U.S.-China Relations
NextFin News - Prime Minister Narendra Modi sat down with the chief executives of Micron, ASML, Intel, Applied Materials, Tata Electronics and roughly a dozen other semiconductor leaders in New Delhi on September 16, 2026, and delivered a simple pitch: India has committed more than $30 billion of public and private capital to building a chip industry from scratch, and it wants these companies to build the rest. The roundtable, held a day before Modi inaugurated SEMICON India 2026, was less a celebration than a down-payment on a bet that has moved from policy announcements to commercial production faster than most investors expected.
The arithmetic behind the headline number is straightforward. The Union Cabinet approved Semicon 2.0 on July 15 with a fiscal outlay of ₹1,27,500 crore, roughly $13.2 billion. On top of that, the India Semiconductor Mission has already attracted about ₹1.6 trillion, or approximately $17.3 billion, of committed private investment across approved projects. Put together, the public-private chip bet exceeds $30 billion. The question for investors is no longer whether India is serious about semiconductors. It is whether the money buys a supply chain that can actually compete with Taiwan, South Korea and China.
The Situation: From Policy to Production
India's semiconductor ambition has a precise starting point: the India Semiconductor Mission 1.0, approved by the Union Cabinet in December 2021 with a ₹76,000 crore incentive framework offering fiscal support of up to 50% for silicon fabs, compound semiconductor facilities, assembly and testing units, and chip design. Five years later, the programme has 10 approved projects spanning six states, with a total investment of ₹1.60 lakh crore as of December 2025, according to the Press Information Bureau. By July 2026, that had grown to 12 approved manufacturing units with cumulative investment above ₹1.64 lakh crore.
The difference between the two phases is structural. Semicon 1.0 paid companies to arrive. Semicon 2.0, announced on July 15, 2026, pays them to build everything around the fab: chip design, fabrication, display manufacturing, advanced packaging, semiconductor equipment, specialty materials, specialty chemicals, R&D, engineering services, and talent development. The modified programme for 2026–27 alone carries a financial outlay of ₹8,000 crore, with targets of supporting one additional fab (₹4,000 crore of investment), nine compound-semiconductor and OSAT units (₹11,000 crore), and 30 design companies under the Design Linked Incentive scheme.
The production milestones are now visible. On July 4, 2026, Modi inaugurated the CG Semi outsourced semiconductor assembly and test facility in Sanand, Gujarat, watching the first batch of commercially produced packaged chips roll off an Indian line. CG Semi is a ₹7,600 crore joint venture between CG Power, Japan's Renesas Electronics and Thailand's Star Microelectronics. Kaynes Semicon opened its Sanand OSAT facility on March 31 with an investment of ₹33 billion and a capacity of roughly 6 million chips a day. Micron's memory packaging plant in Sanand, a ₹22,516 crore project, has been in phase-1 validation. As of March 2026, one unit had commenced commercial production and three were on pilot lines.
The centre of gravity, though, is the Tata Electronics–PSMC fab in Dholera, Gujarat. The ₹91,000 crore, roughly $11 billion, 12-inch wafer fab targets 50,000 wafers a month on the 28-nanometre node and is aiming for its first chip in December 2026. That would make it the first plant to fabricate silicon on Indian soil rather than package imported chips. Tata and ASML signed a strategic agreement on May 16, 2026 to support the development of the facility.
At the September 16 roundtable, Modi framed the next phase in his own words:
During this evening's Semiconductor Roundtable with CEOs, I emphasised: Unequivocal support to the semiconductor industry from the Centre. Predictable and responsive policy environment. Continued focus on skilling, innovation and infrastructure upgradation. Furthering every effort towards making India a leader in emerging technologies.
The CEOs, representing SEMI, Micron, Infineon, Applied Materials, ASML, Merck, Tokyo Electron, Fujifilm, AMD, Intel, Tata Electronics, Lam Research, Rapidus, NXP, Foxconn, Semiconductor Devices and IBM Research, responded by committing to contribute to India's emergence as a globally competitive semiconductor hub. Ajit Manocha, president and chief executive of SEMI, put the ambition more sharply: "India's opportunity is not simply to join the global supply chain, but to become integral to it."
One commitment was immediate and quantified. Fujifilm announced an investment of ₹800 crore for a greenfield semiconductor materials manufacturing facility, with phase one breaking ground in fiscal 2026 and commercial production targeted for fiscal 2028. It was the kind of announcement a host government wants within hours of a leader's pitch.
