India has committed over $10 billion, approximately Rs 76,000 crore, to build a domestic semiconductor industry through the India Semiconductor Mission (ISM) and its linked Production Linked Incentive (PLI) scheme. Four projects have received approval: Tata Electronics’ fab partnership with Taiwan’s PSMC in Dholera, Micron Technology’s assembly, test, marking and packaging (ATMP) facility in Sanand, CG Power’s ATMP unit backed by Japan’s Renesas and Thailand’s Stars Microelectronics, and Kaynes Semicon’s ATMP facility in Sanand. A fifth, the much-publicised Vedanta-Foxconn joint venture, collapsed in 2023, exposing the fragility of political commitments in deep-tech industrial policy. Whether the remaining projects will deliver India’s semiconductor self-reliance ambition, or merely add packaging capacity at the bottom of a global supply chain, depends on execution timelines, talent availability, and a realistic understanding of where India can compete in the decade ahead.
Why Semiconductors, Why Now
The global semiconductor shortage of 2020-2022, which idled automobile assembly lines, delayed electronics shipments, and revealed supply chain vulnerabilities across every advanced economy, permanently altered how governments think about chip manufacturing. What had been treated as a commercial activity efficiently handled by Taiwan Semiconductor Manufacturing Company (TSMC), Samsung, and a handful of other specialised fabs became a strategic national security issue overnight. The United States passed the CHIPS and Science Act (August 2022), committing $52.7 billion to domestic semiconductor manufacturing and R&D. The European Union launched the European Chips Act with a €43 billion target. Japan built a formal partnership with TSMC to establish a domestic fab. South Korea announced $450 billion in semiconductor investments over the decade.
India had compelling reasons to act. The country imports semiconductors worth approximately $24 billion annually, a figure projected to reach $80-100 billion by 2030 as electronics manufacturing scales up under the PLI scheme and domestic demand grows with rising incomes, electrification, and digital infrastructure expansion. Every smartphone, automobile, appliance, medical device, and defence system contains chips that India currently purchases entirely from abroad. Building even a partial domestic supply base reduces foreign exchange outflow, creates high-skill employment, and positions India within the global supply chain in a way that pure assembly manufacturing cannot. The India Semiconductor Mission, launched in December 2021 under the Ministry of Electronics and Information Technology (MeitY), was the government’s response to this strategic gap.
The Four Approved Projects: What Is Actually Being Built
Understanding what India has committed to requires distinguishing between different stages of semiconductor manufacturing. A complete semiconductor supply chain runs from raw silicon and speciality chemicals through wafer fabrication (where transistors are etched onto silicon wafers using photolithography), to packaging (where individual chips are cut from wafers, tested, and encapsulated), to final assembly into consumer products. The technically demanding, capital-intensive, and strategically critical stage is wafer fabrication, the fab. ATMP (assembly, test, marking and packaging) sits downstream of fabrication and is less technically complex, though still valuable.
Tata Electronics-PSMC, Dholera, Gujarat, The most significant of the four approvals, this is India’s first serious attempt at a wafer fabrication facility. Tata Electronics partnered with Powerchip Semiconductor Manufacturing Corp (PSMC) of Taiwan, which will provide technology licensing and technical assistance. The fab will manufacture chips at 28nm and above process nodes, mature technology by global standards (TSMC’s leading edge is at 2nm), but relevant for automotive electronics, industrial controls, power management, and display drivers. The facility is being built at Dholera Special Investment Region in Gujarat, with a projected investment of approximately $11 billion and commercial production targeted for 2026-2027. Capacity at full ramp is projected at approximately 50,000 wafers per month.
Micron Technology, Sanand, Gujarat, Micron’s facility is an ATMP plant for DRAM and NAND memory chips. Micron designs and manufactures its chips offshore (primarily in Japan, Singapore, and Taiwan) and will ship wafers to Sanand for final packaging and testing. The investment is approximately $2.75 billion, with government incentives of around $1.35 billion. This facility does not involve wafer fabrication and does not give India the ability to design or manufacture chip architecture independently, it adds India to Micron’s global packaging footprint. Nevertheless, it is operational and represents real transfer of process technology and employment in a high-precision manufacturing environment.
CG Power + Renesas + Stars Microelectronics, Sanand, This ATMP facility will focus on power semiconductors and microcontrollers supplied by Japan’s Renesas, a major global player in automotive and industrial chips. CG Power brings Indian industrial manufacturing experience; Stars Microelectronics brings ATMP process expertise. Investment is approximately $740 million. Renesas’s involvement is strategically significant because automotive and industrial semiconductor demand in India is growing rapidly as the country’s automobile sector electrifies and Industry 4.0 adoption accelerates.
