Power On with Mark Gurman · Nov 12, 2026 · 40m · 34m saved
With Gerard Williams III, Founder & CEO at Nuvacore (formerly founder of Nuvia, acquired by Qualcomm; former Chief CPU Architect, Apple)
Photo via nuvacore.com
AI has cut the early phase of designing a new computer chip from as long as a year down to two or three months, says Gerard Williams III, founder and CEO of Nuvacore. The industry's dominant narrative is built around Nvidia's GPUs; Gerard, who spent nearly a decade as Apple's chief CPU architect, is betting his new company on the chip Silicon Valley talks about least.
"What would have taken seven months to 12 months to get to a point where you are ready to start the design process, we've actually collapsed that down to a two or three month window." — Gerard Williams III
Gerard shipped the A7, Apple's first 64-bit chip, and worked through the M1 before leaving to found Nuvia, which Qualcomm bought for a reported $1.4 billion, and then Nuvacore, seven months old, told Mark Gurman of Bloomberg on Power On.
The full interview is covered here so you can skip it. 40 minutes of audio, 6 minutes of reading.
Here are the 5 lessons that matter.
Key Takeaways
Milliwatts saved per core, multiplied across millions of cores, become a material change in a data center
A CPU is never one person's work, Gerard says — it's what a team builds
Gerard calls new Apple CEO John Ternus "a good person" and won't say more
He's betting Nuvacore on CPUs, not GPUs, because LLMs still depend on them
AI is compressing Nuvacore's chip-design cycle, but it still needs an engineer to check its work
1. Apple to Chief Architect
Gerard joined Apple's small Platform Architecture team in February 2010, evaluating every piece of silicon the company considered building, and rose to become its chief architect of CPU development.
Gerard's first shipped chip was Apple's first 64-bit processor
"I had the opportunity to become Apple's chief architect of CPU development. The first product that saw the light of day that I worked on was the Apple A7." — Gerard Williams III
He traces Apple's performance-per-watt obsession to his earlier work at TI and ARM, when devices sipped only 50 to 100 milliwatts and every milliwatt saved was a large share of a chip's total budget.
That same discipline scales directly into the data center
"You hear people throw out numbers like gigawatts and multi-gigawatts. I will tell you that milliwatts are still a part of that number... if you can save milliwatts in every CPU that you construct, by the time you multiply that out by tens, hundreds, thousands, if not millions of cores, it's a material change that you can make." — Gerard Williams III
