Arm CEO Rene Haas says artificial intelligence could help cure cancer within his lifetime by speeding up drug discovery and testing. But he warns the chip industry cannot yet build data centers fast enough to supply the computing power AI research needs.
Arm's chief executive says AI could help humanity cure cancer, but the industry first has to solve a chip supply problem. In an interview with the BBC released Tuesday, Rene Haas said health is the killer application for AI because the technology can shorten how long it takes to invent and test new drugs.
Haas, who has run Arm since February 2022 and previously spent seven years at Nvidia, said AI could crack problems humans alone might not solve. In the BBC interview, he said: "I believe in our lifetime, AI will help cure cancer."
Fabs can't keep up with AI demand
Reaching that point, however, requires far more computing power than the industry currently has. Haas said the chips needed for cancer research and humanoid robotics are instead being consumed by AI companies building data centers measured in multiple gigawatts. New fabs can cost tens of billions of dollars and take two to three years to build, so the shortage will not resolve quickly. Haas described the industry as being in a state of extreme supply constraint and said he expects supply to stay tight for some time.
Other chipmakers echo the shortage warning
Haas is not alone in flagging the squeeze. In a January interview, Micron chief operating officer Manish Bhatia called the memory shortage unprecedented, saying high-bandwidth memory demand for AI accelerators was leaving a severe shortage of conventional memory for phones and PCs. Similarly, on a February earnings call, Qualcomm CEO Cristiano Amon said a memory shortage and price increases across the industry would determine the size of the handset market for the rest of the year.
Data centers face growing pushback
Meanwhile, data center construction in the US has skyrocketed, with more than 1,400 built or approved for build-out by the end of last year. Yet a Gallup survey of 1,000 US adults, published in May, found that seven in ten respondents opposed building AI data centers in their local area, with nearly half saying they were strongly opposed.
Even without the chip bottleneck, curing cancer with AI is far from guaranteed. Medical researchers note cancer is not one disease but a collection of more than 100 that vary by patient, and AI's clinically relevant impact so far remains limited.
Source: Business Insider
Trading involves risk.
Risk Disclaimer
