Why it makes strategic sense to incentivise domestic chip research

It will offer India the opportunity to become R&D centre to the world without losing valuable IP rights rather than frittering away capital on chip-making, for which it has no known capabilities yet

The Union Cabinet on Monday approved the proposal of Kaynes Semicon Pvt Ltd to set up a semiconductor unit in Sanand, Gujarat. The government has been providing substantial capital subsidies for joint ventures in semiconductor manufacturing. After some false starts, New Delhi has approved five ventures entailing investment worth about Rs 1.5 trillion. One of these is a manufacturing unit by the Tata group with Powerchip Semiconductor Manufacturing Corporation, Taiwan’s third-largest semiconductor foundry. And four are Assembly Testing Marking and Packaging (ATMP) and Outsourced Semiconductor Assembly and Test (OSAT)  — one by Tata Semiconductor Assembly and Test (TSAT) in Assam; one by the Murugappa group’s C G Power with Tokyo-headquartered Renesas Electronics; and two in Sanand, one by US-based Micron Technologies and another, approved by the Union Cabinet on Monday, by Kaynes Semicon. The government’s capital subsidies for these amount to Rs 50,000 crore. This commitment, which could rise in the coming years, is intended to take India into the global semiconductor arena by establishing a domestic base for chip supplies. But a recent study of the nature of the domestic market for chips suggests that India may be investing in the past rather than the future.

According to a study by the India Electronics and Semiconductor Association (IESA), slated to be released soon, the market for semiconductor chips in India will change dramatically by 2032, with 60 per cent of demand, in value terms, expected to come from chips of less than 10 nanometre (nm). The demand would come from data centres, mobile phones, and computing hardware. All of these will require more memory, which will need cutting-edge, state-of-the-art chips in the 3-4 nm range. The plants coming up in India will manufacture chips in the 28-48 nm category, serving the automobile industry (especially electric vehicles) and consumer electronics. The prognosis, then, is that India will have to rely on import to service two-thirds of domestic demand in the immediate future. Some of this emerging demand could be met by ATMP and OSAT plants but the demand-shift up the value chain implies that domestic manufacturers will have to export to build profitable businesses for their legacy products. This is not necessarily a threatening proposition since the export market for legacy chips is likely to grow at a decent clip. The question is whether the heavy capital that semiconductor manufacturing plants absorb, plus meeting demanding factory standards that are difficult to meet in India — ultra-high sanitation, for instance, and a large continuous water supply — are warranted when they are focused on, essentially, yesterday’s market.

None of this should mean that India must remain a laggard in an industry that is emerging as the new oil of the future. But an opportunity may exist beyond the conventional brick and mortar in the large, high-quality domestic knowledge-based capital at the country’s disposal. This much can be seen from the speedy growth of Global Capability Centres (GCCs) set up in India by the world’s largest corporations, which derive huge value and amass intellectual property from Indian brainpower. All the world’s top chip design companies — Nvidia, AMD, Qualcomm, Intel, and Texas Instruments — have GCCs here. Given this ready-made domestic capability, it may make greater strategic sense for the government to leverage the large indigenous talent pool to incentivise domestic research and development (R&D) GCCs. This would offer India a good opportunity to become the R&D centre to the world without losing valuable IP rights, rather than frittering away capital on chip-making, for which the country has no known capabilities yet.