What should the Indian semicon/electronics industry do now?

March 5, 2010

Borrowing a quote by one of my acquaintances on LinkedIn, as they say, “winners find solution to every problem, while losers find problem in every solution.” Let me attempt to find some solutions for the Indian semiconductor and electronics manufacturing industries and qualify (or highlight) myself as a winner (LoL)!

If you stepped back and took a long, hard look at the Indian semiconductor industry, things look really fine. Design services and embedded are thriving, as always. Lot of senior executives from the overseas industry keep visiting the country every now and then.

However, India lacks in manufacturing, especially, electronics manufacturing. And here, the absence of a fab is glaring! Okay, I know, it is perhaps too late to wish for one, but I am a dreamer. Can’t help it!

To one of my earlier blog posts titles Union budget 2010: Solar, UIDs all the way, there have been several interesting comments on LinkedIn. I’m taking some of those, here, and try and determine what should India do!

First, a friend commended the Government of India’s NSM and focus on solar, and correctly so. India needs solar and the initiatives are great steps in the right direction. According to his calculations: Consider how much energy a 100 W solar panel will generate over its lifetime:
* Rated capacity: 100W.
* Power generated (at 50 percent efficiency) = 50W (because of dusty environments in which our panels live in).
* Number of hours sunlight stays: average of six hours per day.
* Days with no cloud/smog etc: 200 a year.
* Power generated over 25 years: 50x6x200x25=1.5 million watts!!

He says that even if India can store the power and use the power with 50 percent efficiency it will still clear 750 KW of power over 25 years. Is that a bad return for Rs 15000 of initial investment? Quite right! India has taken all the right steps so far that will help it to go on and become a global solar power player in the future.

Yet another friend feels that after 2005, India has had several opportunities for investment in semiconductor manufacturing. However, by the time an investment package was finally put in place, a few had already gone over to China and then it was followed by the economic crisis. He added that investment into semiconductor manufacturing is in terms of a few billion dollars, and therefore, it requires a very reliable and robust infrastructure (mainly power and water) in place. Today, infrastructure is seen as the biggest risk. Again, very valid points.

Notwithstanding the great initiatives made in solar PV in India, the fact remains that it is NOT semiconductors, and can never will be! Even though some say that many of the equipment and processes used in semiconductors are also used in solar, not to mention the personnel and the industry players.

However, how will all of this help the Indian semiconductor industry? We are still at design services and embedded! There are some product development companies, as most saw at the recent ISA Technovation Awards 2010. However, much more needs to be done to really have India on the global semicon map.

Another acquaintance from the US feels we would need a lot of investment for a 22nm fab. It is like building steel! We need semiconductor as the basis for infrastructure. G7 has it. The government of India should have an organization like the MITI (Ministry of International Trade and Industry), Japan, [now, METI (Ministry of Economy, Trade, and Industry). Semiconductor is used in every segment of the food chain. Now is the time to invest and make rapid progress in this field.

This is indeed, interesting! Will India ever have a METI? For that matter, how about having something like Taiwan’s Ministry of Economic Affairs?

I have my doubts as to whether a 22nm fab is even worth considering for India. First, India started rather late in its quest to build semicon ecosystem — by 16-17 years at least, and you may agree or disagree with my estimate. The best it can hope for is to focus on fabless and design services. Second, I’m not sure whether 22nm is the right node to target for, at least in India. India can and should have made some attempts — at least a decade ago, to develop 90nm and even 65nm space. I believe that the time is still not lost or gone!

Later, another friend, a Doctor in the solar/semicon field, said it takes a lot of investment to go to the 22nm node. Utilization of this technology is mostly for the memory market with a smaller subset in logic products with integrated memory (mostly NAND/NOR flash).

With so much glut in the memory market, ask yourself this question: does it make sense for the Indian semicon sector to go for billions of investment for a low pay back (memory) product? Or, would it make sense to implement slightly older technology with a roadmap to evolve to future ones, a la the 193nm node (Intel’s plan) migrating to 11nm node (it is not going to be a whole lot cheaper either). The answer is simple, since much of India’s IP and value addition is in the logic sector.

So what’s the solution for India? What should the Indian semiconductor industry do now?

The ISA Vision Summit 2010 saw some ambitious numbers being thrown up. India is said to be sitting on an electronics consumption opportunity worth $400 billion by 2020. During the Summit, the Karnataka Semicon Policy was announced, and along with it, a state target of $120 billion by 2020 was mentioned.

It is not going to be easy to achieve this target, nor is it going to be easy to achieve the so-called national target of $400 billion by 2020. At least, not with the present state of things. I sincerely hope that at least 50 percent of this is achieved, as it will really help the country make giant strides in semicon.

To achieve these goals, it would require more initiatives on part of the Union government, the state governments, etc., besides the industry and the industry association. Simply announcing a policy and then sit back and wait for things to happen may not help!

For instance, the announcement of the Karnataka Semicon Policy could have been accompanied with news about several other initiatives such as setting up of local plants, or even granting of land to companeis — something that was handled very well during the Solarcon Inida 2009 in Hyderabad last November.

For instance, Solarcon India 2009 saw an MoU signed between the Fraunhofer Institute of Solar Energy ISE and the University of Hyderabad Knowledge and Innovation Park (UoH KIP), in association with the (ISA. SolarCity’s logo was also unveiled and allottment letters were handed to four companies, and later, Moser Baer was allotted land. This is what I mean! All of these were very powerful messages being sent out to the world during Solarcon India 2009. Something similar, though not completely, was repeated at the ISA Vision Summit 2010, where Indian startups were recognized at Technovation Awards 2010. Again, a very powerful message!

