U.S. Semiconductor Industry Veterans Keep Wary Eyes on China

Does the U.S. need its own giant silicon foundry to protect itself from China’s ambitions?

2 min read
Abstract illustration of a microchip on a pale red/yellow gradient
Illustration: iStock

How might the U.S. chip industry solve a problem like China?

A panel of semiconductor industry veterans took up this question at a Churchill Club event this week. The group generally expressed worry about the impact China will have on the future of the U.S. chip industry, and the lack of good ideas about how the U.S. industry can respond to threats posed by China.

“China is the ultimate conundrum,” says Stanford president emeritus and MIPS Computer Systems founder John Hennessy. “It’s a large market that U.S. companies need access to, together with being what will become a major technical competitor. We have never faced that.”

The consolidation of silicon manufacturing into two main foundries raises the threat level, pointed out Diane Bryant, former Intel and Google Cloud executive.

“You really just have TSMC and Samsung left,” she said. “And TSMC is in Taiwan, so you have to be thinking about China and the threat to Taiwan, and what will happen to TSMC.”

China will take over Taiwan “the same time North Korea takes over South Korea,” quipped Hennessy, giving it control over most of the world’s semiconductor manufacturing capabilities.

“What do you do tomorrow if TSMC and Samsung are off limits?” he asked his fellow panel members.

“You can’t go to Global Foundries,” which indeed has some U.S. semiconductor manufacturing capability, said Bryant, “unless you really want Moore’s Law to be dead.” (Global Foundries recently stopped developing the most advanced semiconductor processes.) 

Rodrigo Liang, CEO of SambaNova Systems, argued that fixing this problem can only be done at the level of the U.S. government.

Pradeep Sindhu, founder of Juniper Networks and founder and CEO of Fungible, agreed. “The U.S. government needs an industrial policy,” he said, “and it doesn’t have one.”

The foundry issue is a long-term problem. Perhaps a nearer term question is how the growing capability of China’s tech industry will impact U.S.-based companies.

“China is talking about becoming tech independent, becoming net exporters,” said Bryant. “We can talk about how many years [it will take], but it is inevitable.”

Companies in China will catch up for several reasons, panelists indicated. For one, said Sindhu, they are very hungry to learn.

For another, said Navin Chaddha, managing director of the Mayfield Fund, China’s huge market gives Chinese companies a boost. “Usually innovation happens when you are close to a market,” he said. To date, the U.S. companies and Samsung have benefitted from the boom” in the Chinese tech market, but now “we are seeing Chinese companies benefitting from their local market… and China is the biggest market when it comes to broadband users.”

A solution?

“Invest in that market,” says Chaddha.

That strategy is not without pitfalls, Hennessy indicated. “What happens to your technology when you ship it over there?” he asked.

“To the extent that we can protect it, we will,” Sindhu said.

Hennessy remained skeptical. “Just wait until you sign the deal and send it over,” he said.

“This isn’t a redo of semiconductor wars with Japan in the 80s,” he concluded.  “This is a country that has scale, that has entrepreneurial zeal. They will give us a run for the money.”

The Conversation (0)

The Ultimate Transistor Timeline

The transistor’s amazing evolution from point contacts to quantum tunnels

1 min read
A chart showing the timeline of when a transistor was invented and when it was commercialized.
LightGreen

Even as the initial sales receipts for the first transistors to hit the market were being tallied up in 1948, the next generation of transistors had already been invented (see “The First Transistor and How it Worked.”) Since then, engineers have reinvented the transistor over and over again, raiding condensed-matter physics for anything that might offer even the possibility of turning a small signal into a larger one.

Keep Reading ↓Show less