BANGALORE, INDIA: Is the Indian ecosystem developed enough to enable faster product development cycles to take on competition from the multinational product companies? If no, do we have the economy of scale to have fab in the country that would cater to ideally, the Indian market, besides the global market? Dr. H.V. Ananda, managing director, Synplicity India, addresses these and some other probing questions in an exclusive with Pradeep Chakraborty, Executive Editor, CIOL.

 Dr. H.V. Ananda holds a Doctoral degree (Ph.D.) and Master of Engineering degree in Electrical Communication engineering from Indian Institute of Science, Bangalore. He has over 34 years of experience in the field of semiconductor technology and design. In his current position as the managing director of Synplicity India, Dr. Ananda heads the India Development Centre. Synplicity is the leading supplier of innovative software solutions that enable the rapid and effective design of programmable logic devices (FPGAs, PLDs and CPLDs).

Prior to joining Synplicity, Dr. Ananda served as director (ASIC) at RealChip Communications, where he was responsible for implementation of multimillion gate ASICs. Prior to RealChip, he worked as additional general manager at Bharat Electronics Ltd. He was chiefly responsible for setting up the IC fab and design facilities there. He was a member of the team, which produced the first salable IC in India.

Right through his career, he has been actively associated with government agencies, research labs and academia involved in development of semiconductors in India. He has been a member of several committees set up by the Ministry of Information & Technology, and the erstwhile Department of Electronics, Government of India, to oversee development of microelectronics in the country and review various projects sponsored by the government at academic institutions and research labs.

Dr. Ananda is one of the key members of the University Gateway Initiative of the India Semiconductor Association (ISA) that was responsible for signing an MoU with VTU. He has authored or co-authored several technical articles in national and international journals. He is a senior member of the IEEE. Excerpts.

CIOL: What does the Indian mindset accept – services or products?

Dr. H V Ananda: In my opinion, it is more a question of what makes more business sense.  If you look at the present scenario, I personally do not see any compelling reason for the Indian semiconductor design companies to shift from services to products.
 
Most companies started off as service companies taking the advantage of availability of talented manpower produced by good engineering colleges at a relatively cheaper cost. The entire ecosystem got built up to support this system. People are quite comfortable in doing this kind of work and the business has been quite good. Even today, it appears to be same scenario.

However, going forward, as the trends are changing such as strengthening of the Indian Rupee, increase in wage bills, etc which are affecting the margins, these companies may also need to change and switch over to developing products. It should, however, be borne in mind that selling a product is not easy unless the product has a lot of differentiation.

CIOL: What are the reasons for the lack of product development companies from India?

HVA: At the end of the day, the semiconductor products finally need to end up in a finished system or equipment. There are very few companies in India who develop their own final systems/equipments from scratch, especially in large volume markets such as consumer products. Moreover, this volume is distributed over a number of varieties.

If any Indian product development company needs to add their semiconductor products into many or all of these consumer products, they would need to develop different varieties of products themselves, where volumes for each may not be so high. Economy of scale may not, thus, justify product development. Further, to sustain in the marketplace, given the present open economy situation, it is necessary to keep bringing out newer products in a very short time frame to meet the shrinking market window.

In my opinion, our ecosystem is still not developed to enable such fast development cycles to take on the competition from multinational product companies who service the global market. The above argument holds good for Indian companies who would want to develop and market their own semiconductor products. However, Indian companies have been successful in executing product development on contract basis for other clients, although such companies are few in number.

CIOL: What should semiconductor startups do for becoming product companies?
 
HVA: As I said before, for a product company to be successful, their products should be innovative or cheaper without compromising on the quality. In order to be innovative, they should definitely invest in development of new and emerging technologies. They should look for tie-ups with academic institutions with good research facilities. You would also need to have innovative processes in-house to make the product cheaper. If the startups can keep this going, it may be possible to have successful startups in India.
 
CIOL: Do Indian companies have the wherewithal to enter product domain at all?
 
HVA: As long as they have a product, which can make a difference in the market today, they can enter the domain. There are lots of other companies in niche areas today who can service a particular segment in the market. It may be not a huge market, but then, they show their own differentiation. That is the reason I like to stress upon product differentiation and again it would be difficult to create a huge product company in a short run.

CIOL: All this noise about fabs – do people even realize what it takes to build a fab?

HVA: In my long career of more than 35 years in the semiconductor space, I was associated with building and running fabs for quite some time. Although the fabs I was associated with may not be state-of-the-art from the present standards. However, I know the problems associated with running a fab in India.

Wafer fab facility is extremely intensive in terms of initial capital expenditure, running costs and demands the highest quality of infrastructure like power, water and gases. A good ecosystem consisting of companies supplying specialty gases, chemicals and special materials, many of which have limited shelf life, need to be set up to support the fab. Setting up of this system is expensive and is dependent on the economy of scale of operation.
 
CIOL: In that case, does India even have the capability for sustaining a fab? Where is the market?

HVA: As I mentioned before, the requirement of semiconductor components in India is scattered over many different types and with limited volumes. Hence, the economy of scale may not justify having a wafer fab facility to cater only to the Indian market.

On a global scale, I do not see how an Indian fab can provide advantages in terms of lower costs and better yield compared to the other already established fabs. India has been very successful in the VLSI design sector mainly because of the availability of skilled manpower. However, the type of skill sets needed for a wafer fab facility is totally different compared to that needed for VLSI design, and such skill sets are presently not available in India.

CIOL: Is there a case for homegrown EDA companies – why vs. why not?
 
HVA: First of all, EDA is tough. Not many people are working in the EDA space. Many of the innovations in the EDA space abroad have come out of University research. The elite academic institutions in India should focus more on research in the EDA space and take the lead.

I do agree that a lot of work is going on today on in the semiconductor space, but most of them are on design related issues. However, if you are really looking at developing an EDA product, which involves development of algorithms, optimization techniques etc., then you need to focus on a whole set of different things.

In India, Synplicity works with over 100 colleges all IITs, IISc and NITs for this program and we provide all our world-class products to any college accredited to recognized universities at an extremely affordable price. These products are full-blown unabridged products sold commercially to others. The intent is to enable students and professors to have easy access to the best in class products. I am sure that other EDA vendors also will have such programs.

CIOL: Finally, to what extent can manufacturing help the Indian semiconductor industry?

HVA: Manufacturing will always help. There are many companies starting manufacturing here. If some of these large companies can identify specific semiconductor components and require them in large volumes, then it will make sense for certain companies to manufacture such components in India.