With 100 fab plants globally, India needs a bigger share: SEMI's Manocha

With over 500 companies from across the globe participating in SEMICON India, beginning on September 2 in New Delhi, Ajit Manocha — President and CEO of California-headquartered Semiconductor Equipment & Material International (SEMI), the world’s largest industry association — outlined India’s role in the global semiconductor race in a video interview with Surajeet Das Gupta. Edited excerpts:

 

With a large number of fabs being built across the world, where is India positioned in the global semiconductor sweepstakes?

 

The industry is going through a massive inflection point. Over the past decade, we saw strong growth in this industry because of the IoT (internet of things). But the growth potential is now huge. Last year, industry revenues were close to $600 billion. We project this will double in the next 6–7 years to $1.2 trillion, fuelled by the AI revolution — 50 per cent of the chips being designed today are AI-enabled.

 

How many new semiconductor plants are coming up, and where does India fit in?

 

As we speak, 107 new fabs are set to come on board across the world; 91 already work in progress, including the Tata project in India. However, these fabs will, at most, increase overall semiconductor revenues to three-quarters of a trillion dollars ($750 billion). We should expect another 30–40 fabs to reach $1.2 trillion in revenues by 2028–2030. The growth potential is huge: Around 15 fabs are coming up in the US, nearly 10 in Europe, and the rest in Asia, with China taking a major share.

 

So where do you see India?

 

The world now realises that semiconductors are not just important for national security, but also for economic prosperity. India has made a good start by clearing multiple assembly and test plants, along with one mega fab by the Tatas. India’s trajectory should be much higher, and I would set an aspirational goal of 10 fabs by 2030. 

 

With 100 new fab plants coming up in the world, we (India) need to have a bigger share. It’s not a question of if, but when.

 

India has cleared many OSAT (outsourced semiconductor assembly and test) and ATMP(assembly, testing, marking, and packaging) projects but only one mega fab. Is that the right approach?

 

If you have a fab plant but no assembly and test facilities, it doesn’t serve the purpose. First you produce wafers, then you need assembly and testing. Assembly and test facilities are much easier to set up — in terms of both investment and complexity — compared to wafer fabs. It takes extensive planning and a lot of money ($5–10 billion) to commission a new wafer factory. So wafer fabs will come over time, depending on entrepreneurs’ appetite in the country.

 

Is it necessary — or even possible — to accelerate efforts to get more fabs in India?

 

If you look at China, Taiwan, and South Korea, they didn’t emerge as fab countries in just 4-5 years. It was a multi-year effort. You need an ecosystem to support such plants.  Once the ecosystem was in place, even more companies started moving into 

Singapore. The same will happen in India. We need to execute on what we’ve started and demonstrate our capability to the world.

 

India’s semiconductor demand is expected to cross $100 billion by 2030. What percentage of this will be met domestically by fabs and OSAT plants?

 

My estimate is that by 2030, if India produces $30 billion worth of chips domestically and imports the remaining $70 billion, that will be a good balance. You can never produce every single chip required within the country.