So you see, I really don’t have to touch solar PV! It is a different segment altogether that needs all of our support. Post Solarcon India 2009 and the National Solar Mission (NSM), the initiatives will hopefully trigger off a very significant and ground breaking solar revolution in the country and make India a leading power in the segment.

Solutions for semiconductors and electronics manufacturing in India
Coming back to the solutions for semiconductors and electronics manufacturing in India — first, we definitely need to start developing our markets, focus on developing local products, concentrate on developing a local components ecosystem, nurture inland manufacturing, promote startups, and so on and so forth. Activities such as these will start to attract a few foundry players, I hope.

Now, how do we go about developing our markets? Well, a way could be to identify local problems and develop products built to address those problems — a fact mentioned time and again by several industry players. The difference: these products should be developed and manufactured by Indian companies that are into electronics manufacturing.

Here, we would need the support of fabs and foundries. Perhaps, a foundry like Dongbu HiTek, whom I recently met, could bring an ecosystem to India that could include silicon manufacturing, product testing services, failure analysis service, assembly services and sales channels for the Indian fabless companies. This would allow Indian companies to develop and market products for the global markets. I am sure that other smaller (and bigger) foundry players could do the same. But, they would all need a starting point, and we now need to provide some starting point for such foundry players.

Look, I am not supporting Dongbu here. I was simply lucky to meet up when the company visited India! It could easily have been any other foundry player. Well, I did meet Israel’s Tower Semi back in 2007! But I’ve not heard much about its India plans. Will I? I hope so.

I have also seen what Cre8 Ventures recently had to say regarding their plans to help develop product companies in India. If they come here, and do that, or at least, achieve some of that, it will be absolutely brilliant! India will definitely need more such ventures.

Recently, I heard a bit about National Instruments’ Planet NI Foundation, which desires to solve the world’s problem through engineering by providing access to technology. Planet NI is a unique approach to help the SME’s in India achieve economic prosperity through access to technology. India will deinitely need more of such foundations.

One more thought, can the Indian CEOs of MNCs based in India come together and try and do something, or even put together a plan for developing local products and markets? At least, the like minded Indian CEOs of such companies should definitely have a discussion and see what can be done. I am sure and hopeful that something can always be done!

Okay! Was I able to suggest solutions? Or, am I still stuck on the problems?

  1. Carson
    March 6, 2010 at 9:55 am

    Many thanks for another insightful report and for your positive sentiment towards Cr8.

    I agree with your sentiment toward a regional FAB. I passionately believe the starting point is to bring the right leaders together in the region (as you suggested).

    I really can not wait to get back and spend time with you and the other 16 people we cited to begin to lay down the Indian chapter!

  2. Gary Dare
    March 6, 2010 at 8:55 pm

    Pradeep, going to solar for silicon fabs would be a good idea for India because many solar plants in the West are built from older semiconductor plants (e.g., old Hynix plant in Oregon; NXP conversion in Nijmegen). And, a lot of older tools can be purchased on the used market for cheap, compared to what must be purchased new for cutting edge electronics.

    Reliable water and power supplies are an issue in India, but with solar, that is an industry that can create its own ecosystem. Grow factories by using some of their own solar panel output to create power supplies within their own ecosystem. In other words, some aspects of the plant become self-replicating!

  3. March 6, 2010 at 9:12 pm

    Hi Gary, am sure that the folks in the solar industry in India are aware of all this. However, it all depends on what they actually set out to do! Still early days!

    Thanks for stopping by my blog. 🙂

  4. Asher Bastaker
    March 7, 2010 at 7:41 pm

    I think that Indian companies in semiconductors must expand wisely, worldwide, and become a key player in the world.

  5. Nishit Shah
    March 7, 2010 at 9:08 pm

    Hi Mr. Pradeep, your blogs on indian semicon are always interesting and i read them regularly.

    I do not know how relevant my thought could be as I am just a beginner in semicon industry. Just finishing my graduate studies. But from my stay of 2 years in US, I have learned that major high-tech revolutions being done in US were at first the part of US defense project, which were then commercialized. This is how US infrastructure developed and is where it is today.

    I think India could learn from this and can take steps in right direction. India has all resources needed for conducting researches at PAR with any other nation. More investment should be done in Research and development. I do konw, lack of fab support could be a hurdle in rigorous research and hence government should realize that semicon could be wise investment for india looking 10 years down the line.

  6. March 7, 2010 at 9:50 pm

    Thanks Nishit. We seriously need to look at R&D. I do hope people take due notice of what’s being said and written, and take the appropriate steps.

  7. Ex-Intel
    March 30, 2010 at 1:00 am

    As an experienced Semi professional, here is my two cents on SEMI opportunities in India.
    a) Its not worth the margins to setup a older generation fab in India. Look up Grace Semiconductor. Few folks thought buying old toolset and setting up a fab in low cost China would make it hugely profitable. But they have lost a lot of money and still trying to get off-the-ground.
    b) Its not that a Semiconductor fab would not be successful, but the scales of manufacturing needed to compete with the foundries is huge, and the margins are thin. Unless you have a guaranteed market, or a unique product.. it does not make business sense.
    c) The future is printed electronics. India should invest heavily in printed electronics where there is low capital costs, low costs of manufacturing, huge demand and use of local resources is cheaper than importing it from other countries. Maybe that is the way to go.

  1. No trackbacks yet.
Comments are closed.
%d bloggers like this: