India’s semiconductor mission may need a compass

3D rendering of cyberpunk AI. Circuit board. technology background. Central computer processor CPU and GPU concept. Motherboard digital chip. Tech Science background. , Photo credit: Getty Images/iStockphoto

The United States Department of Commerce and its Indian counterpart recently concluded a memorandum of understanding in March 2023 to ensure that subsidies offered by each country do not come in the way of India’s semiconductor dreams, as announced in December. Backed by the much touted Semiconductor Policy launched in 2021. The US State Department has also worked with India to strengthen sector-specific export control laws in the semiconductor space – as India has agreed to in recent media reports.

While these appear to be pre-conditions in a long dance by bureaucrats on both sides, are they enough to convince a global major like Intel to invest in India to set up a greenfield 300mm wafer fabrication plant costing over $10 billion? Will you inspire? The premise of this question is based on an ongoing two-year dance between Intel and Indian government ministers, who were seen dating Intel CEOs in Davos and New Delhi, hoping that Intel Foundry Services (Intel’s A business unit within Intel formed after its acquisition) US-Israel based Tower Jazz Specialty Foundry) will build a greenfield plant in Dholera, Gujarat.

Facts on the ground do not support a position where Intel is busy setting up fabs inside the US, so where does Indian semiconductor policy need to go from here?

The Semi-Conductor Laboratory (SCL) was set up at Mohali in 1983 by the then Central Government, with a view to create an electronics ecosystem in an era when Keltron, Uptron and Webel were consumer-oriented fledglings in pre-liberalised India . Electronics. However, setbacks, the opening of the market for consumer goods in 1991 and the SCL fire in 1989 dashed these hopes. Some funding was received from the central exchequer to revive the plant into a 180 nm node pilot line to meet the country’s strategic needs, but the facility remains an unfinished dream in its mission to build a domestic semiconductor ecosystem. , SCL Mohali can be viewed as a technology stack similar to others such as Aadhaar, Aarogya Setu and Unified Payments Interface (UPI), which act as a force multiplier effect, to consider designing in India. Encourages many integrated circuit design startups in India.

road ahead

The institutional framework for such a change in focus is already in place with the transfer of SCL back to the Ministry of Electronics and Information Technology (MeITy) after a 15-year stint as a laboratory within the Department of Space as part of the new semiconductor policy. is present. Announcement in December 2021. However, a year after such announcement, no JV partner has been found, leaving SCL employees in limbo. During this period, it seems that the focus of MeitY is to attract Intel to India to set up a fab. This can be judged from the wording of the request for proposals and the indication of MeitY ministers. However, Intel primarily works on the <22nm node and 300mm, which requires SCL to exceed $10 billion in upgrade costs.

An alternative approach may be to take advantage of the human and capital assets that exist in SCL, in a targeted manner, to jumpstart the semiconductor mission by leveraging recent technological breakthroughs in a class of semiconductors requiring advanced lithography equipment. Not there. “More than Moore” segment of the >180 nm node connected to mixed signal analog (BCD and CGE), wide bandgap (GaN, GaAs, silicon carbide) for RF and power markets by leveraging existing lithography capability already in place at SCL . In this scenario, an investment of $50-$100 million could result in the development of Indian solutions for automotive electronics (EV traction inverters/on board chargers), PV-inverters, 5G infra-power amplifiers, railway electronics (traction inverters). Indian counterparts of Bosch, Siemens, ABB, Mitsubishi Electric, Thales and ELTA.

However, upgrades aimed at fabless design houses with proven designs (sales of >$100 million per year) willing to manufacture in SCL in the 180 nm+ node should be supported by subsidies (and possibly transfer process intellectual patents if they have any). The subsidy should be aimed at global design companies aimed at India-specific markets such as BLDC fans or motor drives for e-bike chargers. Unfortunately, existing DLI/PLI schemes offer no such incentive to proven global fabless design companies.

Recent efforts by the India Semiconductor Mission to open up subsidies for global small and medium-sized enterprises in the upstream supply chain are welcome as existing facilities like SCL would benefit from it. But this in itself is not sufficient unless coupled with the incentives defined above and also targeted upgrades at different sets of players.

Read this also | PM Modi meets NXP CEO; discusses the transformative landscape in the world of semiconductors

The stakes are high as a lack of clarity and inaction could see India missing the semiconductor fabrication bus altogether, however, unless there is an improvement on the stimulus targets. Finally, to execute on this vision over the next five years, SCL needs a full-time director with “more than Moore” foundry experience than a career scientist at the Department of Space, as is the case now. This is because there is a multidimensional market that needs to be served.

Saurabh Datta Chowdhary is currently working at Power Integrations (NASDAQ: POWI), and has over 30 years of experience in semiconductor manufacturing. Views expressed are personal