India semiconductor fab business challenges; from the horse’s mouth

Date: 11/12/2013
There are variety of opinions on India's capability in having a locally owned semiconductor fab. To give you more authentic opinion, we are happy to provide you the experience/opinions from a brave semiconductor entrepreneur Mr. Deepak Loomba, Founder, De Core Nano-semicondutors Limited, who risked his career and money in establishing a LED and compound semiconductor fab in Gandhinagar, Gujarat, India. Though the fab, De Core established is not a CMOS fab, but the infrastructure requirements are nearly similar. Below is full email interview (unedited) with Mr. Deepak.

Q1: Semiconductor is a tough area, What are the challenges (in terms of investment, infrastructure, support, and any other) Indian semiconductor manufacturers come across while building semiconductor fabs in India?
I suppose the question asked is what should be the bottlenecks. Investment is certainly a bottleneck because semiconductor is an area which is very capital intensive and hence establishing a global size fab is a very big challenge. Concurrently semiconductor is also volume game, so there are only very few companies who are able to play the niche player role. In most of the cases, the time from niche to mass in case of material science is pretty short. Being a volume game means that by making a few niche products, one cannot make hay. One has to go for volume, which requires a lot of capital investment, because you need utilities that are critical.
Infrastructure – I think maybe things like port and roads are not so important. Supply of quality uninterrupted power supply is very critical. There would be very few places in India where one can generally get power without any problem, that's why we decided to put up a fab in Gandhinagar, Gujarat, because power is amply available there.
Water – may not be much of an issue. in any case fab is going to use any normal water to double, triple reverse osmosis, purification, deionization to bring it to a level where it can be utilized.

Other bottleneck would be – supply of semicon grade materials and gas. We don’t have any supply, because we don’t have any industry – so it is a chicken and egg situation. As there is no industry, there is no stock available in India. Now if you compare with Taiwan as an example, in Hsinchu Science city, companies like BOC, now Linde have actually installed gas pipe lines for supply of semicon grade gases especially nitrogen, argon, hydrogen in their entire area. So if any new semiconductor firm needs to come in – they can just tap into the existing pipeline, so you don’t need to have investment in huge gas storage yard– as we have it in Gandhiangar.

Q2. Are you facing any fab related infrastructure bottle necks in India? Can you elaborate both good things and the things, which need improvement? (Deepak has already answered much of this question in the previous question, he elaborated the infra bottlenecks further below)

As I said – two most critical bottlenecks. First, non availability of materials– whether it's metal organic compounds, semicon grade chemicals & gases, wafers, which we don’t produce it in India or talent (scientists and engineers with knowledge of producing them). We don’t have a school of thought in material science, most of our Indian talent –which actually evolved in the system of IIT, went out of India – did masters, PhD in USA and – went on to redraw the boundary the science and technology in US. Unfortunately there are very few who are in India, and got a job in their required field of interest.
Q3. Now there is support from the Government to set up fabs and also two or more fabs expected to come up shortly, What are the key markets for semiconductor fabs on immediate basis?
An important thing to know is that– it's incorrect to actually club all fabs together. When it comes to market, future prospective in establishing semiconductor fab, one has to understand that fabs come in different hues and colors. I understand that the two fab talked here are going to be CMOS fabs. Then there would be fabs like us, which is optoelectronics fab, we are not as of now into any manufacturing ICs – though it's on the horizon. We are a compound semiconductor fab, so the wafer here is not silicon but saffire. In terms of whether more fabs will come into India is difficult to comment, because silicon and CMOS are very competitive.

I don't have a slightest doubt that, we are at a dawn of semiconductors. You would have refrigerators & air conditioners working on Solid state devices (SSD), because most of SSDs are having very long life and optimal use of natural resources.
The key market on immediate basis is all electronics. There is a huge opportunity in power electronics because there is less competition in power electronics devices as compared to other CMOS devices.

Q4. What is the reason for big players in the industry such as Tata and Reliance reluctant to enter electronics and semiconductor manufacturing in India? India is now emerging as a big market for IC chips now.
I really can’t comment on Tatas or Reliance, they have got money, but maybe not the desire. I can probably assimilate from my experience that vis-a-vis other markets. Semiconductor and electronics are one sector, which has to be driven by people who come from the technology at their helm. Reason: These are very fast evolving areas, so technology that are used today– will not be used 5 months from now. In such dynamic situation– even public companies or even private companies, which are not headed by technopreneurs at their helm– would have a large gap in decision making. Look at Intel, IBM, HP– all are run by technopreneurs like Gorden Moore, Steve Jobs, Michael Dell, etc.

Q5. In India, Engineers are more keen to learn soft design rather than material knowledge of electronics? What are talent gaps in semiconductor manufacturing? and your long term and short terms solutions for that?
Yes, engineers are more keen to learn soft designing. Most of them tend to sit at a place and work rather than get their hands filthy. Then off-course, there would be few self motivated people– who would really want to do this– but earlier they did not had the opportunity. Once the fabs are here in India, lot of these young fellows who wanted to work in these area will get an opportunity to get their hands dirty. But this is all in future, in short term, I think we have to have very strong government support – we have to build centers of excellence.
We need to have exceptionally intensive & rigorous collaborated programs with probably some good university/institutes. But I think university would be a good idea– because IITs anyway get lot of lime light, while lot of good universities, which have very good talent are actually left high and dry. It makes lot of sense to tap talent in pure sciences. We don’t always need engineers, there is an obsession about everything to be done by an engineer. Actually high quality MSc (qualified) pure science guys can be trained to do an excellent job and in fact – what I like more in MSc guys is that– they have more stability and more stable in their attitude.

Q6. Is Indian science and engineering institutes doing significant amount of cutting edge research in advanced semiconductor manufacturing such as 450mm, Graphene and nanotechnology materials, organic semiconductor etc.? Is there any connection established by private entrepreneurs like you with premier institutes in India?

Honest answer to this is– that there is little to no institute, which are doing cutting edge work in any of these areas in India. I am sure there are some oases, some institutes work exceptionally. But, if you look at large in material science, there is practically little to nothing that institutes do in India– as they are more or less inclined to publications and not on ground breaking innovation.
There is a general dis-belief in government that private entrepreneurs are there to just grab money and run away. So even today, we cannot get even one single research grant from government institute. We applied for a government grant and after going through the entire process, we ultimately understood that the grant is available only for Rupees 1 crore. In our industry– Rupees 1 crore grant is not much, because we won’t be able to buy single equipment, not even a characterization equipment within that money.

Q7. Don't you find the speed at which Government is taking decisions for electronics industry is very slow? Any suggestions!
I think its all-right, these are policy decisions, one cannot anticipate that policy decision in country like India– can happen in a MIG. It always take time– because government is accountable to citizens. By the way– even in US, these guys couldn’t even get their government function for three weeks, because of the disagreement between president and their legislature, other than – may be authoritative regimes like China or somewhere else, where decision would have happen at the whim of the leaders. I think the job being done by government in terms of its sheer width is good. Specially the Department of Electronics and IT, it has done momentous tasks by bringing back the policy again which was earlier an issue.

Q8. Tells us something about De Core’ products
De Core houses LED Luminaire products ranging from house to Industry to even streets. It has products like LEDREV for street light, high mast & outdoor lighting. BAYLED as industrial, warehouse and shed lighting. LEDIT covers commercial applications with downlighter, False ceiling LED lighting and LED Panels, while STARRYLED for Park lights. Its upcoming state of art commercial products such as RENE and DANIO may change the outlook of home lighting altogether.

deepak