TSMC’s $100 billion pledge won’t resurrect US chipmaking, says Intel’s ex-CEO

1 week ago 10

“I wasn’t done with the five-plus years when the board made a directional change,” he said.

TSMC has said that the only development work it plans to carry out it in the US will be on the process technology it already has in production, and that its core research and development will remain in Taiwan.

“Unless you’re designing the next-generation transistor technology in the US, you do not have leadership in the US,” said Gelsinger.

He was speaking in an interview this week after becoming a partner at Playground Global, a Silicon Valley venture capital firm that specializes in “deep tech” investments, including in fields like quantum computing and new chip making technologies.

Despite losing out in the cutting-edge process technology used in chip manufacturing, Gelsinger said the US still had a global edge in many advanced technologies that are likely to determine future leadership in artificial intelligence.

He also brushed off suggestions that Chinese AI company DeepSeek, which shocked US technology experts and investors with its low-cost technology earlier this year, had posed a serious challenge to American companies.

“DeepSeek was good engineering, it wasn’t core innovations. It wasn’t major breakthroughs,” he said.

Start-ups backed by Playground include xLight, whose advanced lasers could play a part in future generations of lithography needed to make chips. Other investments include PsiQuantum, the first quantum computing company that has embarked on building a large-scale quantum machine, and d-Matrix, one of a number of start-ups trying to leapfrog Nvidia in making the chips needed to run AI systems, a task known as inference.

While at Intel, Gelsinger failed to make up ground against Nvidia in AI chips, though he said this week that new technologies were needed to bring AI into the mainstream.

“AI, as exciting as it is, is much too expensive,” he said. “We have to have dramatic reductions in the cost of inference for it to be truly deployed in every aspect of humanity.”

Read Entire Article