Why This Time Is Different: The Mechanism Behind the Push
The obvious question is why India's chip push should work now when decades of talk produced almost nothing. The answer is that the mechanism has changed. Earlier attempts tried to subsidise fabs into existence. The current strategy subsidises the entire value chain and lets demand pull capacity online.
India is the world's second-largest smartphone manufacturer behind China, with more than 200 mobile manufacturing facilities operating in the country. Its electronics manufacturing base is large enough to absorb the output of mature-node fabs. The 28-nanometre node that Tata is building is not the leading edge — it will not produce AI accelerators or flagship mobile SoCs — but it is exactly the node that automotive chips, power management ICs, display drivers and microcontrollers require. Those are the chips India's own factories consume. Building for domestic demand first, then exporting, inverts the old export-first model that failed everywhere it was tried without a home market.
The demand numbers support the thesis. Industry estimates cited by the government place the Indian semiconductor market at about $38 billion in 2023 and $45–50 billion in 2024–25, with an expected reach of $100–110 billion by 2030. The national target is to design and manufacture chips for 70–75% of domestic applications by 2029. By 2035, India aims to be among the top semiconductor nations globally, with a roadmap to 3-nanometre and 2-nanometre technology nodes.
The second mechanism is policy continuity across a politically sensitive window. Semicon 2.0 was approved in July 2026, and the roundtable and SEMICON India 2026 followed within weeks. The government's message to investors is that the policy will survive electoral cycles and that the incentive framework is now a permanent feature of industrial policy rather than a one-off scheme. That matters because a fab is a 20-year decision, not a five-year one.
The third mechanism is the shift from "make in India" to "silicon to systems." The 2026 SEMICON theme is not accidental. India is no longer asking whether it can make chips. It is asking how fast it can build everything around them — equipment, materials, gases, chemicals, design IP, and the skilled workforce to run it all. The government has provided design tools to more than 332 academic institutes and 105 startups, and has approved 24 startups under the Design Linked Incentive scheme. Fujifilm's materials plant is the first concrete sign that the equipment-and-materials layer is starting to form.
The Cyclical Call Versus the Structural Call
Investors need to separate two forces that are being blended together. The first is cyclical: the global chip cycle is recovering from the 2023 downturn, memory prices have stabilised, and mature-node capacity is in demand because automotive and industrial customers over-ordered during the shortage and are now restocking. That leg of the story will revert. If the global cycle turns down again, India's new fabs will face the same pricing pressure as everyone else.
The second force is structural, and it is the one that matters for the $30 billion bet. The semiconductor supply chain is being reorganised along geopolitical lines. Taiwan produces the overwhelming majority of the world's advanced logic. China is building self-sufficiency under export controls. The United States, Japan, South Korea and Europe are all subsidising domestic capacity. In that environment, a large, English-speaking democracy with a real electronics market becomes a strategically useful node, not just a cheap manufacturing site. That is a regime change, not a cycle. It will not revert on its own because the driver is the reorganisation of the supply chain, not the level of demand.
The evidence for the structural call is in the policy itself. Semicon 2.0 does not just pay for fabs; it pays for equipment, materials, chemicals and design IP. Those are the layers that make an ecosystem sticky. A fab alone can be undercut. An ecosystem of suppliers, designers and trained engineers cannot be moved quickly. India's target of 70–75% domestic self-sufficiency by 2029 is a demand-side anchor that keeps capacity utilised even when the global cycle weakens.
But the structural call has a hard limit. India has no leading-edge logic fab below 7 nanometres. For AI accelerators, GPUs and advanced mobile SoCs, TSMC, Samsung and Intel Foundry remain the only options, and India will not have leading-edge logic capability before 2030 at the earliest. The bet is therefore a mature-node, packaging-and-systems bet. It wins if the world needs more mature capacity and more packaging. It loses if the value in the chain keeps migrating to the leading edge and India is left with the low-margin layers.
The Second-Order Question the Market Is Not Asking
The first-order story is obvious: subsidies bring fabs, fabs bring jobs and exports. The second-order question is what happens to the subsidy math when the first wave of projects actually starts producing. Semicon 1.0 offers up to 50% fiscal support. That is generous, but it also means the government is underwriting nearly half the capital cost of projects whose commercial viability is unproven. The real test arrives in 2027 and 2028, when Tata's Dholera fab and the other approved units are supposed to be at ramp. If yields are poor, if power and water infrastructure lags, or if global mature-node pricing collapses, the subsidy bill rises and the political appetite for the next tranche could fall.
