Micron Technology Inc has announced plans to set up a $2.75-billion semiconductor testing and assembly facility in India, the first of its kind in the country. Under the government’s “Modified Assembly, Testing, Marking, and Packaging (ATMP) Scheme,” Micron will receive 50 per cent of the total project cost from the Centre and 20 per cent of the from the Gujarat government, implying that Micron will invest $825 million. This facility is expected to create 5,000 jobs.

Given the company’s recent trouble with China, Micron has stepped up its effort to diversify its operations. The company plans to invest $100 billion in building a chip factory in Clay, a city in New York, and to invest $3.6 billion in a chip fabrication plant in Hiroshima, Japan. In these deals, the US and Japanese governments have offered subsidies of $9.5 billion and $1.6 billion, respectively.

A variety of concerns and criticisms have been levelled against this deal. The subsidies offered amount to spending $386,000 per job created (70 per cent of the total cost of the project divided by 5,000 jobs). It is asked, could we not find a better way of creating jobs?

Moreover, one could also ask whether investments should be in the assembly, testing, and packaging of chips or whether we should invest in the design and manufacturing of advanced logic chips as processors if we are to create a vibrant semiconductor industry in India.

Right step

These criticisms miss the mark on several fronts. First, let’s look at the question of jobs being created. The assembly, testing, and packaging of semiconductors are highly automated, unlike the assembly of cell phones. That’s the reason why it creates fewer jobs. However, the semiconductor industry is not about jobs but rather about strategic importance.

Semiconductors are used widely in many commercial applications, critical infrastructure, and defence equipment. During a potential international conflict, India’s economy will suffer immensely without local manufacturing capacity and supply chain disruption. Job creation must come from other industries, not high-technology manufacturing.

The question of whether India has chosen the right entry point into semi-conductors has also been raised. Semiconductor manufacturing is bewildering in its complexity. Creating an ecosystem that is truly free of disruption from external forces is a mammoth task.

One would need to create indigenous champions that manufacture speciality chemicals, software design tools, precision manufacturing equipment, and assembly and testing facilities — all critical in manufacturing semiconductors.

Developing all these at once will take trillions in subsidies and is not a wise course of action.

In contrast, ATMP industry is the more straightforward step to enter. It is geographically diversified and not controlled by a few large players. A state-of-the-art ATMP factory costs less and takes much less time to get fully operational than a plant manufacturing advanced memory or processors chips.

Successful operations in the ATMP segment will build a skilled workforce, drive supply chain indigenisation, and eventually lead to opening other types of manufacturing operations.

One should see this as a strategy akin to what was followed in developing the auto parts industry in India. Let’s start at the low end, master the art, and develop the market before expanding to manufacturing advanced processors and memory parts.

There is a third solid reason for pursuing ATMP manufacturing. For the last decade, the number of transistors on a chip has doubled every few years, as predicted by Moore’s Law. But with the slowing down of this trend, the industry is looking for different ways to pack more in a single chip.

Role of packaging

One of the promising avenues for progress is the integration of heterogeneous chips, such as processors, memory, and sensor processors, in a single package. “Packaging is where the action is going to be,” said Subramanian Iyer, a professor of electrical and computer engineering at the University of California, Los Angeles, who helped pioneer the chiplet concept. “It’s happening because there is actually no other way.”

This is a widely accepted wisdom. Advanced packaging research is one of the areas being funded by the CHIPS Act of the US, which has set aside several billion dollars for that purpose.

Integrating separately fabricated chips to build complex systems in a package has many advantages. It will reduce the power consumed, volume, and weight of the electronic systems built with such a package. In the coming decades, this will be increasingly essential to technologies ranging from cell phones to mainframe computers to automobiles.

However, moving from simple testing and packaging to innovations in technologies involved in these activities will require considerable research. Imagine having a cell phone battery that needs to be charged weekly rather than daily or a cell phone that is as slim as a credit card!

This suggests that we must not simply stop at building the proposed Micron plant. We must use this as a springboard for future industrial development, requiring considerable research and development.

To do this, besides providing subsidies for this deal, the government should set aside a mere 2-3 per cent of the total subsidy to Micron to develop an advanced packaging research facility that takes advantage of the ecosystem being developed by this factory but resides outside Micron facilities, possibly in a technological research park in one of the IITs to undertake more fundamental research and development in the growing field.

The writer is Founder and President, Maker Bhawan Foundation, dedicated to supporting hands-on R&D in Indian technical institutions. Views expressed are personal