Harshad Mehta and his team at Visicon Power are on the verge of setting up India’s first manufacturing plant for Silicon Carbide (SiC) components! The team has already partnered with leading research and academic institutions to ensure that they have access to India’s cutting-edge talent, labs, and resources! Here are some interesting excerpts from our interaction with him…
Q. What is the investment that you have come forward with? What are you investing in?
A. Our plan is to invest in building a total silicon carbide ecosystem in India. This ecosystem will address the weakest links of the silicon carbide industry and all links of the supply chain. For example, starting with the material, device fabrication and followed by device packaging. We are trying to work directly with customers going forward.
Visicon is a subsidiary of a company that I founded in the US. The India plan is completely under the brand name Visicon, and it will address three critical parts of the ecosystem in three phases. The first phase will be silicon carbide epilayering (epi), not only for our own consumption but also for other customers, mostly global. The second phase will be packaging, and the third will be fabrication.
If you look at the process, it is silicon carbide epi first, followed by device fabrication and packaging. But in terms of time, silicon epi and packaging will come earlier than fabrication. It will help us start generating revenue while we are in the process of setting up a semiconductor fab in Baroda.
We acquired a controlling stake in Ruttonsha International in 2006. The technology we are going to be using has been transferred into that name. Silicon Carbide technology will be transferred under the name of Visicon. However, for both Ruttonsha and Visicon Power, Silicon Power will continue to be the umbrella corporation.
Q. Have you planned all the investments in Baroda?
A. Most of the investments are planned in Baroda. While there will be investments in terms of technology and market development, 80 to 85 percent have been earmarked for facility development, infrastructure, and training.
Q. Any estimate in terms of units and/or revenue that you expect from the SiC plant in the next five years?
A. We expect to generate around 300M+ USD in revenue on a cumulative basis over the next 5 years and to make ~50 million devices.
Q. Is this a replica of your business model in the United States?
A. Yes, but this is not the first time. We have had this kind of a business model setup in India over the last 20 years. In 1994, we acquired a silicon power semiconductor fab from General Electric in the US. In 2006, we transported the technology and the process from there to another fab in Halol, Gujarat. So, we have been following the same business model successfully.
About a year and a half ago, we acquired a fab from Microchip for compound semiconductors. We are trying to follow the same model in a more refined way. Our plan right now is to transfer the technology from our Oregon fab to the one in Baroda.
Q. Can you share details about the key manufacturing equipment you have invested in (type of equipment, brand, model, investment) and any that you plan to invest in—in the next one year?
A. SiC epitaxial equipment will be from LPE, both pre and post Epi equipment. On the packaging side, we plan to invest in dicing saws, die attach, wire bonder, moulding and tester capital equipment. Currently, our main vendors are from Disco, K& ;S, ASM, Dage, and a few others. The total investment cost will be over five million dollars.
Q. Have you planned any timelines for the execution of these plans?
A. Yes, our timelines are very clear. We have applied for the Production Linked Incentive (PLI) scheme in India, which has been approved. We are in the process of applying for Scheme for Promotion of Manufacturing of Electronic Components and Semiconductors (SPECS). As we believe in efficiency a lot, we have already placed an order for epi reactors. Normally, it is a six months lead-time. We had placed an order in September 2020. The reactors are expected to arrive here in March or April 2021.
The equipment will be ready but as a part of the whole process, some of the technical support and services have to come from Italy. There have been some issues in Visa applications due to Covid-19. We are trying to move forward as rapidly as we can.
Q. There has been a lot of talk of why India is not fit for a semiconductor fab, but your take on the same is very different. What are the main reasons?
A. About India being ready for a semiconductor fab, my answer would be that semiconductors is a very broad space. They are used in consumer electronics to communicate with microcontrollers. Being a part of this space, what distinguishes us from others is the fact that we are small in size. Secondly, our business strategy is very simple—we target only high-end niche products.
We are not looking at high-volume or smaller-node sizes like 7 nm, 24 nm in terms of the technology. My philosophy has never been to follow me-too products. Big names already exist there and we do not want to be there in that market. Our focus is a highly niche, high-performance and high-reliability semiconductor market. The same applies for power semiconductors, silicon carbide, and for gallium nitride products.
The sector that we are focusing on right now is medium-to-high-power, energy sector, renewables, automotive, defence, and aerospace. We will be selling products under the Visicon brand in India and under Silicon Power brand globally.
Q. Can you explain the three stages and the three phases elaborately?
A. The PLI scheme is one application that includes all phases—phase 1, phase 2, and phase 3. The three phases address the critical components of the supply chain of the silicon carbide ecosystem. These start with the initial material phase, from processing to packaging under Make in India, and then working towards the end use of these products.
