The News: The New York Times reports that TSMC, the Taiwan-based microchip manufacturer, has agreed to build an advanced chip factory in the United States. The Arizona facility, which is expected to employ TSMC’s 5nm manufacturing process, could produce as many as 20,000 wafers per month as early as 2024. While the plant would only be about one fifth the size of TSMC’s massive gigafactories in Taiwan, chips manufactured in the US plant would nonetheless help mitigate the impact of potential disruptions in global supply chains serving US smartphone, technology infrastructure, and defense industries.
While no details have emerged just yet about the types of incentives TSMC may have secured from federal, state, and local authorities to bridge the “cost gap” that such a project would need to overcome in order to become a reality. TSMC’s own investment in the project is reportedly estimated at $12 billion over the course of the next decade. The plant could employ as many as 1,600 people. Read the full article at the New York Times.
TSMC Plan to Bring Advanced Chip Manufacturing to the US Symptomatic of Emerging Geostrategic Shifts, Could Represent a New Inflection Point in US-China 5G Rivalry
Analyst Take: Strengthening critical technology manufacturing in the US, and minimizing the impact of disruptions to a critical technology supply chain: This plan, should it come to fruition, would serve to further reduce US reliance on foreign foundries to provide the US technology market with advanced chipsets used in smartphones, 5G equipment, and critical defense equipment. While not technically part of the US’ current on-again, off-again decoupling-from-China policy, bringing China-adjacent manufacturing capabilities to the US, particularly as it relates to advanced chipsets, would be a strategically sound move for the United States. Why? Because increasing US-based manufacturing capacity for the kinds of advanced chipsets that power critical technologies would help minimize the impact of sudden interruptions in that increasingly vital supply chain. It doesn’t matter if those interruptions might happen as the result of a natural disaster, a pandemic, a trade war, or a military conflict. Disruption to the supply chain is disruption to the supply chain. No matter its cause, increasing the production capacity of critical technology components (like advanced chipsets) in the United States helps mitigate that risk.
This deal, by the way, gives us a fascinating glimpse into the role that the US Department of Commerce often plays in support of US national security interests, particularly when they intersect and/or overlap with critical industry sectors.
- Are we seeing the beginning of a new Taiwan strategy? Taiwan finds itself in a precarious position: somewhat stuck in the crossfire between mainland China and the United States’ occasional trade disagreements, and geostrategic ambitions. Some have argued that that Taiwan’s policy of holding on to its manufacturing capabilities, at least until now, might give the United States an incentive to continue supporting, albeit judiciously, Taiwan’s partial autonomy from mainland China. Expanding production to the United States would certainly appear to be a departure from that strategy. Hypothetically, Taiwan’s bet might be that fostering a closer relationship with the United States, one more akin to a more formal trade partnership, might strengthen US-Taiwan ties at a time when traditional alliances and strategic expectations no longer hold as much as weight as they once did.
This strategy isn’t without risk, but it may be wiser for Taiwan than sticking to eroding paradigms and hoping for the best. Given the ongoing restructuring of global power centers, economies, and technology markets, and how much disruption may yet come our way in the next two to three decades, it would certainly be to TSMC’s advantage to establish new production beachheads in North America and Europe in the coming years.
- Don’t count your chicken until they’ve hatched. It is entirely possible that this project, despite everyone’s best intentions, might not not ultimately deliver on its promises. Delays in approvals and construction could force TSMC to tap the brakes, as could shifting geopolitical and economic conditions. This announcement is a great start, but it is a bit soon to celebrate. (You will recall how Taiwan-based Foxconn’s 2017 announcement that it would build a 13,000 employee plant in Wisconsin fizzled out. After months of controversy and false starts, the company revised its original plans and seemed to downgrade its plans from creating 13,000 jobs to creating only 1,500. Per reporting by The Verge, as of a month ago, Foxconn’s buildings at the site still sat empty.
All of that to say that, while TSMC isn’t Foxconn, best intentions aside, even if TSMC manages to secure the type of investment and incentives that would make its advanced chip manufacturing plant make financial sense, a lot can change between now and 2024. As always, hope for the best, but maintain a healthy dose of caution when it comes to setting expectations this early in the game. A plan is merely that, especially at this juncture.
- Should we expect to see more investments in advanced microchip manufacturing in the United States? It is too soon to tell whether current US efforts to do so will be successful in the long term, or if such efforts will continue to be a priority with future administrations. On the one hand, the strategic advantages of doing so are self-evident, but on the other, building multi-billion-dollar state-of-the art manufacturing facilities to produce advanced chipsets can be cost-prohibitive, and therefore financially untenable without significant government subsidies. Such subsidies can be understandably controversial as plans to build up US-based chip manufacturing capacity could arguably cost bailout-fatigued and subsidy-suspicious taxpayers upwards of a hundred billion dollars over the next decade, and with mixed results. This prospect could become even more problematic if the US decides to continue investing in foreign companies like TSMC rather than in US companies like Intel, NVIDIA, and AMD, for example.
While this latter point may be an issue of optics more than a practical matter (as not all chips and manufacturing processes are equal, and TSMC happens to manufacture chips for US semiconductor companies), providing massive subsidies to non-US chipmakers, seemingly at the expense of US chipmakers, does nonetheless present a sensitive political challenge for individual states, the Departments of Commerce and Treasury, and whichever administration happens to greenlight such investments.
In other words, it is a bit too soon to tell where this is going. The will is certainly there on the part of the US government to pursue this tack, and what tech company wouldn’t be at least intrigued by the possibility of securing multi-billion-dollar incentives and subsidies to build new facilities, but buy-in from the general public may not yet be where it needs to be to make that happen, let alone at scale. This is a complicated topic that may require a good deal of heavy lifting on the part of the US government and select chipmakers to properly educate the public enough to help them understand the importance of such investments, especially if they run into public pushback. “Job creation” has traditionally been far easier to sell to voters than “national security,” but moving forward, success in achieving sufficient buy-in from the public with this particular topic may require all interested parties to look beyond that binary equation.
To be continued.
Futurum Research provides industry research and analysis. These columns are for educational purposes only and should not be considered in any way investment advice.