India’s fab dilemma

With India’s recently-announced semiconductor policy, “to fab” or ‘not to fab’ is a dilemma being pondered in the India semiconductor ecosystem.

A state of the art fabrication unit today requires investments upwards of $3 billion. Shrinking product lifecycles, increasing design complexity, time-to-market pressures, and high investments in setting up fabrication plants have contributed to companies exploring different foundry models.

Fab models
Historically, semiconductors were both designed and manufactured by Independent Design Manufacturers (IDMs). While this business model is still prevalent, the late eighties saw the rise of the dedicated foundry model (pure-play fabs) due to time-to-market pressures, high investments and economies of scale.

Taiwan was early to capitalize on the foundry opportunity by setting up units to cater to the growing demand for contract wafer manufacturing. Taiwan Semiconductor Manufacturing Company (TSMC), established in 1987, is the world's largest dedicated semiconductor foundry. Another major player in niche semiconductor manufacturing is IBM, recognized for its Silicon Germanium (SiGe) BiCMOS technologies for RF/Analog, complex system-on-chip (SoC) and high-performance designs in wireless and networking.

And now, the fab lite model is gaining ground wherein IDMs are retaining the manufacturing of key products in-house while outsourcing the others to pure-play foundries. The impetus behind the fab lite model is that IDMs are reducing their capital expenditure and de-risking themselves from market volatility. In addition, the IDM and the foundry share the cost of process migration giving both a better return on investment.

India as a fab destination
While India is considered a chip design hub, manufacturing has not been a priority so far. The only active fab, the government-owned Semiconductor Complex Ltd., was set up to cater to the local demand for microelectronics; it was recently transferred to the Department of Space (DoS) and is looking at moving to lower process geometries.

Let’s look into whether India has the ingredients to become an attractive fab destination. We score on a potentially big domestic market for electronic products, and the emerging ecosystem of electronics manufacturing service (EMS) providers and original design manufacturers (ODMs). The case becomes stronger with the addition of favourable government policies and rapidly growing infrastructure.

Semiconductor companies would gain considerable benefits if they locate their manufacturing and design operations close to end customers. With a favorable duty regime, this could translate to a cost benefit for system companies with local presence. For example, Nokia has built an effective ecosystem of suppliers around their assembly plant in Tamil Nadu. In such situations, both system companies and the suppliers benefit due to the duty structure and reduced logistical cost.

According to the ISA / F&S Report 2006, by 2015 India will account for 11% of worldwide consumption of electronics equipment. This could mean that the local semiconductor manufacturing will save India valuable foreign exchange which would otherwise have gone in imports.

Recently we have seen a spate of announcements by companies showing interest in setting up a fab in India. The semiconductor policy announced by the government and the growing number of EMS players setting up base in India are key indicators of the growing interest in manufacturing.

Challenges
However, there are a few challenges that need to be overcome before companies we see significant investment flowing into this sector.

One significant shortcoming to India’s prospects of becoming a fab destination is its infrastructure. Not only do chip manufacturing units require basic infrastructure such as power, water, roads, land and ports, they also require trained manpower and a supporting ecosystem. Although the government is taking steps to encourage fabs and ecosystem players to invest in India, we have some distance to cover.

Another challenge is that fabs in India will have to find their own niches and find a viable business model in order to survive and flourish. Fabs in India will be competing with major Taiwanese fabs that offer cost advantage due to their scale, and with the niche (AMS/RF) fabs that have technology as their differentiator.

Alternative Growth Strategies
A growth area for India would be to leverage our existing strength in semiconductor design and embedded software capabilities and move up to system level and product design. North America and Europe are good examples of regions that have focused on product and system level development and have outsourced their manufacturing to foundries outside their geographies.

The advantages of going this route for India are manifold. For one, this could be a parallel process as we seek and build fabs and associated ecosystem. Two, apart from the relatively low investments involved, India could continue to build on an already-strong foundation of design prowess, which means it could start working on it today instead of playing catch-up with companies already well-established in the area of semiconductor manufacturing.

Investment in an ecosystem of assembly, test, mark and package (ATMP) could be another way forward leading to cost and logistics benefit.

Conclusion
While it is desirable to have the manufacturing in the region where there is greatest consumption, globalization has caused regions to specialize and leverage on each others strengths. For example, Taiwan is a hub for semiconductor manufacturing, while countries such as the United States and Japan are well known for product innovation and ownership.

Foundries involve big investments and have unique dynamics. The question of setting up a fab in India has several facets, and investors will carefully consider the business implications and possible returns. This is one question to which there are no easy answers.