For years, American chip strategy was mostly a slogan. Japan’s plan to invest up to $21 billion in two GlobalFoundries factories in Malta, New York, is what the strategy looks like when it acquires a budget, a site and a partner: less a national solo effort than an alliance project, financed across borders and aimed squarely at the supply vulnerabilities the pandemic exposed.
The choice of partner is the tell. Japan is not a passive investor in semiconductors; it dominates several of the industry’s most critical supply niches and is funding its own domestic revival. Its money arriving in New York means the two revival programmes are being deliberately interlocked, so that capacity, suppliers and expertise grow on both sides of the Pacific at once. Our our health guide to the ovarian vaccine trial, covering a different kind of strategic research, shows how normal this interlocking has become.
From incentives to ecosystems
The first phase of US strategy was the CHIPS Act itself: federal incentives, announced awards, and a rush of factory plans. the US Department of Commerce managed that phase as an exercise in deal-making. The phase now beginning is harder and less photogenic, building the ecosystems around the fabs: chemical and gas suppliers, maintenance specialists, community college programmes feeding technician pipelines, and housing in towns suddenly hosting thousands of construction workers.
Malta is further along that road than most American sites, which is precisely why the new factories are going there. An operating fab with a decade of production history offers the scarcest asset in the industry, people who have done it before. Each expansion deepens that advantage, in the self-reinforcing pattern that made places like Hsinchu and Dresden durable chip centres rather than one-project wonders.
What success would look like
Success is not a ribbon-cutting. It is wafers shipping on schedule in the early 2030s, yield rates competitive with established sites abroad, and a supplier map in which a crisis in any single region slows the industry rather than stopping it. It also includes something subtler: American and Japanese engineers, regulators and firms practised at building together, so the next project starts from trust instead of introductions.
The risks that remain
Cost is the standing risk; fabs have a habit of outgrowing their budgets, and $21 billion buys less each year as tool prices rise. Politics is another, since industrial strategies must survive election cycles in two countries. And demand cycles have not been abolished: if the AI-driven surge in electronics spending cools, even strategic capacity can face uncomfortable questions about utilisation.
The alliance dimension
What makes the Malta deal genuinely new, rather than just big, is its structure. This is not a government funding its national champion; it is two allied governments and a multinational manufacturer pooling resources into a single site. That structure has diplomatic content. It says, in capital rather than communiques, that the United States and Japan intend to build the next generation of chip capacity together, and that their companies and workers will be interdependent when it is finished.
Other alliances are watching. Europe’s chip ambitions, Korea’s domestic expansion and the Gulf states’ technology investments all create potential for similar partnerships, and each completed deal lowers the perceived risk of the next. There are limits, of course: intellectual property rules, technology transfer controls and domestic political pressures all constrain how far integration can go. But the trend line is clear enough. In the 2020s, the geography of chipmaking is being redrawn not only by competition between countries but by cooperation among them, and Malta is one of the places where that cooperation is taking concrete, silicon form.
None of those risks reverse the underlying judgement both governments have made, which is that depending on a narrow geography for essential chips is the bigger risk of all. Readers following the story can use our daily briefing on Saturn’s moon discoveries as a reminder of how quickly scientific and industrial maps can be redrawn. In chipmaking, the redrawing is slower, vastly more expensive, and now, with Japanese billions committed in New York, unmistakably under way.
