Chandrayaan-3 LVM3-M4 Mission Post Launch. (Courtesy: ISRO)

 

From a technology policy lens, the success of the Chandrayaan-3 mission in 2023—which saw India become the fourth country to land a rover on the moon and the first to do so near the Lunar South Pole—brings up a pertinent question: If largely government-run efforts could make India a bonafide space power, can some of those learnings help India become a semiconductor power?

Geopolitical competition between the US and China, as well as a perceived overreliance on a seemingly vulnerable Taiwan for the vast majority of advanced chips, has made the semiconductor manufacturing sector the focus of intense industrial policy efforts over the last few years, after decades of it being the poster child of globalization. Countries around the world have doled out state-sponsored incentives and promised favorable policy environments with the aim of establishing a local chip-making industry. India, too, announced $10 billion in incentives under the Semicon India Programme, with hopes of becoming a major hub of chip production. While India has a strong presence in the chip design services segment, it has no commercial chip manufacturing facilities. The Semicon India Programme hopes to build India’s muscle in all segments of this supply chain.

As with space programs, success or failure in these efforts will depend heavily on policy choices made by governments, going well beyond just the amount of money invested. And, as with space programs, actually achieving the aim of building a domestic semiconductor industry would put India into a small, elite club of countries that have tasted chipmaking success. Efforts to pull this off are, in fact, often referred to as “moonshots,” making the comparison to Chandrayaan evident. So, are there lessons for India’s semiconductor mission from its remarkably successful space program?

This question was one of the drivers for our book, “When the Chips Are Down: A Deep Dive into a Global Crisis” (Bloomsbury, 2023). In it, we survey the efforts of various nation-states to build a domestic semiconductor industry. We find significant differences between old-age technologies—space and nuclear being the canonical examples—and new-age technologies, such as AI research and cutting-edge semiconductors. We thus conclude that transposing policy approaches from one domain to another is neither desirable nor effective.

Three characteristics differentiate space and nuclear technology from semiconductor technology. First, the output demanded from space and nuclear programs is small. Companies generally need to produce products for just one buyer—the sovereign government. Contrast this with a semiconductor manufacturing facility, which cannot survive by merely meeting the small government demand. Given the massive upfront capital costs that a chip manufacturing facility requires, it can hope to achieve financial sustainability only by simultaneously contracting out its manufacturing services to several large-volume chip design firms.

Second, because the output required from space and nuclear programs is small, the capital investment needed is manageable for governments. For instance, the budget for the three Chandrayaan missions totals up to Rs. 1,977 crore. By contrast, a single chip packaging plant that Micron Technology, an American company, plans to set up in Gujarat is expected to cost approximately Rs. 16,000 crore.

Third, the supply chain in space and nuclear domains is short and can be substantially indigenized. India’s space program benefited from initial technology transfer from the US and the USSR. After that, a small set of brilliant scientists began the process of indigenization and technology upgrading. Similarly, Pakistan started its nuclear program through technology transfers from China, and its scientists began upgrading the incoming technology. With consistent government backing after the initial technology transfer, such government-run programs could succeed. Such a pathway is not suited to the semiconductor supply chain, which follows a comparative-advantage-based specialization approach. Companies in different geographies specialize in specific segments of the supply chain. The US Semiconductor Industry Association estimates that as many as six major regions (The US, South Korea, Japan, China, Taiwan, and Europe) each contribute 8 percent or more to the total value added by the semiconductor industry. Even a semiconductor genius of the caliber and vision of Homi Bhabha or Vikram Sarabhai, the gifted scientists often credited with the success of India’s nuclear and space programs, will not obviate the dependence on foreign companies for intermediate inputs, talent, and capital.

