Intel CEO Says Nvidia’s Jensen Huang Is Wrong About a Basic Concept of Semiconductors

"Unlike what Jensen might have you believe, Moore's law is alive and well," Intel CEO Pat Gelsinger said at Computex 2024.

Man in suit stands on stage in front of blue screen
Intel CEO Pat Gelsinger delivers a speech during Computex 2024 in Taipei on June 4, 2024. I-HWA CHENG/AFP via Getty Images

Despite falling behind rival chipmakers like Nvidia (NVDA) in recent years, the once dominant Intel isn’t backing down from its competitors anytime soon. Speaking at Computex 2024 in Taiwan today (June 4), Intel CEO Pat Gelsinger discussed a host of new company products and pushed back against claims from Nvidia’s Jensen Huang that the company’s traditional processors are unable to keep pace in the era of artificial intelligence (A.I.)

Sign Up For Our Daily Newsletter

By clicking submit, you agree to our <a href="http://observermedia.com/terms">terms of service</a> and acknowledge we may use your information to send you emails, product samples, and promotions on this website and other properties. You can opt out anytime.

See all of our newsletters

“Unlike what Jensen might have you believe, Moore’s law is alive and well,” Gelsinger said. The phrase Moore’s law was coined in the 1960s by Intel co-founder Gordon Moore and posits that the number of transistors on a computer chip would double every two years, with computing power becoming faster and more affordable. While the sentiment has driven the semiconductor industry for decades, some tech leaders like Huang have claimed it is no longer applicable to the industry and instead believe innovations in A.I. and accelerated computing are what will drive chip performance gains.

Intel for decades was a titan of the semiconductor industry. But the company has been sidelined in the A.I. boom benefiting chipmakers like Nvidia, AMD (AMD) and Qualcomm. Nvidia, whose revenue and profit both hit record highs in the first quarter, is estimated to own more than 70 percent of the global market for A.I. chips and counts major tech companies like Meta (META), Microsoft (MSFT), Google (GOOGL) and Amazon among its customers.

This isn’t the first time Gelsinger and Huang have butted heads on computing methods. Huang over the years has reiterated his belief that Moore’s law has come to an end, often labeling the concept as “dead” and telling Wired earlier this year that “we have to leave Moore’s law behind so we can think about new ways of scaling.”

How Pat Gelsinger missed on the A.I. opportunity

Gelsinger first joined Intel in 1979 and briefly left the company three decades later before returning as CEO in 2021, while Huang co-founded Nvidia back in 1993. In a 2022 interview at the Massachusetts Institute of Technology, Gelsinger claimed he missed out on an opportunity to lead the A.I. wave when Intel canceled Larrabee, a graphics processing unit (GPU) project he was working on before he exited. “When I was pushed out of Intel 13 years ago, they killed the project that would have changed the shape of A.I.,” he said, adding that Huang “got extraordinarily lucky” to branch out into A.I. accelerators at the right time.

Both Nvidia and AMD announced their next generations of A.I. chips earlier this week at Computex 2024. During his keynote today, Gelsinger unveiled Intel’s new Xeon 6 data center processors, describing them as more efficient and powerful than previous iterations. “When I think about this A.I. era I consider it like the internet 25 years ago, it’s that big,” he said. “We see this as the fuel that’s driving the semiconductor industry to reach $1 trillion by the end of the decade.” The global semiconductor industry brought in more than $526 billion in sales in 2023, according to the Semiconductor Industry Association. Nvidia alone claims a market cap of close to $3 trillion.

Gelsinger additionally took shots at fellow tech companies when announcing the pricing for Intel’s Gaudi 2 and Gaudi 3 A.I. accelerators, the latter of which was unveiled in April and will be used by companies like Dell (DELL), Hewlett Packard Enterprise and Supermicro. Kits with eight Gaudi 2 accelerators are listed at $65,000 and marked at $125,000 for Gaudi 3, prices Intel claims are respectively one third and two thirds more affordable compared to that of rivals like Nvidia. “Pretty compelling,” said Gelsinger. “In other words, it crushes the competition.”

Intel CEO Says Nvidia’s Jensen Huang Is Wrong About a Basic Concept of Semiconductors