compute capex supercycle has 12-24 months of runway E288 · 2026-09-04 ▶ 47:23

the reason why it's not going to be stopped is because consumers are adopting it ... Why are the data centers getting built? Why do people want to build them? It's because the demand is there from consumers and businesses. ... the adoption is pulling this whole thing forward

Jason support ×3 ▲ on 🤖 Nvidia + Hugging Face makes NVDA the open-source AI champion E288 · 2026-09-04 ▶ 45:22

I believe right now, starting today, NVIDIA is the number one competitor for OpenAI and Anthropic, not just for tokens, but for enterprise compute. And I think now Jensen is the leading open source provider in America. ... He's going to catch up and blow past China and the Chinese models, I think, by the end of 2027

Chamath support ×3 ▲ on 🤖 Nvidia + Hugging Face makes NVDA the open-source AI champion E288 · 2026-09-04 ▶ 44:06

This Hugging Face thing will go down as one of the most important transactions in AI. Because you are creating now in the largest competitor and the largest most well-capitalized company a bulwark against all of this closed-source oligopoly insanity.

Chamath support ×2 ▲ on 🤖 AI compute capex supercycle has 12-24 months of runway E287 · 2026-08-29 ▶ 41:02

we are supposed to be in the middle of an enormous financial buildout to support AI... thank God, we have companies like Nvidia, and Google, and Microsoft, and Meta, and Amazon who take on that burden... there's so much chirping, by the way, on the internet about the balance sheet of Nvidia, and blah, blah, blah. And I think people completely missed that these guys are putting the entire US economy on their back.

Sacks support ×3 ▲ on 🤖 AI compute capex supercycle has 12-24 months of runway E287 · 2026-08-29 ▶ 29:34

So, what is the takeaway, I think, for the larger market here? It is that the AI boom is continuing. The other narrative that's getting shredded today is that this AI CapEx is a bubble. It's basically gonna end very soon. And what Nvidia is basically saying with its numbers and forecasts is actually, this AI CapEx is gonna continue well into the future. It's got real legs. And the amazing thing is that after these numbers, even despite the 8% bump, Nvidia is only trading at 12 times earnings.

Chamath oppose ×2 ▲ on 🤖 AI compute capex supercycle has 12-24 months of runway E286 · 2026-08-21 ▶ 5:29

That causes people to be less likely to invest in all the data centers. Data centers themselves are being shut down. Where does that leave frontier model companies? They're the most at risk. ... we are now in a very precarious situation that we could have frankly avoided.

Jason support ×2 ▲ on 🤖 AI compute capex supercycle has 12-24 months of runway E285 · 2026-08-14 ▶ 1:13:10

There is no world in which I don't see corporations spending five or 10 percent of the salaries of their employees on the equivalent in tokens... you are looking at $8 trillion in AI spend just in the United States. It is obviously going to happen. And if I am wrong and it is half... that is kind of where these companies are trending.

Sacks support ×2 ▲ on 🤖 AI compute capex supercycle has 12-24 months of runway E285 · 2026-08-14 ▶ 1:03:24

The biggest risk is that you get a glut of compute and you get an overbuild. ... In a weird way, all the political headwinds, ... will almost guarantee that there's not an oversupply relative to the exponentially growing demand.

Gavin Baker support ×3 ▲ on 🤖 AI compute capex supercycle has 12-24 months of runway E285 · 2026-08-14 ▶ 1:11:08

The overwhelming majority of tokens are profitable for everyone in the chain, everyone. Anthropic is generating cash... When I hear these macro and value investors making an assumption that tokens are subsidized and this is all going to collapse in some circular financing bonfire, they're just wrong.

Brad Gerstner support ×2 ▲ on 🤖 AI compute capex supercycle has 12-24 months of runway E284 · 2026-08-08 ▶ 43:40

The demand exists in the world today. I think it will exist in the world for, well, the next 12 to 24 months, ... I suspect this cannot keep up at this pace more than another two years or so. ... Famous last words, I don't see it today over the course of the next 12 to 18 months.

Friedberg support ×2 ▲ on 🤖 AI compute capex supercycle has 12-24 months of runway E284 · 2026-08-08 ▶ 3:42

Capex is high alpha, low beta in data center infrastructure, that capital. And model development theoretically could be high alpha, but it's very high beta. It's a very risky way to deploy capital.

Sacks support ×2 ▲ on 🤖 AI compute capex supercycle has 12-24 months of runway E283 · 2026-07-31 ▶ 6:05

Is this correction driven by fundamentals, or is it driven by momentum? My view is that it's driven by momentum... the key question is the capex being invested in the AI boom — is that real or is it misguided?... My view is that it's real. I think there will be a return on all of this capex... this is temporary market volatility amplified by leverage.

