The legendary architect behind Apple's M1 chip discusses the AI power crisis: building supercomputers with the "frugal" philosophy of iPhone chip design
IT之家 reported on November 13 that in the latest episode of the Power On interview program, Bloomberg's Mark Gurman (Mark Gurman) invited former Apple chief CPU architect Gerard Williams III (Gerard Williams III) as a guest to discuss the current AI wave.
According to public information reviewed by IT之家, Williams is a chip engineer who has worked deeply in CPU architecture for decades and is one of the key figures behind Apple's in-house chips moving from phones to Macs. At Apple, he led the CPU architecture of the A-series and M-series processors, and later founded the AI data center chip company NUVACORE, advocating bringing "performance per watt" from mobile devices to the cloud.
In 2019, he left Apple and co-founded NUVIA, which was later acquired by Qualcomm, where he led the Oryon CPU architecture; in 2026, he co-founded the new company NUVACORE.

During his time at Apple, he participated in the development of multiple generations of in-house processors from the A7 to the M1. The A7 was Apple's first 64-bit mobile processor, while the M1 was the landmark product that moved the Mac from Intel chips to Apple's own Apple Silicon. This experience made Williams a representative figure of Apple's "in-house chips + software-hardware integration" approach.
In this interview, Williams argued that although the entire industry is frantically chasing 英伟达's GPUs, the CPU is the behind-the-scenes protagonist of AI data centers, and the CPU's role as the "brain" is irreplaceable. The GPU is AI's "compute engine," while the CPU is the