It begins with material, product packaging, and finally fabrication.
Q. Will you be providing these services to other silicon original equipment manufacturers (OEMs) as well?
A. In India, there is no company that manufactures silicon carbide. We will definitely look at all the possibilities while defining economies of scale by working with other OEMs. First it would be our company from Oregon, then there may be others who are looking for epi services or other services that we offer, or plan to offer.
At this moment, there is a global shortage of epi-material. Before we start to manufacture, we would provide that service to global players so that we can start generating revenue.
Q. What is the advantage that you have figured out while doing this in India, while a lot of others want to stay away?
A. The secret sauce is the fact that we want to learn to walk first before we start to run. That is what we have done all along. My philosophy is why not!
When in 2000 I wanted to transfer technology from USA to India, the biggest challenge I faced was the cost competition due to the presence of Chinese companies in the market. I had two choices to serve the same customers—either I buy components from China, or I find another high-quality lower-price manufacturing centre. I chose the latter.
We have always tried to set-up a business model where the break-even is not eight to ten years but more like two to three years.
Q. How did you handle the Chinese competition in Ruttonsha?
A. We started by focusing on the quality of the products. I experienced that myself when we had ordered some components from China. Initially when we asked for samples, we got excellent products. But when the bulk of 950 components came, our customers returned 900 of those. This is the problem that we addressed at Ruttonsha.
There is a team of 150 people at Ruttonsha that makes sure that every component is tested before it is shipped to the customer. If we are successful in the plan we have made, we will employ around 80 to 100 more people in the same facility in the next two to three years.
In fact, we have recently inaugurated our facility at Halol and the ceremony was attended by people from different IITs in India. We are trying to coordinate with an integrated resource plan so that we can take care of the things that we need, as we need them.
Q. When you say resource, what all are you referring to?
A. By resources I mean manpower—students, graduates, and post-docs. More importantly, the government of India has invested heavily in providing state-of-the-art equipment to a lot of universities. We also want to make sure that their facility is fully utilised. We can help them in training the students and we can use that equipment on some kind of service basis.
We have already signed contracts with IIT Delhi’s (IIT-D) Rajendra Singh. He is working with us to help develop processes in terms of reliability and others. There is equipment at IIT Delhi that we can use for research and development and, in turn, train the students on a practical basis.
The professors who came to the inauguration ceremony complimented us saying that their students can now practically see how the manufacturing takes place. Emroj Hussain, who is one of our first epitaxial engineers at the Halol plant, got his PhD from Tata Institute of Fundamental Research (TIFR) and was confused whether to stay in India or take up a job in the USA. He was very happy when Visicon offered him a job.
Q. Do you think India has the manpower for such projects?
A. I think India has the resources needed. When there are so many Indians working outside India in this field, I think resources are available.
Now, no matter how big of a university you come out of, there is always an aspect of practical training. What you learn in books and how you implement the same is a little bit different. Being able to work with IITs, I think, is a step forward in teaching those students how to work with the needs of the industry.
Q. Any immediate hiring happening at your firm?
A. We expect to add around 200 people in the next three years.
Q. Is there an opportunity for more universities to collaborate with you? What can they provide and expect in return?
A. In any kind of a research collaboration, technical ability of both the partners is always the number-one aspect. What we need to do is find core expertise of a university and then work collaboratively on a research project. We do not want to do the same things at different places.
Our aim is to work with centres of excellence and guide them in the best possible pragmatic way. We are open to talks with universities and academic institutes, and collaborations in any way we can.
Q. At what stage did your team start interacting with academia to support your research & development initiative? Was this being done at your earlier firm itself, or was a different connection established for the new firm?
A. I believe in efficiency and leverage working together as everyone achieves more as a team. I have always followed this model since my first job at Electric Power Research Institute (EPRI) in the eighties. Working with academia helps us optimally explore state-of-the-art solutions coupled with my pragmatic solution-driven approach. We aggressively look to create manufacturable, usable, sellable technology (MUST). I also believe and follow the integrated globalisation model.
We did initiate our interaction with IIT Delhi, even before our Visicon plan was put on paper. Besides IIT Delhi, we are also engaged with both IIT Bombay (IIT B) and TIFR and shall be leveraging some of their expertise with SiC manufacturing, which is a new technology.
In the US, I also work with the best and the brightest of minds at universities and am honoured to have known and worked with Prof. Agrawal at Ohio State University, Prof. Baliga at North Carolina State University, Prof. DeDoncker at University of Aachen (Germany), to name a few. I also greatly appreciate my mentor Dr N.G. Hingorani, father of power electronics innovations and applications, who has been a great teacher and inspiration for the last 38 years and taught me humility as well as pragmatism.