Kaynes Semicon, Sanand, The fourth approved project is an ATMP facility by Kaynes Technology, an Indian electronics manufacturing services company. Investment is approximately $350 million. Kaynes brings domestic EMS experience; this project extends India’s ATMP footprint and provides a domestically-owned capacity node.
| Project | Type | Location | Investment | Status |
|---|---|---|---|---|
| Tata-PSMC | Wafer Fab (28nm+) | Dholera, Gujarat | ~$11 billion | Construction underway; production target 2026-27 |
| Micron Technology | ATMP (Memory) | Sanand, Gujarat | ~$2.75 billion | Facility inaugurated Sep 2024; production ramping |
| CG Power + Renesas | ATMP (Automotive/Industrial) | Sanand, Gujarat | ~$740 million | Under construction |
| Kaynes Semicon | ATMP | Sanand, Gujarat | ~$350 million | Under construction |
The Vedanta-Foxconn Collapse: A Warning About Industrial Policy
The most instructive episode in India’s semiconductor journey so far is not a success but a failure. In February 2022, Vedanta, the Indian mining and metals conglomerate controlled by Anil Agarwal, signed a joint venture with Foxconn, the Taiwanese contract manufacturer, to build a $19.5 billion semiconductor fab in Gujarat. The announcement generated enormous political attention and was positioned as a transformational moment for India’s technology ambitions. The Prime Minister attended the Gujarat groundbreaking ceremony. The project received environmental clearances and land allocation.
In July 2023, Foxconn withdrew from the joint venture, and the project collapsed. The stated reason was that the two companies could not agree on technology transfer and governance terms. The underlying reality was more structural: Vedanta had no semiconductor manufacturing experience, had not secured a technology partner to provide the fabrication process (unlike Tata-PSMC, which has PSMC), and was attempting to build a leading-edge fab without the technical foundation to operate one. Foxconn, primarily an assembly and packaging company rather than a wafer fabricator, recognised it could not supply the technology Vedanta needed and would have been exposed to liability for a project it could not execute. The collapse revealed the gap between India’s capacity to attract commitments and its capacity to execute them, a gap that the ISM’s design (requiring technology partners and credible execution plans before approval) was intended to close.
You can build a semiconductor factory with money. You cannot buy the 40 years of process engineering knowledge embedded in TSMC or Samsung. That knowledge transfers through people, partnerships, and decades of trial, not through project approvals and ground-breaking ceremonies.
Why 28nm Matters, and Why 7nm Won’t Happen This Decade
The technology node, measured in nanometres, is the single most consequential variable in a semiconductor fab’s competitive position. Leading-edge nodes (3nm, 5nm, 7nm) are manufactured exclusively by TSMC and Samsung and require extreme ultraviolet (EUV) lithography machines made by ASML of the Netherlands, which cost approximately $200 million each and involve supply chains of extraordinary complexity. These nodes power the most advanced processors: Apple’s M-series chips, Qualcomm’s Snapdragon 8-series, NVIDIA’s AI training chips (H100, H200, Blackwell). They are strategically controlled, the United States has pressured the Netherlands to restrict EUV machine exports to China, and these restrictions have materially slowed China’s semiconductor programme.
India will not access EUV machines in this decade. The technology is under strict export controls, and even if those controls did not exist, the capital investment and process engineering expertise required to operate an EUV-based fab are beyond India’s current capacity. The Tata-PSMC fab at 28nm uses deep ultraviolet (DUV) lithography, which is mature technology with no export restrictions. This is a deliberate and correct strategic choice. The 28nm node, while 15+ years old from a design vintage perspective, remains deeply relevant for a large share of global semiconductor demand, automotive (engine management, ADAS sensors), industrial controls, power management ICs, microcontrollers, IoT devices, RF chips, and display drivers all run on mature nodes where cost per wafer matters more than transistor density. This market is large, structurally stable, and not concentrated in the hands of the two leading-edge fabs that control the advanced node market.
The question is not whether 28nm is inferior to 3nm, it obviously is, for compute-intensive applications. The question is whether India can build a commercially viable and scalable fab business in the segment where it can actually compete. On that question, the answer is a qualified yes, provided the Tata-PSMC execution stays on track and the yield rates achievable with PSMC’s process technology are competitive with the global alternatives (primarily TSMC’s mature node capacity and GlobalFoundries).
The Talent Gap: India vs Taiwan and South Korea
Building a semiconductor fab is not primarily a capital problem, it is a human capital problem. The Tata-PSMC fab, when operational at full capacity, will require approximately 4,000-5,000 direct employees, of whom a significant share must be process engineers, fab technicians, equipment engineers, and materials specialists with experience in semiconductor manufacturing environments. India does not currently have this workforce at scale.
