Ex-Intel CEO Gelsinger warns TSMC's $165B investment will not restore U.S. semiconductor leadership

Intel
(Image credit: Intel)

TSMC's plan to spend $165 billion on its American manufacturing capacity and an R&D facility will certainly increase U.S. semiconductor production market share. But it doesn't guarantee that the country will regain leadership in process technologies, said Pat Gelsinger, former chief executive of Intel, in an interview with the Financial Times.

"If you do not have R&D in the U.S., you will not have semiconductor leadership in the U.S.," Gelsinger told the Financial Times. "All of the R&D work of TSMC is in Taiwan, and they have not made any announcements to move that [to the U.S.]."

If you don't have R&D in the U.S., you won't have semiconductor leadership.

Pat Gelsinger

Gelsinger argued that manufacturing alone is not enough to regain technology leadership, even though this will undoubtedly improve the semiconductor supply chain in America. Nonetheless, the U.S. cannot lead in this field without designing next-generation process technologies domestically.

Under the current plan, TSMC plans to build six Fab 21 modules to process wafers using various fabrication technologies, two advanced packaging facilities, and an R&D center in the U.S. The company hopes to house everything at its Fab 21 site near Phoenix, Arizona, though no final decisions about the locations have been made as of last month, when Tom's Hardware spoke with TSMC.

While TSMC's R&D center is certainly planned in the U.S., its focus remains to be seen. TSMC has developed its fabrication processes in Taiwan for decades, although many of its engineers come from the U.S. As manufacturing technologies get more complex and require longer pathfinding processes with each generation, TSMC may offload part of its R&D operations from its Taiwan facilities to the U.S. facility.

TSMC has reportedly hinted that its U.S. development will focus only on refining existing processes, though it did not elaborate. Contract chipmakers like TSMC tend to develop performance-enhanced versions of their process technologies for their alpha customers (e.g., N3 => N3P => N3X) as well as specialty versions of their production nodes for customers that prefer to use mature nodes (e.g., N4 => N4e, N4c). This ensures that their fabs are always filled with orders and their utilization rates are high.

The former Intel executive also acknowledged that tariff threats by President Donald Trump helped the U.S. semiconductor industry by motivating foreign chipmaker TSMC to build additional U.S. facilities. Still, he suggested this was insufficient without deeper R&D operations in the U.S.

Anton Shilov
Contributing Writer

Anton Shilov is a contributing writer at Tom’s Hardware. Over the past couple of decades, he has covered everything from CPUs and GPUs to supercomputers and from modern process technologies and latest fab tools to high-tech industry trends.

  • leoneo.x64
    Admin said:
    Although TSMC will build an R&D facility in the U.S., its core R&D will remain in Taiwan, so the U.S. will keep following Taiwan even after TSMC builds out its Fab 21 site with six fab modules and two packaging facilities, says former CEO of Intel.

    Ex-Intel CEO Gelsinger warns TSMC's $165B investment will not restore U.S. semiconductor leadership : Read more
    Go sniff "Gloo" dude. No one asked for your opinion. Your time is over.
    Reply
  • hotaru251
    leoneo.x64 said:
    Go sniff "Gloo" dude. No one asked for your opinion. Your time is over.
    ...except hes right.

    if you only follow what others do you arent leading you are following & in event relations sour you have no R&D and fall behind very fast.
    Reply
  • leoneo.x64
    hotaru251 said:
    ...except hes right.

    if you only follow what others do you arent leading you are following & in event relations sour you have no R&D and fall behind very fast.
    He's a self-promoting noise maker. Nothing rules out R&D from this investment. Intel had decades to do R&D but they rode the wave. Intel lost their fab edge and steady node progression is one way to achieve optimisation.

    He's dishing out advice to new CEO and now this. Moans about Nvidia being lucky. No one asked him. Lol. I agree that he was the tiniest bit better than the horrible stagnation-DJ krzanich but it's been mostly words from him.

    nVidia uses TSMC. They do their own R&D. Mr Gelsinger will yell anything to get heard.
    Reply
  • derekullo
    Getting the usb orientation correctly on the first try is luck.
    Becoming the most valuable company in the world isn't luck!
    Reply
  • Pierce2623
    leoneo.x64 said:
    Go sniff "Gloo" dude. No one asked for your opinion. Your time is over.
    Yeah the last true engineer CEO of Intel has zero idea about fabrication tech….Get real guy.
    Reply
  • jlake3
    Pierce2623 said:
    Yeah the last true engineer CEO of Intel has zero idea about fabrication tech….Get real guy.
    While there's a point buried in what Pat said, during his tenure Intel struggled on the fab side, had issues with CPU design, missed the golden window to get into gaming GPUs, and seems to have mostly missed the boat on AI accelerators as well. Sure it was a mess when he arrived, but he didn't exactly knock it out of the park on the turnaround.

    After his dismissal he landed at a cloud platform for churches... which I'm sure is important, but for someone coming from a chip design and foundry company it feels like maybe an indicator that no company doing hardware was interested.

    Despite the fact that he was fired and people are regularly speculating whether Intel has the runway left for a turnaround and if they could be forced into a sale, he seems to regularly be talking to anyone who puts a mic in front of him about how great his plans were and offering unsolicited advice to his replacement while almost lashing out at certain things. Like, why does he need to weigh in on this issue at all now that he's the head of a faith-based cloud service company, other than that he's either got an axe to grind or really likes hearing himself in interviews?
    Reply
  • derekullo
    Intel's market cap is $102 billion. (A figure Gelsinger is hopefully well aware of)
    Hypothetically TSMC buying Intel would give TSMC a U.S. semiconductor leadership.
    Reply
  • usertests
    Another lame Pat in exile opinion. The point is for parts of the supply chain to be moved here before it's too late. TSMC may even be forced to become an American company in the end, with the Taiwanese fabs bombed to rubble.
    Reply
  • phead128
    Mr. AMD in the rear window and Mr. NVIDIA was lucky. Nobody asked for his opinion
    Reply
  • Pete Mitchell
    Why is anyone wasting headlines on this guy? He was at the helm when Intel faceplanted, so he really needs to just go quietly off into retirement.
    Reply