Bridging The Service-Product Gap In Indian Semiconductor Industry

Three things are important to build a semiconductor product company from ground up ­ availability of well-researched ideas, access to talent, tools and technologies, and most importantly an ability to fail often. Let us look at each one of them closely. With institutes of global repute like IISc and IITs, we always had access to high quality research. With democratization of knowledge globally and easier access to technologies, more and more engineering institutes are carrying out active research at the leading edge of the technology. On talent front, India always had this as its primary advantage due to high availability and lower cost. However, majority of the chip design talent today is largely concentrated within the captive outsource units of global semiconductor companies. There is a big change required here in terms of risk taking ability of people to move out of the comfort zones of these outsourcing units to unexplored territories of product start-ups. For this to happen at scale, one or two big success stories may be needed from India-grown chip design start-ups.

The problem starts, however, with tools. Unlike the software world, where start-ups have easy access to a wide range of open-source components and cloud-based infrastructure for rapid prototyping, the chip design world is crippled with high cost of experiments. EDA tools are very expensive for start-ups to afford. There is no common design infrastructure that students, researchers and entrepreneurs can leverage to carry out experiments. Most institutes have low-cost access to design tools, but little know-how to build a design platform integrating these tools. For rapid experiments to happen, the need of the hour is to build a cloud-based common design platform that can be accessed by anyone who wants to carry out design experiments. No entrepreneur should be blocked just because there is no access to cell libraries, PDKs, flows or tools. Once we democratize infrastructure, just like what we did in the software world, we can expect a tsunami of experiments and may be a few successful ones that can lead to products and companies.

The capital has always been a problem with chip design start-ups. Unlike software and services, chip design start-ups carry higher risk of failure while at the same time it is expensive to build a product prototype. This is an area where industry needs some help from government. We have been seeing some activities from government on this front in terms of setting up dedicated funds. The cost of prototyping can be significantly lowered by creating common design platforms with voluntary development participation from eco-system players such as students, researchers, faculties, entrepreneurs and industry experts. The research funding in institutes can be funnelled more towards consumable research activities incentivizing product launches.

While all these are necessary to build a product company, they are still not sufficient because of huge entry barrier in any established market. What we need is a trigger that will create a level-playing field between the established players and the newcomers. Twenty years back, wireless was the trigger that gave the world many new global companies. We are back to a whole new inflection point for semiconductor industry ­ `Connect everything and make each one of them autonomous'. It is a huge space for innovation and customer value creation. We are just scratching the surface with autonomous vehicles, smart homes and autonomous factories. With massive computing, storage and communication power in our hand, we are only limited by our imagination on what is possible. Many of these applications may not even need expensive leading-edge technologies to build products, but may need deep domain knowledge, system-level thinking and right skills. We have painfully built the necessary skills in India over past two decades. Now is the time to leverage all our strengths to Make in India, for the World!