India needs pathway for its semiconductor goals

SOON after taking over as the Union Minister for Heavy Industries and Steel, HD Kumaraswamy made a statement while addressing workers of his party, Janata Dal (Secular), in Bengaluru. Referring to subsidies being granted to Indian and foreign semiconductor companies, he said, “One company has invested $2.75 billion. The state and Central governments are giving a subsidy of 70 per cent to the company. How many jobs will they create? 5,000. For the 5,000 jobs that they will create, the subsidy that will be given by the state and the Centre is $2 billion. It means for every job they are creating we are providing them a subsidy of Rs 3.2 crore.”

The statement embarrassed the Central Government because the subsidy-driven semiconductor programme, euphemistically called the Production-Linked Incentive, is a pet scheme of the Prime Minister. Kumaraswamy retracted his statement the next day.

Subsidies, however, are an important element of the India Semiconductor Mission (ISM), which was launched in 2021. Speaking on the same subject, Electronics and Information Technology Secretary S Krishnan mentioned last week that Rs 70,000 crore out of the planned Rs 76,000-crore financial incentive allotted for the development of the semiconductor industry had already been committed. While the Central Government provides a subsidy of 50 per cent to companies starting operations in the ‘semiconductor ecosystem’, some states are offering top-up subsidies. This way, almost 75 per cent of the cost of setting up a unit under the ISM is subsidised. The Indian initiative, according to Krishnan, is ‘one of the most generous’ globally. Kumaraswamy mentioned the same figures, but he looked at them from the prism of job creation, which he felt was not commensurate with the investment required. He perhaps ignored the fact that the employment generation criterion is not a factor in semiconductor manufacturing because it is highly automated.

The ISM was launched to boost semiconductor manufacturing when the world suffered a short supply of different types of chips due to disruption in manufacturing in Taiwan, Taipei and China during the Covid-19 pandemic. The disruption hit the manufacturing of cars, computers, mobile phones and a range of consumer goods globally. Everyone realised that dependence on one source for supplies was dangerous. That’s when the US and Europe launched programmes to diversify the semiconductor supply chains. The ISM, too, was a response to the situation. Semiconductor chips go into everything — from computers and cars to missiles and battle tanks. They are a key driver of the global economy, much like oil. All of us consume them in some way or the other, and continued dependence on one or two sources of their supply is not sustainable. This makes semiconductor manufacturing a strategic area.

The challenge with semiconductor manufacturing is manifold — it is knowledge-driven, meaning it requires fundamental research; it is technology-driven, needs highly sophisticated machines for production and a special environment to operate them; and it is hugely capital-intensive. On top of that, chip manufacturing requires massive amounts of water and electricity. The daily water requirement of a mid-sized fab (as a chip manufacturing unit is called) could be as much as the whole year’s supply for a town or a small city. The water should be ultrapure — way beyond the drinking water’s purity and quality. This industry is a power guzzler too. Environmental issues, therefore, are as important as fundamental research, technology and investment when it comes to chip manufacturing.

The IT secretary wants more funds from the Finance Ministry for the ISM as the Rs 76,000 crore allotted has been almost exhausted. At this stage, we don’t how much more subsidies are necessary and for how long, and for what purpose in terms of any defined outcome after five or 10 years. The US, for instance, has committed $50 billion in government investment for the semiconductor sector and has taken a long-term view with well-defined goals and landmarks.

According to Dr Rahul Sud, a US-based semiconductor industry veteran, India would require billions of dollars of investment if it wishes to make a dent in the chip industry, which will be a trillion-dollar industry by 2030. The investment should be focused on achieving leadership in a niche area. The third phase of the China Integrated Circuit Fund, announced recently, envisages government investment worth $47 billion to boost manufacturing in identified segments like flash memory and dynamic random-access memory. The current market for these two types of memory chips is worth $200 billion and is projected to be $500 billion in a few years.

In contrast, the subsidies under the ISM have gone to companies in the segments which are considered the low-end of the chip market — outsourced semiconductor assembly and test or OSAT and not in high-end silicon fabs. A semiconductor fab manufactures high-end chips, while OSAT vendors are engaged in post-fabrication work. They cut the manufactured wafers into chips, package them and do the testing. All this work too needs a certain level of sophistication and investment, but it is not in the same league as the cutting-edge fabs. The same goes for companies like Tata Electronics, which has collaborated with Powerchip Semiconductor Manufacturing Corporation of Taiwan to make chips for power management, display drivers, etc. This work is at the bottom of the semiconductor chain.

There is no denying that India should have a stake in the semiconductor industry, given the economic, strategic, geopolitical and foreign exchange compulsions. However, we need a clear set of goals and vision, and the pathway to achieve them. The emphasis should be on leveraging India’s advantage of having a skilled workforce in chip design, promoting microelectronics education and research and developing an industrial base. Of the $50 billion the US has committed for the semiconductor sector, as much as $11 billion will be spent on R&D. We see no such focus in India. On the other hand, the overall R&D kitty appears to be dwindling. The government also needs to build political consensus, like in space and atomic energy programmes, because the investments required are long-term and need continuity.