India’s Chip Ambitions Need an Engineering Renaissance, Not Just a Design Boom

Story

India’s Chip Ambitions Need an Engineering Renaissance, Not Just a Design Boom

India’s semiconductor push is often narrated as a talent story: the country already supplies a fifth of the world’s chip-design engineers, so surely fabrication is just one more policy nudge away. That framing understates the problem. Design is knowledge work that travels well across borders and requires little more than a workstation and software licenses. Fabrication is not. It is an intensely physical discipline, comprising vacuum systems, plasma chemistry, precision metrology, contamination control and several other domains, that only exists where decades of accumulated, tacit, shop-floor knowledge have been laid down. India has world-class chip designers precisely because its post-liberalization growth model rewarded computationally intensive, capital-light, export-oriented services. That same model left the physical-engineering trades (process, materials, equipment, manufacturing systems) comparatively starved of talent, capital and institutional prestige. The semiconductor mission has now made that imbalance impossible to ignore.

Where the mission stands

By mid-2026, India’s semiconductor push has visible momentum. Under the roughly Rs.76,000-crore India Semiconductor Mission (ISM 1.0), 12 projects have been approved with combined investment exceeding Rs.1.6 lakh crore, including the Tata-PSMC wafer fab at Dholera, Micron’s assembly and test facility at Sanand, and OSAT units from CG Power, Kaynes and Tata Electronics. Commercial production has already begun at several of these ATMP and packaging units, even as the first true wafer fab remains years from commissioning. A second phase of the mission, ISM 2.0, has been floated to widen fiscal support and pull in equipment and materials suppliers, precisely because early experience has shown that approved capital and operational capacity are not the same thing: disbursement has lagged approvals, and the domestic supply chain of gases, chemicals, substrates and tools that any fab depends on remains thin.

This gap between announced capacity and functioning ecosystem is not primarily a financing problem. It is a knowledge-and-people problem. A fab is not a factory that runs once installed; it is a facility that has to be continuously operated, debugged and improved by people who understand vacuum systems, thin-film deposition, statistical process control and materials chemistry at a visceral level, that is, knowledge acquired only through years of hands-on work, not coursework.

India’s engineering-education system expanded dramatically over the past three decades, but its expansion became increasingly uneven. During the 2010s, computer engineering and related computing fields captured a disproportionate share of student demand, while traditional branches such as mechanical, electrical, chemical and metallurgical engineering remained substantial but generally experienced much less growth. The resulting imbalance reflected, among other factors, differences in student demand, industry opportunities and graduate earnings, while investment in laboratories, research capacity and industry linkages varied considerably across institutions and disciplines.

The Missing Middle: Why the market alone won’t fix this

This is a coordination failure, not simply a mistaken policy choice. A student choosing computer science over metallurgy or process engineering is behaving rationally: software offers higher starting salaries, faster international mobility and dramatically cheaper training infrastructure: a computer lab serves hundreds of students cheaply, while a cleanroom or materials characterization facility is capital-intensive and slow to scale. Weak industrial demand discourages students from specializing in physical-engineering trades; the resulting shortage of specialized graduates then discourages manufacturing investment, which reinforces weak demand. Breaking this loop cannot be done through moral appeals to study “core” engineering. It requires making industrial careers genuinely competitive, in pay, working conditions, research opportunity and career mobility, while simultaneously building the laboratories, pilot lines and industry partnerships that make specialized engineering degrees valuable in the first place.

What the semiconductor industry actually needs

Global benchmarks are instructive here. Roughly 60% of new US semiconductor manufacturing jobs are expected to require something short of a four-year degree (technicians, equipment operators, maintenance specialists) rather than design engineers. They generally require post-secondary technical training, certificates, or two-year degrees, rather than necessarily a bachelor’s degree. Taiwan’s TSMC, similarly, employs thousands of technicians alongside its professional engineering staff, and its university partnerships deliberately span electronics, materials, chemistry, mechanical and chemical engineering rather than concentrating in computer science. TSMC’s workforce data explicitly distinguishes technicians from professionals. For example, its 2019 workforce structure lists 17,160 technicians and 24,416 professionals. The occupational reality of chip manufacturing is a pyramid, not a spike: process engineers, equipment engineers, materials scientists, reliability engineers, industrial-automation specialists and, in large numbers, skilled technicians who keep tools running. India’s policy response risks recreating its old imbalance in miniature if every new “semiconductor” program is simply a design or AI-adjacent curriculum grafted onto CSE departments, while the far less glamorous disciplines, e.g., vacuum engineering, plasma physics, thin-film science, contamination control, precision machining, remain neglected.

The institutional task ahead

Programs like Chips-to-Startup, which has pushed electronics design automation (EDA) tools and FPGA infrastructure into well over two hundred institutions, are valuable but address the incremental question of adding semiconductor content to existing departments. The harder task is structural: building applied-engineering universities, regional colleges and advanced technical institutes that are geographically and institutionally tied to actual fabs, OSATs, equipment makers and chemical suppliers, so that laboratories teach real industrial problems rather than simulated ones, and so that a technician training pipeline exists alongside the engineering-degree pipeline.

India’s own workforce analyses have long identified a mismatch between the skills produced by its education and training system and the specialized requirements of electronics manufacturing. Government-backed sector studies have highlighted technical competency gaps among new entrants and curriculum deficiencies in ITIs, particularly for niche manufacturing, supervisory and multi-skilled roles. More recently, India has responded by revising ITI curricula and introducing new-age trades, including semiconductor-related training, while overall ITI enrollment has increased.

None of this argues for suppressing computer science, which is itself becoming indispensable to modern fabs through automation, digital twins and process-control software. The correction needed is compositional: treating semiconductor manufacturing as an integrated system spanning design, device physics, process engineering, materials science, equipment engineering, packaging, manufacturing systems and skilled trades, and funding that system as a mission rather than as isolated departmental grants. India’s current aggregate R&D intensity, still under 0.9% of GDP (up from 0.64% of GDP, as per the Economic Survey 2025-26), will not stretch to cover this on its own; the fabs themselves, once operating, have to become the primary schools of industrial knowledge, absorbing engineers, training suppliers, and feeding real problems back into universities.

The country that built a globally competitive software industry without first building a manufacturing base achieved something rare. But that achievement should not be mistaken for having already built a complete technological, “full-stack” economy. Chips are forcing the distinction into the open. Whether India’s fabs become durable sources of accumulated industrial knowledge, or remain capital-intensive islands dependent on imported equipment, materials and expertise, will depend less on how many chip-design engineers it produces than on whether it can finally build the unglamorous, physically grounded engineering base that chip manufacturing has always required.