The cross-asset transmission channel runs through the rupee and India's fiscal deficit. A $30 billion semiconductor bet is large relative to India's industrial policy budget, and it comes alongside other subsidy programmes. If the projects deliver, the current-account benefit from displaced imports and new exports strengthens the rupee over the medium term. If they stall, the fiscal cost shows up in the deficit and the rupee carries the risk premium. That is the second-order trade embedded in the headline: this is not just an industrial policy story; it is a credit story for India's public finances.
There is also a competitive transmission channel. Taiwan, South Korea and China are not standing still. Taiwan is expanding mature-node capacity overseas precisely because customers want geographic diversification. China is subsidising its own mature-node build-out at a scale India cannot match. The window for India to capture share is real but narrow. It closes if the global mature-node market becomes oversupplied before India's fabs reach scale.
The Strongest Counter-Thesis
The strongest argument against the bull case is not that India lacks talent or demand. It is that semiconductor manufacturing is a craft business where yield is everything, and yield is learned over decades, not bought with subsidies. Taiwan's dominance rests on a density of process engineers, supplier relationships and tacit knowledge that no amount of capital replicates quickly. China has spent far more than $30 billion and still cannot reliably produce leading-edge chips under export controls. On this view, India's most likely outcome is a collection of packaging plants and a single mature-node fab that struggles to compete on cost with established players once the subsidies phase out.
This counter-thesis is backed by the hard reality of the technology gap. India's roadmap to 3-nanometre and 2-nanometre nodes by 2035 is an aspiration, not a plan with a named equipment supplier and a qualified process. The gap between 28-nanometre and 3-nanometre is not a straight line; it is a series of exponentially harder problems. A sceptic would say the $30 billion buys India a seat at the mature-node table, not a position among the top semiconductor nations.
The counter-thesis is correct on the leading edge, and investors should not confuse the two. But it may underweight the packaging and mature-node segment where India actually has a chance. Advanced packaging is becoming more valuable as Moore's Law slows, and India's OSAT base is already producing. The falsifying signal for the bull case is specific: if Tata's Dholera fab does not deliver its first chip by mid-2027, or if approved projects fail to move from pilot to commercial production at the pace the government has set, the structural thesis weakens materially. Watch the 2027 production milestones, not the 2035 roadmap.
What's Next: Three Horizons
Short term (6–12 months): The market will trade on announcements. SEMICON India 2026, running September 17–19, is likely to produce more project commitments and more MoUs. Tata's Dholera first-chip target of December 2026 is the single most important near-term catalyst. A slip there would be read as a warning; a hit would validate the entire policy framework.
Medium term (2–3 years): This is where the thesis is proven or broken. The base case is that India brings one 28-nanometre fab and several OSAT facilities into commercial production, captures a meaningful share of mature-node packaging, and becomes a credible alternative supplier for automotive and industrial chips. The upside case is that the ecosystem layers — equipment, materials, design IP — form faster than expected, pulling in more foreign investment and pushing self-sufficiency toward the 70–75% target. The downside case is that yield problems, infrastructure bottlenecks or a global mature-node glut leave capacity underutilised and the subsidy bill rising.
Long term (5–10 years): The structural bet either pays off as a geopolitical supply-chain node or settles into a modest domestic-serving industry. The 2035 target of top-nation status and 3-nanometre/2-nanometre capability should be treated as directional, not as an underwritable forecast. The long-term winner in India is not the fab alone; it is the companies that supply the fab — the materials, gases, chemicals and equipment providers that sit upstream of production.
The beneficiaries are clear: the approved project sponsors (Tata Electronics, Micron, CG Semi, Kaynes), the equipment and materials suppliers that win Indian contracts (ASML, Applied Materials, Tokyo Electron, Fujifilm), and the domestic electronics manufacturers that gain a local chip supply. The exposed are the taxpayers underwriting the 50% fiscal support and the investors who price India's chip stocks as if the 2035 roadmap is already de-risked.
Modi's pitch to the CEOs was that India offers a predictable policy environment and unequivocal support. The investors' reply will come in capital commitments over the next 12 months. The $30 billion bet is already on the table. What happens next is whether the chips follow the money.
India is betting $30 billion not that it can make chips, but that it can build everything around them — and that is the difference between a subsidy and an ecosystem.
Explore more exclusive insights at nextfin.ai.
