FABulous or FABless India

Will India be a ‘FAB’ulous or a ‘FAB’less country? This is not a million dollar question but a question of millions of dollars. Many industry stalwarts feel that India should have a fab considering the boom here and other experts feel that it is too early for India to take the plunge.

Fabrication is a capital-intensive business with cyclical in nature. So if we are going to have a fab in India we should be clear on the cyclical nature of the business and the risks associated with it. Like any other business it is very important that we succeed. To succeed we need to invest in a fab during the up trends and down trends. This will help gain the momentum in the long run with minimal or zero failures. While we may not be prepared right now, we would be so in about seven or eight years. This time frame can be reduced to three or four years if some of the big Indian companies consider acquiring technologies from good fabs around the world, and then migrate or replicate the same here.


Capital intensive
Having a fab facility is capital intensive with large expenditure and demands of water, power and gas. This should also be supported by a good ecosystem of suppliers that support such kinds of facilities. Building such an infrastructure is expensive and is largely dependent on economies of scale. For fabs to be operational and profitable it is essential that the loading is optimal or above optimal. Here the demand for semiconductor components is in limited volumes and in different types. So the economies of scale do not support a fab facility in India. The only alternative would be to ensure a regular influx of fabrication orders by the ecosystem. But as of now India is not geared for such uniform demand.

An Indian fab will not be very competitive when compared with the other established fabs across the world in terms of costs, scalability, flexibility, etc. India has been successful in the software segment and off late, it has become an important destination for hardware design. There is a greater demand for hardware, systems and board design due to the availability of skilled manpower and the supportive ecosystem. These skill sets will not be of great help for fab facilities, instead there are different skill sets that are required for fabs and this is presently scarce in India.

On the flipside you can see many Industry experts in India highlighting the Moore’s law that transistor density doubles every 18 months or so. This essentially means that either the chip would keep getting smaller or would be capable of more or more functions. As such, a fab unit would need to keep investing to keep pace with industry developments.

The general view is that all fabs are in the decline stage or in the loss-making stage except one. The fab industry is currently entering a phase of consolidation and India would do well to wait until this consolidation settles.

The truth, say experts, is that India is really only suited for testing and assembly facilities, not for full-blown fabs. There is enough chip fab capacity in the world, so India’s effort would only create a surplus. Countries such as Singapore and Malaysia have testing and assembly units, so India should continue doing what it does best: software. Furthermore, the country does not have the environment to accommodate the development of fabs. The country’s infrastructure is exceptionally poor, and the highly mechanized fabs are not big employers. Finally, fabs require heavy government subsidies, which the Indian government cannot afford.

Cost efficiency
Experts feel that having a fab really does not mean cost efficiency, cost competitiveness, cost savings or cheaper chips. The cost of setting up a fab is about $3.5 billion capital investment and with increasing prices the costs would become very much higher because of the newer technological processes.

India would do well with a fabless model because of its competitive strength in the talent pool and design activities. The entire industry here is changing with times and the current trend is ‘innovation & reinvention’.

On the face of it, having a semiconductor fab in India will make a great business sense but looking at the real challenges, the future of having a fab in India is further than what it seems. The country should wait for at least five to seven years to have full-fledged fab.

You can reach Raghu Panicker at raghu_panicker@mentor.com.