SEMICONDUCTORS – FAB OR DESIGN – THE ROAD AHEAD

Different countries are working on strategies to ward off this crisis, which they feel is impending and is likely to throw off a large number of manufacturing / economic activities out of gear. India has lots of misses in this area for a variety of reasons. Given the gravity of the situation, India seems to have taken a call. It has decided to make a serious foray into chipmaking. It has announced a $10 billion effort and we can see some initial signs of success. How much and how well can we see the trajectory to that high end, at the end of the fab tunnel which we are yet to initiate.

The need no sane person can question. Making mobile electronics domestically has been cited as a case in point, the value of mobile phone production grew over eight times in the last eight years. Does it provide the very much needed glimmer of hope or the chip fab is a different world altogether? The story of the chip is a long tech tale of its manufacturing, and has a lesson to be learnt, however audacious our thoughts and plans may be. As it has to happen finally. But for that it is a failed attempt and that too after long years of fruitless toil, may not be worth it. China’s bold attempt has made very marginal gains in over two decades of persistence. Through huge tax breaks , China has funded over $50 billion in the last decade.

From none to 16% of the world’s production between 2000 to 2020. But even today they are five years behind the technology leaders in this field. ”Chinese chips are mostly memory chips as opposed to high end semiconductors. US/EU export controls will widen the gap. It might develop it internally, but the gestation period would be over a decade. China proves that decades of effort and billions in capital investment aren’t enough to assure success. Chip manufacturing is the hardest technical exercise; ”correctly etching and connecting tens and hundreds of billions of transistors 50 times smaller than a virus – and is getting harder.” From 30 chipmakers, twenty years back, we have only two now. Is it a fair challenge for a new entrant to get into?

Fab may not turn out to be a fabulous dream, if we go by history. Miracles do happen. Chip design can be a winning area, Indian companies are already providing low end chip design services to third parties, with over 2,000 chips being designed here and 20,000 engineers working on it. Support and preferential access,  can create a cutting-edge fabless chip design industry, the Israel way.  Outsource manufacturing to third party and move ahead in the semiconductor value chain. Design is not the only area to claim a bigger stake in the semiconductor pie, India can get into OSAT (Outsourced Semiconductor Assembly and Test) and ATMP (Assembly, Test, Marking and Packaging) industries, which together make up for 50% of the semiconductor supply chain revenue. These activities need much less investment and have much higher employment potential, might turn out to be more suitable to India.

Sanjay Sahay