There's still talk of an artificial intelligence bubble fills the air these days, fearing overinvestment that could pop At any moment, a contradiction is brewing on the ground: companies like Google and OpenAI can hardly build infrastructure fast enough to meet their artificial intelligence needs.
During an all-hands meeting earlier this month, Google's head of artificial intelligence infrastructure, Amin Vahdat, told employees that the company must double its capacity every six months to meet demand for artificial intelligence services. reports CNBC. Vahdat, vice president of Google Cloud, presented slides showing the company needs to scale “the next thousand times in 4-5 years.”
While a thousandfold increase in computing power sounds ambitious on its own, Vahdat noted some key limitations: Google must be able to deliver this increase in performance, computing resources and storage networks “at essentially the same cost and increasingly with the same power and the same energy level,” he told employees during the meeting. “It won’t be easy, but through collaboration and co-design, we will get there.”
It's unclear to what extent this “demand” Google mentioned represents organic user interest in AI capabilities versus how the company integrates AI features into existing services like Search, Gmail and Workspace. But whether users use these features voluntarily or not, Google isn't the only tech company struggling to keep up with a growing user base of customers using artificial intelligence services.
The biggest tech companies are racing to build data centers. Google Competitor OpenAI planning to build six massive data centers across the US through its Stargate partnership with SoftBank and Oracle, committing more than $400 billion over the next three years to reach nearly 7 gigawatts of capacity. The company faces similar restrictions serving its 800 million weekly users of ChatGPT, with even paid subscribers regularly exceeding limits on features such as video synthesis and simulated reasoning models.
“Competition in AI infrastructure is the most important and also the most expensive part of the AI race,” Vahdat said at the meeting, according to a CNBC viewing of the presentation. The infrastructure chief explained that Google's mission goes beyond simply outperforming competitors. “We're going to spend a lot,” he said, but noted that the real goal is to create infrastructure that is “more reliable, more performant and more scalable than what's available elsewhere.”
