Google's AI infrastructure boss Amin Vahdat has new goal for employees: We must…

Google's AI infrastructure boss, Amin Vahdat, has informed employees that the company needs to double its compute capacity every six months to keep pace with the demand for artificial intelligence (AI) services. According to a report by CNBC, Vahdat, who is also a VP at Google Cloud, made the statement during a recent all-hands meeting in a presentation titled “AI Infrastructure.” One of the slides in the presentation on “AI compute demand” said: “Now we must double every 6 months.... the next 1000x in 4-5 years.” As per the report, Vahdat also said, “The competition in AI infrastructure is the most critical and also the most expensive part of the AI race.”

Google-parent Alphabet CEO Sundar Pichai and CFO Anat Ashkenazi were also present at the meeting. The presentation followed Alphabet's better-than-expected third-quarter results, which led the company to raise its capital expenditures forecast for the second time this year, to a range of $91 billion to $93 billion, with a “significant increase” expected in 2026. Competitors such as Microsoft, Amazon, and Meta have also increased their capital expenditure forecasts, with the four companies collectively anticipating spending over $380 billion this year.

What Google exec Amin Vahdat said about the company’s AI investments

At the meeting, Vahdat reminded employees that Google’s “job is of course to build this infrastructure, but it’s not to outspend the competition, necessarily.”


“We’re going to spend a lot,” he added, explaining that the aim is to offer infrastructure that is “more reliable, more performant and more scalable than what’s available anywhere else.”

Along with expanding physical infrastructure, Vahdat said Google will increase capacity by developing more efficient models and using its custom silicon.

Last week, Google announced the public launch of its seventh-generation Tensor Processing Unit, Ironwood, which the company says is nearly 30 times more power efficient than its first Cloud TPU from 2018.

Vahdat also noted that Google benefits from DeepMind’s research, which helps the company understand how AI models may evolve in the coming years.

Google needs to “be able to deliver 1,000 times more capability, compute, storage networking for essentially the same cost and increasingly, the same power, the same energy level,” Vahdat noted.

“It won’t be easy but through collaboration and co-design, we’re going to get there,” Vahdat added.

Earlier, Google CEO Sundar Pichai told employees that 2026 will be “intense” due to AI competition and pressure to meet demand for cloud and compute resources.

At the meeting, Pichai also addressed a question about a possible AI bubble, an idea gaining attention as investors question whether the expected levels of AI spending are sustainable.

One of the employees asked Pichai, “Amid significant AI investments and market talk of a potential AI bubble burst, how are we thinking about ensuring long-term sustainability and profitability if the AI market doesn’t mature as expected?”

Responding to the question, Pichai said he understands the concern. “It’s a great question. It’s definitely been in the zeitgeist, people are talking about it,” he said.

He repeated a point he had made before about the risks of investing too slowly. He pointed to Google’s cloud business, which recently reported 34% annual revenue growth to more than $15 billion for the quarter, with a $155 billion backlog.

“I think it’s always difficult during these moments because the risk of underinvesting is pretty high. I actually think for how extraordinary the cloud numbers were, those numbers would have been much better if we had more compute,” Pichai noted.

He added that Google takes a disciplined approach, supported by strong core businesses and the company’s balance sheet.

“We are better positioned to withstand, you know, misses, than other companies,” Pichai added.