Amid claims that Micron is not a “US manufacturer,” its manufacturing sites have long been concentrated in Asia, with US facilities accounting for only a small amount of production. Following the CHIPS Act, the company shifted toward building facilities from the ground up in the US, bringing its dependence on technologies such as HBM and the long-term investment risks into focus.
Micron has long been viewed by the market as a US manufacturing company, but in reality most of its manufacturing sites are concentrated in Asia, including Taiwan, Japan and Singapore. Its facilities in Boise and Virginia accounted for only a small fraction of production. The program’s host has argued that the core of its engineering operations was in Taichung and Hiroshima, rather than Idaho. This suggests that the US has a comparatively weaker technological advantage.
At the same time, its acquisition strategy over the past 25 years had served as a means of “survival,” including the absorption of TI’s memory business in the late 1990s. An attempt to acquire Hynix ended in failure, while political and economic barriers were also significant. This has prompted the view that the strategy of acquiring inexpensive existing fabs is reaching its limits over the long term. Seeking new growth, Micron will be forced to decide whether to shift toward building fabs from the ground up in the US.
Current performance, centered on HBM, is strong. Revenue from data centers and cloud customers has risen sharply, HBM revenue is on the order of $10 billion, and HBM4 shipments will be handled by US facilities. However, HBM is not viewed as a long-term solution, and concerns have been raised that future architectural innovation could render it obsolete. This casts doubt on the economics of plans to build new fabs, particularly in New York.
New fabs in the US are being considered at multiple sites, including New York, Virginia and Idaho, and their distance from overseas ecosystems could become an issue. Unlike the previous Asia-centered model, the higher costs of operating in the US, along with the characteristics of its materials and labor markets, will have a major impact on the investment case. Those involved want to determine how competitive a production system built from the ground up can remain through the next downturn. The view presented is that a conclusion will be reached in 2035.