Gerstner predicts next year will see more compute deployed globally than in all of human history combined.
“We're going to deploy more compute on a global basis next year than all the compute ever deployed by humanity combined. That's how much compute is going to drive these training models.”
Gerstner says MANG (Microsoft, Amazon, Nvidia, Google) went from zero to $25 billion in VC investing in six years.
“He called them MANG. Microsoft, Amazon, know, Nvidia and Google. And so as you can see on this chart, we go from almost no venture capital investing, you know, out of these folks six or seven years ago to all of a sudden this $25,000,000,000, last year in VC investing.”
Gerstner reports Super Micro raised full-year guidance 40% to $14 billion.
“Super Micro took their full year guide up after the bell today by 40%. 40% to $14,000,000,000.”
Gerstner highlights NVIDIA data center grew over 25% versus consensus forecast of minus 6%.
“Data center revenues were expected to shrink by 6%. Right? We thought we we had huge variant perception by being up 16%, and it ended up over 25%.”
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.”
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.”
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.”
Gerstner says only a few companies can afford 100,000 GPU clusters that Microsoft, Elon, and Meta are building.
“I think the latest discussions, whether it's Stargate out of Microsoft, a 100,000 GPU cluster, Elon's talking about a 100,000 GPU cluster, Mark is talking about that.”
Gerstner identifies four ingredients to compete in AI: capital at $40B/year, data scaling, compute infrastructure, and distribution.
“I think there are four important ingredients to compete in this market. Number one, you have to have capital, and the leaders are spending $40,000,000,000 a year.”
Gerstner says Meta grew net income from $22B to $55B while cutting headcount from 85,000 to 69,000 in two years.
“In two years, that company has gone from $22,000,000,000 in net income to $55,000,000,000 in net income. They've reduced their headcount from 85,000 people to 69,000 people.”
Gerstner says hyperscalers' total capex is now at $200 billion, raising questions about return on investment.
“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 and saying, when are you gonna get a return on the dollars that you're spending?”
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”
Gerstner argues AI wave differs from internet era because hyperscalers are deploying massive capex directly, not customers.
“because all AI tools are being delivered as a service in the cloud, there's just a massive amount of CapEx happening.”
Gerstner notes Satya called AI a supply-driven wave with much experimental work, creating reset risk.
“Satya very quickly said this is a supply driven wave. And I think up to date, that is true. We're we're building ahead of demand.”
Gerstner argues this would be the first major technology phase shift with capex and revenues perfectly aligned.
“the idea that we're going to go through this phase shift with CapEx and revenues perfectly aligned and perfectly matched, right?”
Gerstner says big four tech companies will spend $220 billion on capex this year.
“if you look at the big four, they're gonna spend $220,000,000,000 this year on CapEx.”
Gerstner reports Masa Son said $9 trillion cumulative capex on 200 million GPUs is reasonable, possibly too small.
“Masa at FII today said 9,000,000,000,000 of cumulative capex on 200,000,000 GPUs is very reasonable. He said, in fact, I think it may be too small.”
Gerstner notes Google Cloud grew 35% year-over-year, adding $2.5 billion ARR in one quarter, equivalent to Databricks.
“So just to put that in perspective, they added about 2,500,000,000 of new ARR in the quarter. That's like adding a Databricks in the quarter.”
Gerstner cites CoreWeave's $12 billion OpenAI Stargate deal announced yesterday as evidence AI buildout continues.
“CoreWeave announced yesterday a $12,000,000,000 deal with OpenAI on Stargate. Those things are getting constructive. We're investors there. I I can tell you those things are getting built.”
Gerstner cites Jensen Huang estimating $1.5 trillion in AI compute buildout discussed in Paris.
“Jensen Huang said over in Paris this week, he estimates a trillion and a half of AI compute build out”
Gerstner says Anthropic raising $5B at $170B valuation on $5B revenue, unprecedented venture scale.
“I think that Iconic is going to lead a $5,000,000,000 round into Anthropic at $170,000,000,000 In the case of Anthropic, it's, you know, 170,000,000,000 on rumored $5,000,000,000 in revenue.”
Gerstner reports X.AI targets 50 million H100 equivalent units, requiring 4 million GPUs and 11 gigawatts.
“If you look at this one out of Elon talking about the X dot AI goal is 50,000,000 in units of H100 equivalent.”
Gerstner suggests Apple should acquire AI capabilities given their $2 trillion market cap and cash position.
“What's your opinion on the build it versus buy it debate around Apple? I feel like for $2,025,000,000,000, they should just buy it, just given how much cash they have”
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.”
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.”
Gerstner quotes AMD's Lisa Su calling this year two of a ten-year compute buildout supercycle.
“Lisa said we're in year two of a ten year super cycle to build out the compute in America. If you're not on the AI wave, then you're not gonna participate in all that upside.”
Gerstner compares AI buildout to Manhattan Project, calling it both national and economic security imperative against China.
“We are in a global AI race with China. This is like the Manhattan Project. We have to build out compute. It it's a it's not only a national security imperative.”
Gerstner argues that $100-200 billion in revenue by 2028-29 makes the capex spend responsible.
“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 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.”
Gerstner says $250B NVIDIA spend implies $500B total capex requiring $1 trillion in AI revenue to justify.
“Assuming that is just 50% of the CapEx, you're spending about 500,000,000,000 on CapEx, and then you've gotta earn about a trillion dollars of AI revenue for all of this CapEx to be worth it.”
Gerstner argues AI must generate $1 trillion revenue to justify current CapEx versus $400 billion for all software.
“then you've gotta earn about a trillion dollars of AI revenue for all of this CapEx to be worth it.”
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.”
“In 2023, these companies spent $150,000,000,000 on capex building out data centers. This year, over 500,000,000,000. Though, that's not speculative. That is purchase orders. Those are buildings. That's power.”
Gerstner says capex jumped from $150 billion in 2023 to over $500 billion this year in committed spending.
“In 2023, these companies spent $150,000,000,000 on capex building out data centers. This year, over 500,000,000,000. Though, that's not speculative. That is purchase orders.”
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.”
Gerstner says AI companies will spend over $500 billion on capex this year, up from $150 billion in 2023.
“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. That's power.”
Gerstner says the major tech companies will spend over $500 billion on AI capex this year.
“The spending, the CapEx from that group alone is gonna be over $500,000,000,000 this year.”
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.”
Gerstner notes the $700 billion AI buildout is entirely funded by profitable private US companies.
“That is the size of the interstate highway system, and it's all funded by profitable private companies in America.”
Gerstner compares the next three years of AI investment to building the interstate highway system, all privately funded.
“I think for the next three years, this is like building the interstate highway system. Right? This is significant. It's important to The United States Of America, by the way.”
Gerstner says the Mag Five will spend $800 billion in total capex this year.
“This year, we're gonna have 800,000,000,000 in total CapEx by the Mag five. And the question was, would there be revenues to justify that?”
Gerstner warns that 30-40% of compute is delayed this year, which could slow the entire AI trade if delays persist.
“And you heard the rumors and seen the headlines that 30 or 40% of compute now is delayed this year. So keep your eye on power and compute.”
Gerstner says the US is financing nearly a trillion dollars in annual AI capex privately, unlike China's government subsidies.
“I think it's extraordinary that we're financing almost a trillion dollars of capex a year to build out the the entire AI infrastructure without any money from the US government.”
“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.”
Gerstner claims AI investment is contributing over 100 basis points to US GDP growth this year.
“This is the single most important contributor to the GDP growth of The United States Of America this year. It's contributing over a 100 basis points of our GDP growth.”
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”