A Tenth of One Import Bill: What Semicon 2.0 Is Up Against

The Union Cabinet approved ₹1,27,500 crore for Semicon 2.0 on 15 July 2026 — about $13.2 billion, to be spent over six years. India spent almost $150 billion importing semiconductor products in the nine years to FY2025. The programme’s entire outlay is roughly a tenth of one past import bill, and the government’s own projections have the bill growing.

The Cabinet decision is the second phase of the India Semiconductor Mission. Semicon 1.0, approved in December 2021 with ₹76,000 crore, concentrated on fabs and on ATMP and OSAT packaging plants. Semicon 2.0 adds the layers above and below them — chip design, machines and materials, research and development, and talent, alongside more fabs and further packaging capacity. Six pillars, as the government’s own briefing sets them out. Support runs on a pari-passu basis: 30% of capex for equipment and materials, 40% for logic and memory fabs, 35% for compound semiconductor, display and ATMP/OSAT units. The scheme runs six years. The government expects it to draw about ₹4 lakh crore of investment, produce about ₹2 lakh crore of output and about ₹1 lakh crore of exports.

The arithmetic the programme is up against

A note published by the Press Information Bureau in September 2026, around the Semicon India gathering, sets out the other side of the ledger: India “spent almost USD 150 billion on semiconductor product imports during FY17–FY25”, with those imports growing at a compound annual rate of 23% over the period. If the trend holds, the note says, annual imports could reach USD 240 billion by 2035, against demand of USD 110 billion by FY2030 and above USD 200 billion by FY2035.

Three calculations make the gap legible. USD 150 billion across nine years is USD 16.67 billion a year — our division of the two published figures. At ₹96.5 to the dollar, the mid-October 2026 level, USD 150 billion is ₹14.475 lakh crore — our multiplication. And ₹14.475 lakh crore against the new programme’s ₹1.275 lakh crore is 11.35 times — our division. Converted at the same rate, the entire outlay is about USD 13.2 billion: six years of programme against one nine-year import bill eleven times its size.

That is not an argument that the money is misspent. It is the ratio in which every other question about Semicon 2.0 has to be asked.

What the first ₹76,000 crore bought

In 2021 India had no commercial chip plant. By 2026, on the government’s own account: twelve semiconductor manufacturing units approved, carrying cumulative committed investment of over ₹1.64 lakh crore — one silicon fab, one silicon carbide fab, one gallium nitride micro-LED display fab and nine ATMP/OSAT units — across Gujarat, Assam, Uttar Pradesh, Odisha, Punjab and Andhra Pradesh. Five have begun commercial production. Twenty-four chip design projects have been approved, with a total value of about ₹900 crore.

That last number is the honest measure of the design programme. India hosts about 7% of the world’s semiconductor-domain global capability centres and employs close to 20% of the global chip design workforce. Twenty-four funded projects and ₹900 crore is what the state has so far put behind the one part of the chain where India’s comparative advantage is not in question.

Where the money leaves

Electronics imports crossed the hundred-billion-dollar mark for the first time in FY2025-26, reaching USD 116.18 billion, up 17.76% from USD 98.65 billion — Commerce Ministry data. Against a total import bill of USD 774.98 billion, electronics was 14.99% of everything India bought from the world — our calculation from the two published figures.

Exports of electronic goods reached USD 47.69 billion over the same year on the same source; the government’s own briefing rounds it to USD 48 billion, and so do several reports of the release. The gap is about USD 68 billion — our subtraction of export from import — and electronics is now 10.79% of India’s merchandise exports, our calculation of USD 47.69 billion against the USD 442 billion total the government publishes, which is the same “about 11%” its own slide deck states.

Inside that import bill, CRISIL’s breakdown puts close to USD 40 billion of electronic components in FY2025-26: integrated circuits about USD 30 billion, recorded media USD 5.3 billion, semiconductors and electric accumulators USD 4.9 billion each, circuit apparatus USD 2.1 billion. Electronic products buy in 29.8% of their total supply, among the most import-intensive manufacturing industries in the country. And the supplier of first resort is unchanged — China was the source of 48.9% of India’s semiconductor imports in FY2025-26, down from roughly 64% a year earlier, as our component analysis set out.

Two documents, four months apart

In May 2026 the government published a semiconductor roadmap whose stated aim is to build a USD 120–150 billion semiconductor value chain by 2035. In September 2026, the note that accompanied Semicon 2.0 projected annual imports of USD 240 billion by 2035.

The two figures are not contradictory. A domestic value chain and an import bill are different measurements, and a country can build a large one while still buying a great deal. But they describe different countries, they were published four months apart by the same government, and reading them together is the clearest statement available of how much of the gap the current level of ambition is designed to close. That reading is ours, not the government’s.

The layer that decides it

The genuinely new money in Semicon 2.0 is aimed at machines and materials, which Semicon 1.0 did not touch at all, at 30% of capex on a pari-passu basis. The market India is buying into is one we size ourselves: the global semiconductor manufacturing equipment market was USD 118.70 billion in 2025 and is forecast to reach USD 213.85 billion by 2033, a 7.63% CAGR — the layer that captures the value when fabs are actually built, and the layer whose latest outlook we published in September. India’s entire six-year programme is 11.13% of one year of that market — our division of USD 13.2 billion by USD 118.70 billion.

None of which is a verdict on the strategy. It is the size of the entry ticket.

What to watch

The ratio. Component imports divided by electronics exports. If it is flat in 2028, the boom was an assembly boom; if it falls, India has started moving up the chain. The design count. Twenty-four funded projects is the baseline; whether the next twenty-four are companies rather than projects is the test. Whether five producing units become five producing at scale. Commercial production and volume are different milestones, and the fab support is 40% of capex precisely because the capital is heavy and the gestation long. And the political variable. India’s goods surplus sectors are already named in a live American investigation into structural excess manufacturing capacity, and India has signed the statement that keeps it alive; semiconductors appear in that statement rather than in India’s own paragraph — the two tracks are separate, and only one has produced a duty so far.

Pheonix Research covers semiconductors and electronics — components, equipment, fabrication and end markets — through the Electronics and Industrial Machinery & Automation segments, including the Global Consumer Electronics market, USD 1.18 trillion in 2025 to USD 1.92 trillion by 2033 at a 6.28% CAGR. The series continues with The Phone, Not the Commodity, The Component Test and The Map of the Boom.