Gerstner predicts 2024 will see more compute deployed than all previous years combined for AI.
“We're gonna launch more. We're gonna deploy more compute in 2024 than in all previous years combined.”
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.”
Gerstner identifies hyperscaler total CapEx at $200 billion as the real concern drawing criticism and return questions.
“In fact, I think if people are gonna be critical of anything, a lot of people are looking at the total CapEx of the hyperscalers now at $200,000,000,000”
Gerstner argues Amazon spent over $100 billion on AWS for eight years before profitability, paralleling AI investment.
“Amazon spent over a $100,000,000,000 on AWS, invested for over eight years before they saw profitability in that business.”
Gerstner says Amazon spent over $100 billion on AWS for over eight years before seeing profitability.
“just Amazon spent over a $100,000,000,000 on AWS, invested for over eight years before they saw profitability in that business.”
Gerstner claims 30% annual infrastructure expense growth requires comparable inference revenue growth on both sides.
“So, you know, if you think that infrastructure expenses are going to grow at 30% a year, then I think you have to believe that the underlying inference revenues, right, both on the consumer side and the enterprise side are gonna grow somewhere in that range as well.”
Gerstner argues GPU capex sustainability depends on hyperscalers generating commensurate inference revenues, citing Satya's disciplined approach.
“It all comes down ultimately to the revenues that are generated by the people who are making the purchases of the GPUs. Right?”
Gerstner posits 30% annual infrastructure expense growth requires matching inference revenue growth from enterprises and consumers.
“if you think that infrastructure expenses are going to grow at 30% a year, then I think you have to believe that the underlying inference revenues, right, both on the consumer side and the enterprise side are gonna grow somewhere in that range as well.”
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.”
Gerstner argues capex growth depends on enterprise adoption rates and consumer willingness to pay, not just model improvements.
“It's the rate of adoption of consumers and what consumers are willing to pay to use ChatGPT or to use Clawd or to use these other services.”
Gerstner says TSMC pledged $100 billion more, contributing to $1.5 trillion in total U.S. reinvestment commitments.
“I mean, the list of companies, I think we're up to a trillion, a trillion and $0.5 of reinvestment.”
Gerstner says NVIDIA raised total compute buildout estimate to $3-4 trillion from $2 trillion.
“Last night on the call, they said 3 to 4,000,000,000,000 of total compute. He told us that inference was gonna billion x. Right?”
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.”
Gerstner calculates OpenAI needs at least $150 billion revenue in 2030 to justify announced CapEx commitments.
“And so, like, the question is, Bill, how much revenue do they need in 2030 to justify $150,000,000,000 in capex? Well, I think you would need at least $150,000,000,000 of revenue.”
“Tech companies are investing over the course of the next four or five years, it's about 10 times the size of the Manhattan Project on an inflation adjusted or GDP adjusted basis.”
Gerstner contrasts OpenAI's $13 billion revenue with its $1.4 trillion compute commitment over four to five years.
“OpenAI's revenues are still a reported $13,000,000,000 in 2025. And Sam, on your live stream this week, you talked about this massive commitment to compute, right? 1,400,000,000,000 over the next four or five years”
Gerstner argues $100-200B annual capex is responsible if OpenAI hits $100-200B revenue in 2028-29.
“Now if they're doing a 100 to 200,000,000,000 in revenue in 2829, then now we're in the zone of responsible.”
Gerstner breaks down OpenAI's $1.4 trillion capex: half borne by partners, remainder spread over five years.
“And if they're doing that and you spread the CapEx out, so we talked about 1,400,000,000,000, half of that's gonna be borne by their partners.”
Gerstner calculates partners will bear half of $1.4 trillion capex, leaving $600-700 billion spread over five years.
“And if they're doing that and you spread the capex out, so we talk about 1,400,000,000,000, half of that's gonna be borne by their partners.”
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.”
Gerstner says frontier model revenues are needed to fund $1-2T annual capex buildout of American AI
“Narrative violation. Have been hearing for 18 months frontier models dead. Competition w open source is not zero sum. If you want American AI to win u should cheer for open & closed - we need these revenues to pay for the $1-2 T annual capex build out of American AI. 🇺🇸🚀💰 https://t.co/5swFA78sel”
Gerstner characterizes data center infrastructure capex as high-alpha, low-beta, while model development is high-alpha but very high-beta and risky.
“I would frame it as capex is high alpha, low beta in data center infrastructure... model development theoretically could be high alpha but it's very high beta — it's a very risky way to deploy capital.”