We find empirical proof for this reasoning in the fact that many regimes with robust state-run space and nuclear programs couldn’t achieve similar success in semiconductors. Consider China’s case. In 1945, Mao Zedong popularized the slogan “zili gengsheng” — “regeneration through one’s own efforts.” In 1956, a landmark twelve-year plan was announced to provide a roadmap for the state’s science and technology efforts. Semiconductors were among the twelve technologies identified as a top priority under the plan. Universities started semiconductor programs, and government-run factories and research labs began operating in the sector. By 1965, China had developed its first chip before Taiwan and South Korea. However, after some initial successes, this primarily government-run effort fell behind other countries. These firms started faltering without the discipline that market finance and competition bring in. These firms couldn’t meet the demands of constant upgrading and capital infusion. Being disconnected from the American semiconductor industry during the height of the Cold War also played a role in slowing technological upgrading.

Nevertheless, whatever progress China made came to a rude stop during the Cultural Revolution. When an attempt was made to revive the industry in the 1980s, China’s semiconductor industry was a laggard—most companies failed to reach production targets, and the technology was several years behind. Things began to change only when Foreign Direct Investment started flowing into the Chinese semiconductor industry in the 1990s.

The USSR’s case is also quite similar. Like the Star City for its space program, the USSR hoped that a new city, Zelenograd, would become a scientific paradise that would excel in semiconductor manufacturing. It was supposed to be the Soviet “Silicon Valley.”

The USSR followed a “copy it” model by smuggling chips from the US and trying to reverse engineer and manufacture them locally. Because the USSR was cut off from the global semiconductor ecosystem during the Cold War, these companies focused on supplying chips to the Soviet military rather than chasing the ever-growing global chip market. Eventually, these projects fell behind. Russia still doesn’t have a single commercial semiconductor manufacturer today.

India’s case is not too different. Two of India’s public-sector firms, Bharat Electronics Limited (BEL) and Semiconductor Complex Ltd. (SCL) were able to strike technology transfer agreements with competent Western chip-making firms. While they met initial successes, they faded away by the 1980s.

These initial experiences of the USSR, China, and India provide insights into the public policies needed for semiconductor manufacturing success. First, government-run chip companies had no incentive to compete in a hyper-competitive domain that demanded constant capital infusion and technology upgrading. They started well but couldn’t keep themselves in the race for long. Even their customers within the government eventually turned away and sought better technology at cheaper rates through imports.

Second, these companies were shielded from internal competition. Competition forces companies to seek differentiation. Without it, these companies had no reason to chase new customers beyond government departments. BEL and SCL were a part of the race to make chips. But from the government’s perspective, this competition was undesirable—having two companies perform the same task wastes resources. While SCL was chosen to manufacture chips, BEL was confined to assembling them. In Taiwan, on the other hand, the government-led ERSO could spin off multiple private companies and foster competition successfully. While India’s approach may have saved the government precious money, it perpetuated a structure that was fundamentally at odds with innovation.

Third, inward-looking trade and business policies proved costly, as in the case of the USSR. The dominant economic narrative was to save foreign exchange and dollars from leaving the country. This meant strict import controls and exorbitant tariffs. Even after paying these duties, equipment remained stuck at ports awaiting government approvals. The cumulative effect was that products from BEL and SCL couldn’t compete internationally. Seeking self-reliance, the government was neither interested nor confident in exporting chips.

These experiences suggest that policy recipes from the space domain cannot be replicated in semiconductors. Nevertheless, success in space and nuclear fields remains an immense source of psychological motivation. It reminds us that with the right policy ingredients, India can make technological leaps happen.

Authors

(L-r) Pranay Kotasthane is Deputy Director at the Takshashila Institution and chairs the High-Tech Geopolitics Program. Abhiram Manchi is pursuing an MBA plus MS in Digital Technology at Boston University.

 

This article first appeared in India in Transition (IiT), published by the Center for the Advanced Study of India (CASI) of the University of Pennsylvania. All viewpoints, positions, and conclusions expressed in IiT are solely those of the author(s) and not specifically those of CASI.

© 2023 Center for the Advanced Study of India and the Trustees of the University of Pennsylvania. All rights reserved.