While $10 billion sounds like a reasonable investment, it would not be enough to compete with global players, said Raj Kumar, Founder and Group CEO, IGSS Ventures, a tech investment holding company.
Though India was slightly late in the semiconductor game, Kumar said this could be used to the country’s advantage. “There’s one key advantage if you’re late, that is to learn from the mistakes and the positives of other semiconductor-building nations. They started earlier and there are many valuable lessons that India can learn from that. For example, India’s budget is $10 billion. I think it’s a reasonable budget to start with, but not enough to compete globally. China, for example, has in the last 30 years or so easily invested half a trillion dollars in semiconductor. The pioneers in semiconductor and IC chip design are even today spending $50-$70 billion. So, India’s $10 billion is a decent number,” he said, speaking at Semicon India 2022, which was held from April 29 to May 1 in Bengaluru.
The founder went on to explain how the country could best optimise this investment and do things in a much better way than other nations. “There is the $10 billion from the government and if we assume another $10 billion comes from the private sector, then $20 billion can now be used in the same way a $40-billion investment was used earlier.”
The country should first look at the technology space, he suggested. Semiconductor fabrication plants require massive capex. There are two groups in semiconductor technologies. “First is the digital technologies, better known as bleeding-edge technologies, which can cost $20 billion or more. Note that $20 billion or more is two times the total budget (given that earlier learnings ensure we don’t have to reinvent the wheel). We have another set of technologies which is known as more-than-Moore (MtM) technologies, which doesn’t follow Moore’s law. It has a lower starting project cost,” he said.
In simple terms, Moore’s law states that our computers would become more powerful and less costly. MtM refers to technologies that use digital and non-digital functions on the same chip.
Kumar said that in 2021, 45% of the semiconductor products were of more-than-Moore technologies. This included devices such as CMOS sensors, MEMS sensors and pacemakers. “So, with India starting late with $10 billion, one key strategy here can be any focus initially on the more-than-Moore technologies.”
Another way to best optimise the $10-billion investment would be having fabrication plants of different wafer fabs. There are more than 200 wafer fabs in the world of 200-300mm and it is important that at this stage, India participates in full-service capabilities. “It is important we do this in the first 5-10 years to have different fabrication plants. Otherwise, why would anyone come to India? We have heard knowledgeable speakers saying that India has very large design capabilities. That is true and we need to leverage that,” he said at the event.
For this to happen, his suggestion was to focus on strong global collaborations. He cited the examples of Taiwan and China where collaborations have been on with global experts for decades now. Some 30 years ago, a group of IDM semiconductor experts came to Taiwan and eventually started semiconductor manufacturer TSMC. They had a good partner in Philips, which later became an expert.
“Intel has a satellite factory in China. Don’t take this wrongly, it doesn’t mean you do not need localised support. I’m a firm believer in global collaborations. We are pushing local universities and research institutes with the subjects and also working with global universities at the same time. So, you need to have global progression as well,” he added.