Taiwan’s semiconductor workforce advantage is structural and generational. TSMC, founded in 1987, has been running fabs for nearly 40 years. The country has trained engineers specifically for semiconductor manufacturing through dedicated programmes at National Taiwan University, National Chiao Tung University (now NYCU), and a system of vocational institutes that feed directly into fab operations. Approximately 300,000 people work directly in Taiwan’s semiconductor industry; the cluster effect, suppliers, equipment vendors, process specialists, multiplies this workforce by several times. South Korea’s Samsung and SK Hynix have built comparable depth over the same period. These countries have engineers who have spent entire careers inside fabs, accumulated process knowledge that is not in textbooks, and developed the institutional culture of continuous incremental yield improvement that semiconductor manufacturing requires.
India’s semiconductor education and training pipeline is near zero. Indian engineering colleges produce large numbers of electronics and electrical engineers, but virtually none with semiconductor process engineering specialisation. The IITs have limited semiconductor research infrastructure, the Indian Institute of Science Bangalore (IISc) has a National Nanofabrication Centre (NNfC) and IIT Bombay has a Nanofabrication Facility, but these are research tools, not production environments. MeitY has announced plans for semiconductor lab infrastructure at multiple institutions, and the Tata-PSMC project includes provisions for workforce development, but translating institutional announcements into a pipeline of trained fab engineers will take 5-10 years at minimum.
The near-term solution will be heavy reliance on PSMC process engineers, expatriate specialists, and diaspora talent repatriation. This is a workable bridge but not a scalable long-term answer. India’s semiconductor ambition requires a dedicated national programme in semiconductor engineering education, similar to what IIT’s founding represented for software engineering, that does not currently exist at the required scale. For comparison, look at how Vietnam’s manufacturing success, examined in our analysis of Vietnam’s Doi Moi manufacturing transformation, was built on deliberate workforce development over decades, not just on foreign direct investment.
India vs the World: The Semiconductor Comparison
Placing India’s programme in global context reveals both the ambition and the gap. Taiwan’s semiconductor sector contributes approximately 15% of GDP, employs 300,000+ people directly, and manufactures over 90% of the world’s most advanced chips through TSMC. South Korea’s Samsung and SK Hynix together account for roughly 60% of global DRAM production. China, despite US restrictions, has invested over $150 billion in its domestic semiconductor programme since 2014 and now has a credible mature-node fab industry through SMIC, though its advanced-node ambitions remain constrained by EUV restrictions. The United States, through the CHIPS Act, is spending $52.7 billion over five years specifically to rebuild domestic leading-edge manufacturing capacity, a recognition that even the world’s largest economy allowed itself to become dangerously dependent on concentrated offshore production.
India starts from essentially zero in wafer fabrication. The $10+ billion ISM commitment is large by Indian government standards but modest compared to the capital intensity of the global competition. A single leading-edge TSMC fab costs $20-30 billion; TSMC’s planned Arizona fabs represent a $65 billion total commitment. India’s entire semiconductor programme is roughly comparable to one mid-size TSMC facility. The appropriate comparison is not with Taiwan or Korea, where decades of compounding investment have created deep competitive moats, but with countries at a similar starting position: Israel (world-class in fabless design, minimal fab capacity), Malaysia (strong ATMP, growing fab ambitions), and Vietnam (currently ATMP-focused with long-term fab aspirations).
- Taiwan: 90%+ of advanced chip production; TSMC at 3nm and beyond; 300,000+ semiconductor workforce
- South Korea: Samsung + SK Hynix control global memory market; leading edge fab capability at 3nm
- China: $150B+ invested since 2014; SMIC at 14nm (mature); blocked from EUV/advanced nodes
- USA: $52.7B CHIPS Act; Intel, TSMC Arizona, Samsung Texas fabs being built
- India: First fab (28nm) under construction; 3 ATMP facilities coming online; $10B committed
The Realistic Timeline: What India Can Achieve by 2030
A realistic assessment of what India’s semiconductor programme will deliver by 2030 requires separating early wins from structural limitations. On the positive side: Micron’s Sanand ATMP is operational and will generate real output, employment, and process learning. The CG Power-Renesas facility, once operational, will supply a segment of India’s automotive semiconductor demand domestically, reducing import dependency in a high-growth segment. The Tata-PSMC fab, if construction and ramp proceed on schedule, will be manufacturing 28nm chips by 2027-2028 and could reach meaningful scale by 2030.
What India will not have by 2030: leading-edge (sub-7nm) fabrication capability; an indigenous chip design ecosystem capable of designing complex SoCs (system-on-chip) for smartphones or AI accelerators; a domestic photoresist, specialty gas, or semiconductor materials industry of scale; or a workforce deep enough to staff a second generation of fabs without continued heavy dependence on foreign technical expertise. These are structural gaps that take a generation, not a five-year plan, to close.
