Market Walk

Semiconductor PLI – Why this plan is going to be easier said than done?

During the week, the government went ahead and announced an aggressive PLI plan for making semiconductors locally. Is that really a feasible idea or would it be too impractical at this juncture.

Think chips, Think Taiwan

When you think of the esoteric world of precision semiconductors, the one name that immediately comes to you mind is Taiwan. That is largely thanks to the giant, TSMC, which has become the second most valuable chip company in the world after Nvidia. But the making of semiconductors is most likely going to be easier said than done. The needs of this business are just too stringent and the business too capital intensive.

Complexity of the Fab

Taiwan Semiconductor Manufacturing Corporation (TSMC) manufactures chips on behalf of other companies. Its largest customer is Apple, which account for nearly 20% of TSMC sales. But the issue here is about the complexity of a Fab. A fab is a place where the actual semi-conductors are made on behalf of the big players. It is estimated that a typical Fab unit or cleanroom is supposed to be 10 times as clean as a pharma research laboratory. Maintaining air purity is key as the chips are extremely sensitive to dust and other particles. The room for even the smallest error in cleanrooms is zero and that is the complexity that would be hard to manage in India.

Cost is another consideration

Fabs don’t come cheap. In fact, the complexity of a Fab cleanroom ensures that these Fabs required huge outlays. It is estimated that a typical Fab room can cost anywhere between $3 billion and $6 billion. The actual cost would be based on technology node and wafer capacity. The controlled environment of Fabs has to be supplemented with high quality air supply, high quality water supply, zero downtime etc. The cost is not just about the investment, which is largely into plant and equipment. The costs of maintaining and sustaining a Fab room is also extremely prohibitive.

Where is the opportunity?

While the Rs.76,000 crore PLI allocation is a good starting point, the opportunity matrix is important. Large consumers of Fab chips like Apple and Samsung will not want to disrupt their existing set-up unless there is a very strong base case. Then there are the big chip makers like Intel and AMD, which also outsource a lot of work to the Fab units. However, this is a business which is technology intensive with limited room for error. Hence, building trust and comfort will be much easier said than done. Taiwan has spent 40 years setting up the ecosystem for semiconductors. From equipment to labour to port efficiency, there are a plethora of issues to be looked into. It is good to be ambitious but India must approach this more pragmatically!

Show More

Related Articles

Back to top button