L&T Semiconductor has a pipeline of 20 automotive clients so far: CEO Sandeep Kumar

Sandeep Kumar, CEO of L&T Semiconductor Technologies

Back in 1986, five fresh-out-of-college students who had joined semiconductor giant Texas Instruments (TI) were about to make a breakthrough. When Apple, Microsoft, and the like were making microprocessor units (MPUs) used in servers, desktop computers, and mobile computing devices, Dallas-based TI formed a new team and tasked the young guns with making microcontroller units (MCUs) that go into refrigerators, washing machines, and cars, among other devices. TI's focus was more on microcontrollers, as they don't require complex operating systems to function, whereas microprocessors do.

“We were successful in making the first cell phone chip, the Adaptive Voltage Scaling (AVS) microcontroller, and the engine controller. We built it from zero to roughly $700 million in business in 10 years,” recalls Sandeep Kumar, CEO of L&T Semiconductor Technologies. After completing his BTech from IIT Delhi, Kumar, who hails from Chandigarh, moved to the US and did his MS in Electrical Engineering at the University of Cincinnati before joining TI.

Cut to 2021, and life came full circle when Kumar's classmate from IIT, heading L&T’s defence business, suggested he build a semiconductor company in India. Kumar put together a plan and approached multiple companies, including L&T. Two years later, Larsen and Toubro showed interest, but instead of investing in the company, they asked him to start the whole business for them. In 2023, L&T Semiconductor Technologies (LTSCT), a fully owned subsidiary of L&T, came to life, and Kumar assumed the role of CEO in November.

LTSCT is a fabless semiconductor company that will focus on designing chips. They will outsource the manufacturing to US-based Global Foundry, Taiwan’s TSMC, and Vanguard. Their product portfolio includes smart devices across MEMS sensors, power, analog, mixed-signal, and radio-frequency (RF) products to support automotive, industrial, and energy verticals. LTSCT has onboarded six automobile companies, and the total value of the deal is $150 million per year. “We have a pipeline of roughly 20 automotive clients in Europe, Japan, and India, and out of these, we are signing six. The rest will also be signed soon,” says 61-year-old Kumar, who lives in Austin but is currently in India.

The engineering and construction conglomerate is taking the route of "chip design" rather than "chip making" as there is a dearth of an indigenous chip design company that can build designs and intellectual property (IP) in India. There are a number of small startups working in the semiconductor design space, but none of them has made it big. LTSCT’s focus on designing custom application-specific integrated circuits (ASIC)-based chip solutions would help them carve a niche, especially with the growing boom for compute as we enter the AI era, says Neil Shah, vice president of Counterpoint Research.

In a conversation with Forbes India, Kumar spoke about the areas India needs to focus on while building the upcoming semiconductor ecosystem, L&T’s entry into the sunrise sector, and why the government’s Design-Linked Incentive (DLI) Scheme needs amendment. Edited excerpts:

Q. The Indian government is promoting chip-making in India. Why did L&T choose to foray into chip design?

Three years ago, the Indian government was very big on manufacturing, and the incentives were geared towards bringing in seven nanometres, which is a very advanced technology. My view was that nobody will bring that, and we need to build a product company. I had put together a plan for the same. After a while, the government also realised that wasn't happening. They changed the incentives to 180 nanometres for bringing the technology to the country anyhow. But the fundamental gap is still the same in my mind, which is never build a factory unless you know what product to make inside the factory.

Q. Can you tell us more about why the focus is specifically on the automobile, energy, and industrial sectors?

Most semiconductor players don't like to touch these markets because it takes three years to design the product and qualify it, and real sales will start in five years. Plus, the quality and reliability requirements are very high. The positive is that they will keep buying from you for 10 years. Also, there is a lot of room for new innovations because there are elements of sensing, edge AI, and control systems, and it is not just computation. You have a human interface, a mechanical interface, going into an electronic interface. We have to figure out the electromechanical system in the case of industrial or automotive. In the case of energy, we have to understand power distribution, power generation, and turbines. So it is not just data manipulation, like it is in the cell phone or in the PC. Here, you have to have cross-functional domain expertise.

We are building a capacity to do 10 to 15 products in parallel. Each of these products, on average, takes 30 engineers to execute in a two-year period. It takes 12 months to design, six months to test, and three months to qualify. So it will take about a year and a half to two years before you are production-ready with a new product.

Q. What is the size of the chips LTSCT will be designing?

In the automotive sector, smart analog and smart power will generally run at 130 to 45 nanometers. The smart RF and the compute engines will run somewhere between 28 nanometres and seven nanometers. And then the power devices for EV electrification get built on silicon carbide or gallium nitride, where the geometry is not so relevant. They are large devices and measured in microns, not nanometres.

Also Read: Keeping focus on Make in India and sunrise manufacturing sectors like semiconductors crucial for new coalition government

Q. Is LTSCT planning to gradually move into manufacturing chips?

This is more of a math game. A silicon fab takes $10 billion of investment. That is the lower end. It can take more. In order to make money, if you build a factory with that much capital, you need to be able to sell at least $1 billion worth of product from that factory every year. So I need to know how to sell $1 billion of product before I build a factory. I can get there in seven-eight years; that's when I need a factory. Now if I know how to get there in the next three-four years with high confidence, then I can start building the factory in three-four years, so that it is ready in seven-eight years, because it takes four years to build it.

Q. LTSCT has invested $100 million so far. Have you taken support from the government's Design-Linked Incentive (DLI) Scheme?

The answer is no, and partly the answer is no because the DLI scheme actually, at least until now, doesn't allow companies like L&T to take that funding. It is restricted to some startups, and there is a very strong definition of that. Maybe they will revise it, and then we will see if it makes sense to leverage it. L&T has deep pockets and enough capital, so I don't think capital is the issue for us. We easily have a three-year runway, out of which we’ve consumed nine months.

We're looking for acquisitions. We recently acquired a semiconductor design startup, SiliConch Systems. We are evaluating multiple more, and all those are driven to accelerate our product portfolio, revenue ramp, and talent acquisition on all three fronts.