A 28-nm fab gets the political wind
On May 16, 2026, in The Hague, Tata Electronics and ASML signed a Memorandum of Understanding to equip India's first commercial 300-mm semiconductor fab in Dholera, Gujarat. The signing was witnessed by Prime Minister Narendra Modi and Dutch Prime Minister Dick Schoof. ASML's release describes a "holistic suite of advanced lithography tools and solutions" — industry shorthand for the ArF immersion DUV systems used to pattern features down to roughly 28 nm. Tata Electronics CEO Randhir Thakur said the partnership will "ensure the timely ramp of our Fab in Dholera" and build a "resilient and trusted supply chain." The $11 billion facility will run 28-nm to 110-nm processes, with technology partnership from Taiwan's PSMC.
The political subtext is the more interesting story. ASML cannot sell its top-end DUV immersion tools — the TWINSCAN NXT:1970i, NXT:1980i and NXT:2000i and successors — anywhere it pleases. Since September 7, 2024, Dutch government export-licensing has captured all three, bringing them in line with the already-controlled EUV systems and aligning with US Export Administration Regulation 734.4(a)(3). The express purpose, as the US-Dutch-Japanese trilateral has made clear over the last three years, has been to deny these tools to China's advanced-node foundries. A bilateral signing ceremony in front of two heads of government is the clearest signal yet that India sits comfortably on the permitted side of that line.
What the export-control regime is really doing
Steelman first. There is a coherent case for the existing Dutch and US controls on DUV immersion equipment. ASML's NXT:2000i and successor systems are the only tools capable of patterning sub-7-nm logic at volume; if they reach a strategic competitor before allied economies have built their own redundancy, the resulting asymmetry would be structural and slow to unwind. From the Bureau of Industry and Security's perspective, the September 2024 expansion was not protectionism dressed in security clothing — it was a narrow, end-use-focused intervention applied at the upstream chokepoint where the unit count is small and the technology is genuinely irreplaceable.
But that same logic also implies the regime is supposed to be permissive to trusted partners running non-frontier nodes. That is exactly where Dholera sits. Its 28-nm to 110-nm output is automotive, industrial, mobile, and AI-adjacent — segments that the US, EU, and Japan are all explicitly trying to diversify out of Taiwan. ASML President & CEO Christophe Fouquet's pledge to establish "long-term partnerships in the region" reads as confirmation that the Dutch licence framework, in practice, is operating as advertised: gating, not blanket prohibition.
India's industrial policy bet
The Tata-ASML signing slots into a wider Indian policy architecture that has taken roughly four years to assemble. The India Semiconductor Mission's Fiscal Support Agreement with Tata Electronics and Tata Semiconductor Manufacturing committed ₹91,000 crore (~$11 billion) to the Dholera fab, with the Centre underwriting 50% of eligible project costs on a pari-passu basis and a planned capacity of 50,000 wafers per month. The Cabinet has now approved 12 fabs and packaging facilities across six states under the ISM umbrella, with cumulative investment of roughly ₹1.64 lakh crore. On April 9, 2026, MeitY notified a Special Economic Zone covering the Dholera site.
Each of those decisions is individually defensible. Stacked together, they amount to a substantial public-money bet that India can compete at trailing- and mature-node manufacturing within the trusted-partner perimeter the US and EU are drawing around their own reshoring efforts. The TRUST initiative (Transforming the Relationship Utilizing Strategic Technology), launched by President Trump and Prime Minister Modi in February 2025, and India's subsequent invitation to the US-led Pax Silica AI-and-semiconductor coordination grouping, give the bet diplomatic cover.
Where the proportionality argument gets harder
There are two worries on the road from here. First, the Dutch licence regime has been transparent about what it restricts but opaque about to whom. Industry would benefit if the Netherlands and the United States published, even at high abstraction, the criteria distinguishing permitted destinations from restricted ones. The Tata-ASML deal effectively reveals India's classification by demonstration, but the absence of a published framework leaves smaller, similarly aligned economies guessing.
Second, as Carnegie analyst Shruti Mittal has noted, US-India semiconductor cooperation already shows licensing asymmetries: demonstration licences "appear to be processed faster for U.S. companies entering India than for Indian companies entering the United States," and BIS is publicly under-resourced. Without Validated End-User mechanisms and routinised bilateral verification, even well-designed controls can become unnecessary friction on the legitimate trade they were never meant to block. The reciprocal worry — what Mittal calls the "second China" concern — is best addressed by transparent end-use frameworks rather than rhetorical reassurance.
The Tata-ASML MoU is the clearest single signal that the export-control regime is doing what its designers said it would: tilting the supply chain toward trusted producers without indiscriminately freezing it. New Delhi's job from here is to keep the bet inside that perimeter — by hard-coding end-use verification, IP protection, and transparent diversion controls into Dholera's operating model — rather than treating tier-2-trusted status as a permanent endowment.