India’s semiconductor imports grew from $45.3 billion in 2017 to $54.87 billion in 2021. This import growth trend will likely reach US$ 80 billion by 2026 and US$ 110 billion by 2030. Besides, the global semiconductor market is expected to spur demand to register US$ 692.5 billion in sales by 2025. It will likely grow to US$ 1 trillion by 2030. Hence, India’s semiconductor dream has been from chip taker to exporter.
Unlike the past import substitution strategy, India has been after alluring foreign semiconductor firms (FDI) instead of giving protection to domestic firms. But as access to the domestic market, tax incentives, and free land are not attractive enough, India has come up with $10 billion cash. As high as 50% subsidies to capital expenditure of global firms are on the table. On top of it, India has a large pool of engineering graduates; a growing number of IITs have been producing an increasing number of high-caliber engineering graduates.
On the other hand, major multinational semiconductor companies (MNCs) have already set up captive chip design centers in India. Consequentially, MNCs have nurtured 20,000+ chip designers. To take it forward, the Indian Government has been implementing a program to train 85,000 additional chip designers. For offering hands-on experience, India also has a semiconductor fab at Mohali. But are these strengths sufficient to meet India’s semiconductor dream to graduate from chip taker to exporter? In retrospect, creating a flywheel effect through a flow of ideas is at the core of success. It’s vital for increasing the quality and reducing the cost to be competitive. Due to it, the semiconductor industry has a history of rising and falling as waves. Despite all other strengths, this capability will determine how far India’s semiconductor dream will be met.
- The US-China race makes India’s offers insignificant
- Import substitution strategy weakens the merits
- Human resource is not good enough to meet India’s semiconductor dream
- India’s failure to use SCL as a seedbed to spin off commercial ventures
- Fabless companies for fulfilling India’s semiconductor dream
- India’s lack of track record in creating flywheel effect weakness India’s semiconductor dream
- Focus on discontinuity and creating a flywheel effect from the humble beginning
The US-China race makes India’s offers insignificant
India’s growing domestic market, high-caliber engineering graduates, and $10 billion subsidy could have been strong enough to attract MNCs. But in the global race between US and China, this package has been failing to allure reputed firms. Hence, TSMC or Intel is reluctant to set up semiconductor fabrication plants in India.
The USA has come up with the Chips and Science act to counter China’s $150 billion state financing in attaining semiconductor independence. In addition to USA’s $52 billion, European Commission, in February 2022, expressed commitment to mobilize €43 billion in ‘policy-driven investment’ for the EU’s semiconductor sector by 2030 (bruegel, Oct 2022). On top of it, individual member countries have incentive programs to facilitate semiconductor investment.
Consequentially, Intel announced an initial investment of over 33 billion euros for establishing R&D and manufacturing facilities in member countries of the EU. It includes an investment plan of an initial 17 billion euros into a leading-edge semiconductor fab mega-site in Germany. Intel’s plan focuses on developing R&D capacity in France. Intel’s EU manufacturing and foundry service facilities will be in Ireland, Italy, Poland, and Spain. Rome’s decision to offer a 40% subsidy has already closed a deal. As a result, Intel’s $5 billion investment is to set up an advanced semiconductor packaging and assembly plant in Italy (Reuter, Aug 04, 2022).
The rapid expansion of semiconductor manufacturing in the USA
To take a pie of the US’s fund, Intel has broken ground in Ohio, initially to set up two fabs costing $20 billion. On the other hand, TSMC and Samsung have been busy shrinking their facilities in China and expanding in the USA, Japan, and their respective home countries. For example, TSMC announced in Feb 2022 to develop a new plant in partnership with Sony in Japan costing $7 billion. TSMC is also setting up a plant in Phoenix, Arizona, USA, at the cost of $12 billion. To expand US facilities further, TSMC has decided to set up a 2nd plant in Arizona (Bloomberg, Nov. 2022). Besides, Samsung has laid out a potential plan to spend almost $200 billion on 11 plants in Texas, USA (Bloomberg, July 2022).
Due to expansion in the USA, Europe, and Japan, India is taking a slow lane
It seems that India’s semiconductor dram is caught in Chip War. Due to incentives and other measures for developing a semiconductor blue supply chain, MNCs like TSMC, Intel and Samsung are expanding facilities in western countries. As a result, India’s subsidies and domestic market are failing to attract MNCs to set up fabs in India. Hence, India has formed a technology collaboration with Belgium’s IMEC to set up a 28nm fab (Indian Express, June 2022).
