The pandemic created enormous challenges for semiconductor companies.
Unprecedented demand drove up prices and backlogs for chips and equipment as supply chains faltered. At the same time, concerns grew about dependence on China and Taiwan. As a result, policymakers passed the $53 billion Chips and Science Act.
With globalization’s golden age potentially in the rearview, we are seeing a rise in nationalistic policies. Some of this due to uncontrollable circumstances, such as Xi Jinping’s microaggressions against Taiwan and Vladimir Putin’s war on Ukraine. Even Taiwan Semiconductor Manufacturing Co.
founder Morris Chang is calling globalization all but dead.
The passage of the Chips and Science Act is the beginning of a longer process of building greater resiliency and lower dependence on Taiwan and China. It’s also about protecting domestic technology leadership and dealing with threats to national security. These three items make up the trifecta of critical concerns for passing the law: national security, supply-chain resiliency and technology leadership.
And despite the bill being passed, there is a second wave of activity among semiconductor companies, and that is how the dollars are appropriated. That begs the question of how the U.S. “should” be distributing subsidies to achieve the best outcomes for the U.S. and our global trading partners.
How funding should flow
I believe that Intel
the largest semiconductor manufacturer in the U.S., should receive the largest portion of the $52.7 billion total, which was headlined by the $39 billion in manufacturing incentives.
An American company, Intel has made massive commitments to expand manufacturing in the U.S., including foundry services. That would satisfy the critical infrastructure and defense requirements. With questions around Intel’s capabilities looming, there will be many doubters, making this the ultimate moment for Intel to show its plans under CEO Pat Gelsinger to regain technology leadership.
The next biggest share of the Chips Act should go to Micron Technology
in that order.
Micron has made massive domestic commitments, including a recent $40 billion investment in memory manufacturing, which will create tens of thousands of jobs in the long term following this temporary down-cycle for semiconductors. The company is single-handedly driving U.S. production of memory from the low single digits to near 10% globally over the next decade.
GlobalFoundries and Tower Semiconductor
are going to be critical in expanding manufacturing of the lagging edge of semiconductors. The Chips and Science Act grossly underinvested in this area, but semiconductors over 14 nanometer (nm) make up the vast majority of all semiconductors used in manufacturing. The lack of expansion of this kind of chips is causing continued supply-chain issues for items like automobiles and appliances.
With customers including Samsung Semi and Intel, IBM plays an interesting role in semiconductor manufacturing, and research and development in the U.S., which isn’t talked about often. The company’s New York-based research centers continue to put out critical ideas that drive semiconductor manufacturing, which will be implemented in future process innovations, such as its 2 nm wafers.
With Sen. Chuck Schumer seeking to put New York on the map as a tech hub, IBM, GlobalFoundries and Micron benefit from his overtures, and I expect all of them to see significant dollars from the Chips and Science Act.
Best of the rest
Finally, I support continued significant investment in leading fabless chip makers that commit to manufacturing more semiconductors in the U.S. to receive among $13.2 billion in subsidies earmarked for R&D and workforce development.
and others are significant contributors to the U.S’s strong global technology leadership in data center, edge, AI, automotive and devices. And their investments in R&D are critical to sustain that leadership.
We should want U.S.-based companies to continue to pour investment into critical research and development. While the act is more focused on manufacturing, it would be short-sighted not to see the interdependence of strong fabless chip makers, manufacturers and global technology leadership.
How does Taiwan Semi figure into this?
None of this is to suggest Taiwan Semi won’t remain a strong leader in manufacturing massive volumes of leading-edge semis for U.S.-based fabless leaders such as Apple
Qualcomm, Nvidia, AMD and Marvell
However, funding TSMC, a company that has already benefited massively from globalization and offshoring, does very little to support the critical tenets of the Chips and Science Act. Our largely favorable trading relationship with Taiwan has already enabled the company to become a global semiconductor juggernaut — further subsidies not required, in other words. With the fragility of China-Taiwan relations, further investment into Taiwan Semi seems too risky and uncertain for Chips and Science Act dollars to support.
Daniel Newman is the principal analyst at Futurum Research, which provides or has provided research, analysis, advising or consulting to Oracle, Cisco, Juniper and dozens of other technology companies. Neither he nor his firm holds any equity positions in companies cited. Follow him on Twitter @danielnewmanUV.