Q. Have you been initiating such tie-ups in the US too, or was it something you tried in India for the first time? If in US too, how was the Indian experience different?
A. In the US, both Silicon Power and SiCamore Semi have been working with Ohio State, NC State, and RPI, to mention a few. We lean on these organisations for materials, devices, packaging, and system development. We expect similar relationships in India as well. I truly believe that being born In India has greatly helped and enhanced my overall experience with my personal as well as professional journey. I strongly believe in and have ingrained in my life two eternal and core teachings of the Gita—Do your best and do not worry about the results! I also blindly believe that whatever happens is good for the long term.
Both my experiences in the US and India have been rewarding as we got to work with some of the sharpest minds in academia. We hope, over time, to be able to hire more students from India’s institutions.
Q. Going back to the business model, do you see any challenges in creating the material?
A. The challenges are going to be there. With the start of anything new, its qualification as a process itself is a big challenge. Even though the semiconductor processes are modern and detailed, challenges exist there too. You have to keep debugging how small things work in a new environment.
Things are also quite dependent on humidity, temperature conditions, the environment, cleanliness, and a lot of other factors. Those processes take time to perfect in a new environment. We have been planning for the same and are sure that we will be able to get over it.
The other challenge that we may come across is the availability of things like chemicals in as pure a state as we need in India. We might have to import such materials in that case. With chemicals, there is a lot of planning needed as they are not allowed to be imported through air. It becomes necessary to plan intelligently in all such cases as we have to make sure to have lead-time.
Q. By when do you plan to start the packaging part?
A. Joint secretary Gaur inaugurated the clean room only a few days ago. We already have a shell in Halol that is ready to accept the equipment. The epi system will arrive in about two-to-three months from now and we will then start putting everything together. I would say packaging equipment would start to arrive in about six months from now.
Q. Do you think the ‘not so pro-China’ mood will help you in any manner?
A. Certainly! The geo-political conditions that have developed over the last one or two years will help us. I call them tail winds. While the tail winds will help us propel forward, they are not primary factors. I am sure that even without these tail winds, we will be able to fly successfully.
Q. Is it possible that you will also be offering services to high-quantity players?
A. We would provide value-added services, materials, packaging and so forth. Where we feel we may not be able to go after is very high-volume, relatively lower-power silicon carbide chips which other companies are doing through billion-dollar equipment right now.
Q. Which companies would you be typically competing against in this case?
A. CREE, STMicroelectronics, and the likes of them. They are high-volume ones. Whenever a new entity comes, the bigger companies, whether intentionally or unintentionally, try to cut prices publicly. This is done to diminish the progress of the new company. As we go high-power, our first few steps would be lower-power, which will compete with them.
Our strategy would be to not compete head-to-head but provide some smart intelligent solutions to the customers that would use our products.
Q. Does that mean that you are looking to provide customers with complete solutions including reference designs?
A. Absolutely! Silicon Power is a vertically-integrated company today. Not only do we design our own components but also manufacture them. Silicon Power also packages them and provides a full system that ultimately goes into the end use. We have been installing complete power-management systems back in the US.
Q. Do your plans include gallium nitride components?
A. What we plan to do is look at a technology and a product from the technical as well as the economic viability angles. When you look at the semiconductor sector, silicon material has matured over the last fifty years. The next was silicon carbide, and that is the reason it is far ahead in terms of the material and maturing technology in comparison to gallium nitride.
The latter is not there yet; you want to make sure that instead of going after one defect-free wafer, you go after manufacturable and reliable product supply chains. The second thing is when it comes to the applications, silicon carbide, in terms of performance, meets all the criteria that high-performance systems generally require. Though people are trying to push gallium nitride for power convertors, I do not see that happening, unless some big breakthrough happens.
Q. Do you find the recent policy announcements to be helpful?
A. The government policies introduced in the last eighteen months are very industry-friendly. These are very pro-growth. I think they want to make ‘Make in India’ as friendly as it can be. I understand that the government has a different set of organisations that execute the whole policy. Yes, we are very happy with Ministry of Electronics and Information Technology (MeitY), Invest in India, and others, but to work with IFCI (the erstwhile Industrial Finance Corporation of India Ltd) on the project side, you have to learn from your mistakes and implement the learnings.
Q. Who would you want to reach out to through this interview?
A. One of the most important sectors that can help us is the customers. They are buying these products and buying from outside India. We have already approached a lot of big companies working in the electric vehicle space in India. Our approach is to tell these companies that let us try to work together and develop the most optimum product for your application. This would ensure that it is not merely a customer vendor thing but a strategic relationship.