Besides, a 65nm analog wafer fab in India involving foundry Tower Semiconductor Ltd (eeNews, May 2022) is moving forward. Indian oil-to-metals conglomerate Vedanta and Foxconn have signed an MoU with the Indian state of Gujarat to set up a $20 billion semiconductor and display plant(TC, Sept 2022). Such investments will unquestionably not enable India to produce chips using the latest nodes (5nm or 3nm) to power the most lucrative smartphone market segment. Furthermore, as Indian companies are not innovating and designing electronic gadgets, finding customers for those fabs has become an issue. Even if India had fabless companies like MediaTek of Taiwan, those fabs could have leveraged them to take off.
Import substitution strategy weakens the merits of meeting India’s semiconductor dream
After independence, India adopted the import substitution strategy to build the industrial economy. The strategy has been to import capital machinery and design for making copies of the imported products. Hence, to offer a profit-making pathway, India provides an incentive for the import of capital machinery, tax differentials, and protection to exploit the domestic market. But such a strategy fails to encourage producers to keep developing in-house R&D capacities to improve themselves. Instead, they keep making copies based on imported technologies. But the global race keeps progressing in improving products through incremental innovation and reinvention. As a result, products produced by India as import substitution keeps falling behind. Hence, despite early success, such import substitution strategies, thinking, culture, and policies have been failing to build a globally competitive industrial economy in India.
One notable example of failure is India’s Hindustan motors and automobile industry as a whole. Despite some success in labor-intensive value addition, India’s automobile industry has failed to attain a global edge. Besides, due to the reinvention of automobiles as electric vehicles, India risks high-level erosion of its labor-based competence in making automobiles.
Within two to three years, the semiconductor industry substitutes matured nodes with the next one, making products better and cheaper. Hence, success in the semiconductor industry is not about acquiring the capability of making for import substitution. Instead, the focus should be on improving to move to the next node faster than the competitors. So far, India’s import substitution through incentives, subsidies, and protection does not create such a pathway. Hence, import substitution strategic thinking appears detrimental to India’s graduation from chip taker to an exporter. Due to it, India’s semiconductor dream runs the risk of being unmet.
Human resource is not good enough to meet India’s semiconductor dream
Of course, human resources matter. But so far, India’s human resources have been in hard-core engineering. Undoubtedly, Indian engineers have been showing excellence in the global semiconductor industry. They have a strong footprint in Silicon Valley. Starting from designers to directors and CEOs, international semiconductor firms depend on Indian engineers. But that is not sufficient for meeting India’s semiconductor dream. In addition to engineering competence, success demands R&D management and decision-making capability. Due to its weakness, engineering competence fails to be relevant. For example, Intel has high-density Indian engineers. But due to failure to outperform the competition through improving performance and making intelligent decisions, Intel has been compelled to lay off as high as 20 percent of engineers, including many Indian engineering graduates.
India’s failure to use SCL as a seedbed to spin off commercial ventures
As early as 1984, the Government of India (GoI) founded a 100% state-owned enterprise – the integrated device manufacturer, Semiconductor Complex Ltd (SCL) at Mohali. SCL started the journey with 5-micron process technology. Subsequently, it upgraded the facility for fabricating micro-electronic devices in a 0.8-micrometer node. Later, its name has also been changed to Semi-Conductor Laboratory. In addition to making chips for space and defense programs, SCL’s aims include research and development in semiconductor technology. As the global race has made SCL’s 0.8 microns obsolete, GoI plans to spend $1.25-$1.30 billion to modernize and upgrade SCL (ET, Nv 2022).
Not surprisingly, SCL is a typical story of India and many other less-developed countries. India did set up SCL with imported capital machinery. It was upgraded with the import of know-how from Tower Semiconductor, Israel. Highly likely, its plan upgradation will take place through the import of expertise and machinery. Perhaps, India’s linkage with Tower Semiconductor and recent development with IMEC will play a significant role. They keep operating the imported machinery without acquiring the capability of improving by themselves. Such a reality has been at the core of the failure of leveraging SCL in developing the semiconductor industry in India.
Not surprisingly, we observe different growth paths in Taiwan, and Japan. For example, after taking Transistor’s license in 1952, Sony did not keep producing it as per Bell Labs’ know-how, and waiting for updates. Instead, Sony focused on improving the design and production process, making the Transistor increasingly better and cheaper. Consequentially, a semiconductor cluster, consisting of innovators across the value chain, started to form in Japan. Similarly, instead of operating RCA-supplied fab as per the training, engineers of ITRI of Taiwan focused on learning and improving the process by themselves, making it better than what RCA had.
