Intel Is Making a Very Big Mistake by Dissing Asia
(Bloomberg Opinion) -- The Western world has a problem, and Intel Corp. is here to fix it.
That’s the apparent message from Chief Executive Officer Pat Gelsinger, who announced plans in March to return his company to the made-to-order chip business and in the process rebalance the supply chain “for something that’s critical to humanity.”
“The majority of leading edge foundry capacity is concentrated in Asia, while the industry needs more geographically balanced manufacturing capacity.”
National security concerns have been rising for some time, and it may be smart business to play on the kind of subsidies they can bring. Yet underneath I’d describe a kind of cultural mindset positing that Western nations are safer than Asian countries, despite some of the U.S.’s closest military, political and economic allies being in the Pacific region. Such an approach is mistaken and could paint the company — as well as America — into a corner just as it seeks more flexibility and security.
It’s true that Taiwan has become the most important player in the global chip industry and is uniquely vulnerable. But the notion that Asia is less stable or predictable than Europe or the U.S. is without merit. Japan and South Korea are among the world’s strongest democracies, with solid rule of law and IP protection, not to mention relatively open and healthy economies. There’s absolutely nothing more predictable than the outcome of Singapore’s next election.
In his initial announcement of Intel’s new foundry plans, during an investor call the following month, and then in an interview with CBS’s 60 Minutes, Gelsinger continuously asserted that expansion is needed in Europe, in addition to the U.S., while leaving out Asia.
Having a dearth of capacity is unhealthy and results in a nation losing the innovation that goes along with the underlying capacity, Greg Slater, Intel’s head of global regulatory affairs, told me. There’s solid logic to this argument, and the American semiconductor ecosystem would benefit by having more fabrication plants. But it doesn’t explain why development in one cluster of foreign nations — such as the European Union — is in the best interests of U.S. national security and close allies in Asia is not.
When asked what made Germany, for example, a good prospect for chip expansion, Slater pointed to a “common culture.”
One tie that does bind the U.S. and Europe is an eagerness by political leaders to spend money in the hope of recapturing former chip glory. This urgency is heightened by a temporary shortage of components for cars, a politically important industry. The E.U. has a goal of meeting 20% of its semiconductor demand from within its own borders, up from 9%, and is willing to spend 30 billion euros ($36 billion) to make it happen. President Joe Biden and his predecessor have spoken of the need to invest more to reinvigorate the U.S. chip sector.
So this has been an opportune time for Intel to announce it will spend $20 billion on two additional plants in Arizona and to appeal to governments for money to build up capacity. The auto parts that are currently scarce are made with technology that’s more than a decade old. By the time any new Intel fabs are up and running — it takes two to three years to build one from scratch — this shortfall will be largely solved.
What Intel really wants to make are the most advanced chips used by Apple Inc., Qualcomm Inc., Nvidia Corp. and Advanced Micro Devices Inc. — American companies without their own factories. Gelsinger is right that most high-end, made-to-order chipmaking is in Asia — Taiwan, to be precise. Taiwan Semiconductor Manufacturing Co. has the market cornered on this supply for the simple reason that it’s spent three decades building the know-how.
TSMC’s problem is its stubborn reluctance to expand capacity outside a 160-kilometer (100-mile) radius. While it does have plans to set up in Arizona, that facility will only account for around 2% of its global supply. And besides the not-insignificant matter of Beijing’s threats of invasion, its home base is afflicted by earthquakes, typhoons, droughts and unstable electricity generation.
It’s not as if Intel is unfamiliar with Asia. The company has a chip factory in Dalian, China, and assembly and test facilities in Chengdu, in addition to Vietnam and Malaysia.
Yet Gelsinger wasn’t talking up other parts of Asia in an April investor call when he argued that “we need a more geographically dispersed resilient supply chain.” Not South Korea, which is already home to Samsung Electronics Co., nor Singapore, once a global semiconductor player. A new foundry in the city-state would give the world an additional supply base 3,000 kilometers away from TSMC’s headquarters.
Intel’s other assertion is that more supply chain diversity is needed to enhance U.S. economic security. The company is a big backer of legislation that’s been drafted in support of the government handing out money for chips. Slater said that there’s a good opportunity for Washington and the E.U. to develop a partnership, and that European companies could also contribute to U.S. security.
Shouldn’t Japan, a close economic and military ally with a solid industry background, be a worthy member? Slater responded that such an alliance should be focused and it’s best not to have “too many players.”
For Intel, playing the national security card is a potentially lucrative one. While Gelsinger and Slater note that expansion and foundry plans would move ahead even without subsidies, getting government support would be a crucial factor in how much and how quickly the company can add capacity. Both the U.S. and Europe feel vulnerable in the global chip race. That offers Intel the chance to be the hero they need.
Yet the focus on semiconductors ignores the fact that chipmaking is only the start of the supply chain. Once it leaves the Class 1 cleanrooms of Taiwan or Arizona, such components get tested and packaged before heading elsewhere to be soldered onto a board alongside other parts, then packaged with a screen or communications ports. The majority of that assembly is done well outside the borders of the U.S. and Europe — in Asia, Eastern Europe and Latin America. These jobs are destined to remain there.
In fact, the U.S. will never be able to replace the global electronics infrastructure. The best it can do is manage the risk and work with trusted nations to ensure every step in the process is reliable. That’s going to be a lot harder if leaders discount an entire region and cling to the notion that common culture is what matters in a secure supply chain.
Spoken in his presentation given that day.
This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.
Tim Culpan is a Bloomberg Opinion columnist covering technology. He previously covered technology for Bloomberg News.
©2021 Bloomberg L.P.