The right frame for India’s semiconductor mission is not “matching Taiwan”, a goal that is neither achievable in a decade nor strategically necessary. The right frame is: can India build a viable, growing domestic semiconductor manufacturing base that (1) reduces critical import dependency in strategic segments, (2) creates a platform for deeper integration into global supply chains, (3) generates the workforce and process knowledge that compound over time, and (4) positions India to participate in the next generation of semiconductor growth driven by automotive, industrial IoT, and AI inference chips, where mature nodes remain competitive? On those terms, the ISM programme, if executed, represents a credible foundation rather than a vanity project.
What Citizens and Policymakers Can Do
Five Levers That Matter
India’s semiconductor ambition will not be determined primarily in government offices or fab construction sites. It will be determined by the quality of engineering education, the depth of the research ecosystem, and the business environment for high-precision manufacturing. Five levers that citizens, educators, and policymakers can push:
- Semiconductor education pipeline: Advocate for dedicated semiconductor engineering programmes at IITs, NITs, and regional engineering colleges, curriculum designed around fab process engineering, not just chip design.
- R&D investment accountability: Track MeitY’s semiconductor R&D commitments against actual disbursements. India’s science missions consistently announce budgets and underspend; semiconductor R&D requires multi-year sustained funding, not annual budget negotiations.
- Fabless design ecosystem: India’s strongest near-term semiconductor opportunity is fabless chip design, designing chips that are manufactured offshore. Companies like InCore Semiconductors and Signalchip are early movers; the ecosystem needs more. Support domestic fabless design through IP protection, procurement preferences for domestic designs in government electronics, and IIT research spinout infrastructure.
- Supply chain localisation: Semiconductor fabs require ultra-pure chemicals, speciality gases, precision equipment, and industrial utilities at specifications far above general manufacturing. Developing domestic suppliers for these inputs is a decade-long programme that needs to start now.
- Talent repatriation: Thousands of Indian semiconductor engineers work at TSMC, Intel, Qualcomm, Micron, and Samsung abroad. The ISM should be explicit about the conditions, compensation, research environment, career prospects, that would bring a fraction of this diaspora back to lead India’s domestic programme.
India’s $10 billion semiconductor bet is the right call for a country that imports $24 billion in chips annually and is building towards $80 billion in annual demand by 2030. The gap between the bet and the outcome will be closed, or not, by the quality of execution in the years ahead. The investment is committed. The execution is the question.
What You Can Do at Each Level
Personal: If you are an engineering student or recent graduate, consider semiconductor process engineering as a career path – the Tata-PSMC facility will need thousands of trained engineers by 2027 and the field offers internationally competitive salaries with state support. RWA/community: Connect your resident welfare association or local professional network with MeitY’s India Semiconductor Mission portal; community-level awareness of ATMP career pathways in Gujarat can redirect talent into a high-growth sector. Ward/city: Engage your local municipal body on industrial land-use planning: the Dholera Special Investment Region model of dedicated semiconductor infrastructure zones is replicable in other cities and requires civic advocacy to prioritise. National: Write to your MP or the Parliamentary Standing Committee on Science and Technology demanding an annual public audit of ISM disbursements versus committed investments – large industrial missions in India historically underspend their allocations without public accountability mechanisms.
The Bigger Picture: Technology Sovereignty
India’s semiconductor programme sits within a broader strategic shift toward technology sovereignty that is reshaping industrial policy globally. The COVID-era supply chain crises, the US-China technology decoupling, and the growing use of technology access as a geopolitical lever have collectively convinced governments that strategic dependencies in critical technology supply chains are unacceptable risks. Semiconductors are the most acute version of this dependency, the chips that power every piece of advanced military, communications, and computing hardware. A country that cannot manufacture semiconductors domestically is, at some level of strategic analysis, dependent on the foreign policy decisions of the countries that can.
India’s approach, building mature-node fabrication capacity, developing ATMP capabilities across multiple segments, and investing in design ecosystem development, is a credible path toward reducing the most acute dependencies, even if it cannot achieve complete semiconductor self-sufficiency. This is similar to how India’s space programme, examined in our analysis of India’s Quantum Mission, built strategic capability through sustained institutional investment rather than attempting to match American or Soviet resources directly. The semiconductor mission, if it avoids the execution failures that derailed Vedanta-Foxconn and sustains its funding commitment through political cycles, a challenge that mirrors how Israel built world-leading R&D density from a resource-poor starting point, has the potential to be a comparable achievement, a foundation laid in this decade that compounds into genuine industrial strength in the next.