Unlike Taiwan, India kept operating SCL and making it obsolete
Upon getting the research fab in 1984, India pursued the strategy of how to master the operation for producing custom-designed chips. Despite having strength in engineering education, India did not aspire to keep improving fab equipment and yield optimization by itself. India also did not use SCL to acquire the competence to embark on being a globally competitive chip producer—let alone spin-off firms like TSMC.
Unlike India, Taiwan focused on using the research fab they got from RCA in 1974 at the cost of $10 million to master the learning and yield optimizing capability. Subsequently, they spun off the expertise as a commercial operation in 1980—UMC. Later, Taiwan proceeded to scale up the learning further by setting up TSMC in 1987. Taiwan focused on developing a snowball effect through learning, idea generation, and integration to increase yield. They further grew their competence in envisioning the future, patterning with equipment makers and chip designers to keep moving to the next process nodes. This learning performance reached the pinnacle making TSMC the best silicon processor in the world.
The sharp contrast between India and Taiwan has been that India focuses on learning existing technologies and operating them to deliver outputs as per order. And once acquired technologies get obsolete, they go for importing the next node with growing investment. But unlike India, Taiwan has developed the culture and management practice to keep learning, creating a snowball effect for exploiting unfolding commercial opportunities. Hence, India’s lost opportunity of turning SCL into the springboard of the developing semiconductor industry is no exception. As self-learning and improvement capability is vital, India appears to be in a weak position to meet the semiconductor dream.
Fabless companies for meeting India’s semiconductor dream
Over the last two decades, MNCs like IBM, Texas Instruments, Intel, and QUALCOMM have set up captive chip design centers in India. Of course, this is a testimony to the quality of engineering education in India. As a result, India has been blessed with more than 20,000 experience chip designers. In addition to design expertise, they also got a chance to get exposed to MNC’s culture and management practices for developing next-generation chips. But unlike Taiwanese engineers, Indians kept drawing a salary from their MNC employers. India could have leveraged this window of developing indigenous chip design and fabless chip companies. Hence, like Taiwan’s MediaTek, India has none. Therefore, India’s initiative of developing fabs faces the barrier of finding customers. On the other hand, Taiwan’s home-grown fabless companies have been growing as major customers of TSMC and UMC.
India’s lack of track record in creating flywheel effect weakness India’s semiconductor dream
Although India produces quality engineering graduates, India runs short of nurturing humble beginnings through a flow of ideas. Hence, they keep focusing on making. Similarly, in service, India keeps delivering as per customers’ orders. It seems that Indians are good engineering employees. But they do not have the management and cultural abilities to envision and pursue the future by creating a flywheel effect out of a flow of ideas. Unfortunately, such weakness makes all national-level initiatives to focus on operating imported equipment, making them obsolete.
For success in the semiconductor business, whether in the fabless or fabs, there is a strong need for management capability of creating a consistent flow of ideas for persistent improvement. Due to its critical importance and high-level specialization, TSMC’s Arizona plant will have 95% Taiwanese managers. TSMC finds prospective American managers as “giant babies.” Until India addresses this critical deficiency, any amount of public or private investment will keep pushing India in the slow lane.
Missing management and cultural competence of creating flywheel effect for meeting India’s semiconductor dream
As India’s domestic semiconductor consumption has been growing, there is a concern about increasing import bills. Hence, India has been desperate for import substitution. As protection to access the domestic market and tax incentives are insufficient, India has been offering massive subsidies. Ironically, India got access to fabrication technology as early as 1984. Besides, Indian engineering graduates have shown excellence in the global semiconductor market. On top of it, MNCs have been offering chip design expertise to India through captive design facilities.
Despite all these, India has failed to develop the domestic semiconductor industry as Taiwan did. The underlying reason has been India’s failure to have the management and cultural capability to create a flywheel effect out of a humble beginning. Until India addresses this critical issue, perhaps, India’s import substitution-based semiconductor industry development agenda will, at best, replicate the success of India’s automobile industry. Hence, import substitution will be making poorer chips at a higher cost than the global players will be offering, pushing India into the slow lane.
- Chip War
- China’s Semiconductor Independence–prematurely caught?
- Semiconductor Value Chain–globally distributed ecosystem
- Semiconductor IndustryWaves
- Intel Falling Due to PC and Mobile Waves
- ASML Lithography Monopoly from Sustaining Innovation
- Taiwan’s Semiconductor Monopoly – How did it arise?
- ASML TSMC Nexus Fuels Semiconductor Monopoly
- ASML Monopoly in Semiconductor — where is magic?
- SEMICONDUCTOR MONOPOLY DUE TO WINNING RACE OF IDEAS