When veteran engineer and executive Pat Gelsinger returned to Intel as CEO in 2021, the once-great chipmaker was in a slump. After failing to adapt to the mobile era and then missing several steps in cutting-edge microprocessor manufacturing, it was now also falling behind in supplying chips to feed the tech industry’s growing hunger for artificial intelligence.
With optimism that at times seemed reckless, Gelsinger promised that Intel would make an epic comeback. He vowed to shake up its sleepy corporate culture, refocus on core engineering, and deliver a revitalized manufacturing plan that would put rivals TSMC and Samsung on notice.
This week, Gelsinger declared Intel’s comeback plan well and truly on track. He announced a rebrand of the company’s “foundry” business, which manufactures chips designed by other companies, saying that Intel’s latest manufacturing process would later this year yield silicon chips as efficient and capable as ones from TSMC. Microsoft is the first big customer for this new chipmaking technology—a key coup for Intel as it tries to convince the industry that it can offer competitive products suited to the age of AI.
Pat Gelsinger spoke to WIRED senior writer Will Knight about Intel’s AI reboot over Zoom from his home in Santa Clara, California. The conversation has been lightly edited for length and clarity.
Will Knight: You announced this week that Intel will relaunch its business that manufactures chips on behalf of other companies as an “AI-era system foundry.” What does that mean?
Pat Gelsinger: I began Intel’s strategy two plus years ago, and for the company, generative AI has been this unexpected surge. This has been the land of Nvidia, but we’re the one company that actually has the opportunity to participate in 100 percent of the AI market. We know how to connect up networks and memory and [provide] supply chains and all of these other elements that we’re finding customers are super excited to take advantage of.
Speaking of the AI surge, what did you make of reports suggesting OpenAI’s CEO Sam Altman wants to raise $7 trillion to develop and manufacture chips needed to guarantee progress in AI?
My first reaction was, that’s a mind-bogglingly big number. And then I had to do the math. Today, the biggest AI models were generated on about 10,000 GPUs. The belief is there that we probably need to be 10 million for the biggest AI models that get produced in the future.
We’re already saying we may spend a couple billion dollars training the most advanced models today. Plus, the math in the $7 trillion also includes power and data centers.
This week you said Intel is on track to deliver its new “18A” manufacturing process, which will compete with TSMC’s best offerings. What else are you doing to regain an edge?
The whole industry is pursuing this next generation transistor, what we call ribbonFET. I think everybody’s [asking] who’s going to produce the next best transistor on the planet.
But the thing that everybody is giving us credit for is backside power, this new way of delivering power into the device, which gives you better current resistance performance, but it’s also improving the density of the chip. That means the same wafer, instead of producing 100 chips, can produce 120 chips. It’s a huge value proposition.
You announced Microsoft as a customer of your foundry business. But Intel previously fell behind the competition in this market. How will you convince customers that things are different this time?