Chamath support ×2 ▲ on 🤖 AI compute capex supercycle has 12-24 months of runway E281 · 2026-07-18 ▶ 1:07:09

The most incredible thing is how extreme the price is at the front end of the curve, when you have verifiable, energizable power today. ... about 40 percent of all these projects are getting mothballed and stopped, and so it's creating this massive deficit of available energy to ... the extent that you actually want drug discovery, or ... we may actually not be able to service it

Chamath oppose ×2 ▲ on 🤖 AI compute capex supercycle has 12-24 months of runway E280 · 2026-07-11 ▶ 5:16

I sat down with my CTO today... right now, our token costs are doubling every 45 days... [and the downstream productivity?] maybe 5% max... you need to use a lot more tokens to get to this next iteration of improvement because we've effectively already asymptoted... everybody in the next three or four years will for sure go through [this reckoning]... [and per Fable's own numbers] the actual ROI [of AI on S&P-493 EPS] was somewhere between 1 and 2%.

Chamath support ×2 ▲ on 🤖 AI compute capex supercycle has 12-24 months of runway E276 · 2026-06-13 ▶ 17:14

Two years ago, I bought 2000 acres in Arizona with a partner. ... allowed to build a two gigawatt data center. ... I just put in an offer for another ... when I started this project, it was like four or five billion and it's increased by

Chamath oppose ×2 ▲ on 🤖 AI compute capex supercycle has 12-24 months of runway E275 · 2026-05-29 ▶ 51:56

There was an economic and capital mode to training that is going away. ... we're getting these domain specific architectures at the silicon layer and then second, we're rebuilding all of the core components. ... the entire training complex in C and it's an order of magnitude increase ... why would we stick to the $10 billion training runs when we can have the $10 million training runs?

Gavin Baker support ×2 ▲ on 🤖 AI compute capex supercycle has 12-24 months of runway E274 · 2026-05-22 ▶ 4:34

if OpenAI and Anthropic call it $100 billion of ARR now, with 80%-ish gross margins on inference, the returns are there. ... if we add in Gemini, we add in Cursor, we add in XAI, we add in Open Source, it's not hard to see $200, $300, $400 billion of ARR at the end of this year at high margins. ... there's going to be a really strong ROI this year, even excluding

Chamath oppose ×2 ▼ on 🤖 Nvidia risk: special-purpose silicon + Huawei E274 · 2026-05-22 ▶ 1:19:00

at the end of this Grok transaction last year, my ... we're going to move to these domain-specific architectures. I thought that was like a fait accompli, ... But the reality is that that DSA market evolution is actually happening inside of Nvidia. That's what's so insane to me. ... These guys actually have domain-specific architectures because they're doing these design programs with every,

Gavin Baker oppose ×3 ▼ on 🤖 Nvidia risk: special-purpose silicon + Huawei E274 · 2026-05-22 ▶ 1:14:22

there's a narrative that Nvidia is losing share to the TPU, ... within the Western AI world, within data centers that are being built, ... Nvidia's AI business is growing faster than Broadcom's, and faster than a lot of other companies that are seen as part of this ASIC sharegate story. ... these other ASICs are not being submitted for benchmarks. ... I think the reason they're not being submitted is they will lose.

Sacks support ×2 ▲ on 🤖 AI compute capex supercycle has 12-24 months of runway E271 · 2026-05-01 ▶ 46:55

The bull thesis for AI just got validated in a single afternoon... $725 billion in CapEx guidance in 2026 from four companies... more than 2 percent of GDP... AI was 75 percent of GDP growth [last quarter]... AI is now synonymous with the growth of the American economy.

In just two years, computation went up by a factor 10,000x... We haven't even started scaling yet. We are absolutely at a million x... If that $500,000 engineer did not consume at least $250,000 with the tokens, I am going to be deeply alarmed.

the $50 billion factory will generate for you the lowest cost tokens. ... even when the chips are free, it's not cheap enough. ... is coming to Nvidia, and the growth of open models is incredible. And that's all on Nvidia.

Sacks support ×1 ▲ on 🤖 AI compute capex supercycle has 12-24 months of runway E264 · 2026-03-13 ▶ 39:04

there's a tremendous latent demand for the ability to generate code in large quantities, ... To be able to buy code on a metered basis as the cost per token keeps going down, it's kind of an amazing deal.

Chamath support ×3 ▲ on 🤖 AI compute capex supercycle has 12-24 months of runway E264 · 2026-03-13 ▶ 44:54

I am building a one gigawatt data center in Arizona. ... When I greenlit that project, I thought it was going to be a four or $5 billion investment. ... now it's upwards of $50 billion for the powered shell, ... it's about a five to six-year payback just to get into the money. And then it's about 10 billion a year.

Chamath support ×2 ▲ on 🤖 AI compute capex supercycle has 12-24 months of runway E263 · 2026-03-06 ▶ 55:17

I have a small software company called 8090... our costs have more than tripled since November of 25. Between the inference cost that we pay AWS, which is ginormous, between our cost with Cursor, between Anthropic, we are just spending millions.