Tech companies are rushing to build out their infrastructure as their increasingly resource-intensive AI products overwhelm capacity, drain chipmakers’ supplies, and require more power. Google, once dubbed “”king of the web,” is one of those companies, and a high-level executive at The Big G has told employees that the company needs to rapidly expand its service capabilities if it wants to meet demand for its AI services.
CNBC recently caught a presentation given by Amin Vahdat, vice president of machine learning, systems and cloud AI at Google. The presentation included a slide on “AI compute demand” that claims Google should “double every 6 months… 1000x over the next 4-5 years.”
“Competition in AI infrastructure is the most important and also the most expensive part of the AI race,” Vahdat reportedly said at the all-hands meeting where the presentation took place. “Google’s job is certainly to build this infrastructure, but not necessarily to outcompete the competition,” he said. “We’re going to spend a lot more,” he said, in an effort to build AI infrastructure that is “more reliable, more performant, and more scalable than what’s available anywhere else.”
Since the CNBC story was published, Google has objected to the reporting. While CNBC originally quoted Vahdat as saying the company would need to “double” its compute capacity every six months, a Google spokesperson told Gizmodo that the executive’s words were taken out of context. The spokesperson further said that “Wasn’t talking about capital formation close to the magnitude suggested. In fact, he simply said that demand for AI services means we are being asked to provide significantly more computing capacity, which we are driving through efficiencies in hardware, software and model optimization in addition to new investments.”
CNBC has since updated its reporting from a “counting” to a “serving” capacity. Service capacity would refer to Google’s ability to handle a rising tide of user requests, while compute capacity would refer to the company’s overall infrastructure dedicated to AI, including what’s needed to train new models and other expenses. When asked for further clarification about the difference between the two, the spokesperson said that the original headline “read as if it was saying that we’re doubling the amount of compute we have – measured either by the number of chips we run or the amount of megawatts of power.” Instead, “the increase in capacity described by Amin will be achieved in a number of ways, including new and more capable chips and model efficiency and optimization,” he said.
Whatever is going on secretly, it appears that Google—like its competitors—needs to scale up its operations to support its emerging AI infrastructure business. Vahdat’s comments come shortly after the tech giant reported some huge profits from its cloud business, with the company announcing that it plans to increase spending in the coming year.
During his presentation, Vahdat also reportedly claimed that Google needs to be able to provide “1,000 times more capacity, compute, storage, networking”. [than its competitors] Basically for the same cost and more, the same power, the same energy levels.” He acknowledged that it “will not be easy” but said that “through collaboration and co-design, we will get there.”
The race to build data centers—or “AI infrastructure” as the tech industry calls it—is becoming insane. The likes of Google, Microsoft, Amazon, and Meta all claim they’re going to increase their capital spending in an effort to build the future of computing (cumulatively, Big Tech is expected to spend at least $400 billion over the next twelve months). As these facilities continue to grow, they are creating all kinds of drama in the communities where they live. Environmental and economic concerns abound. Some communities have begun to oppose data center projects – and, in some cases, they are successfully knocking them out. Still, given the enormous amounts invested in this industry, it will be an ongoing battle for Americans who don’t want an AI colossus in their backyard.