Arm’s CEO says AI could help humanity cure cancer within our lifetimes. First, the industry needs to figure out how to make enough chips to power it.
“I’ve always thought that the killer app for AI is health,” Rene Haas said in an interview with the BBC released on Tuesday. “AI is going to not only shorten the amount of time that those drugs can be invented, it’s going to shorten the amount of time that you test them.”
Haas has led Arm since February 2022, after joining the British chip designer in 2013. Before Arm, he spent seven years at Nvidia, where he was vice president of its computing products business.
In the interview, the Arm CEO said that AI could crack problems that humans alone might not.
“I believe in our lifetime, AI will help cure cancer,” he said.
Getting there, however, will require enormous amounts of computing power, Haas said. And the chips needed for cancer research or humanoid development are being gobbled up by AI companies planning data centers measured in multiple gigawatts.
“Right now, it’s quite constrained. We need more fabs before we can put a data center in space, I’ll tell you that much,” he said. “We’ll probably only put data centers in space when the biggest impediment to data centers is the cost of the data center.”
Haas said AI’s appetite for chips and memory is at a scale the semiconductor industry hasn’t dealt with before. Building new factories isn’t a quick fix: fabs can cost tens of billions of dollars and take two or three years to construct.
He said the industry was in an “absolutely supply-constrained environment” and expected supply to remain tight for some time.
Even without the chip bottleneck, AI curing cancer in the next few years is far from guaranteed. This is because medical researchers say that cancer is not a single disease but a collection of over 100 diseases that vary in patients.
While AI can help researchers identify targets and predict how drugs might behave, evidence of its clinically relevant impact remains limited. The wide range of variables in each body also makes it difficult for AI models to reliably predict how a potential drug will behave in humans.
Data center boom
Other chip executives are sounding similar alarms.
In a January interview, Micron’s chief operating officer, Manish Bhatia, called the memory shortage “really unprecedented” and said high-bandwidth memory for AI accelerators was consuming so much manufacturing capacity that it was leaving a “tremendous shortage” for conventional products such as phones and personal computers.
In a February earnings call, Qualcomm CEO Cristiano Amon said “industry-wide memory shortage and price increases” would determine the size of the handset market through the year, adding that “the whole sector is impacted by memory.”
Data center construction in the US has skyrocketed in recent years. By the end of last year, more than 1,400 had been built or approved for build-outs. They have become increasingly divisive due to their impact on air and water quality in the neighborhoods in which they are being built.
A Gallup survey of 1,000 adults in the US, published in May, found that seven in 10 respondents opposed constructing data centers for AI in their local area. Nearly half indicated they were “strongly opposed” to construction.
