Having just secured Rs 107 crore in a funding round led by Zoho Corporation and Unicorn India Ventures, Jyothis Indirabhai, Co‑founder and CEO of semiconductor company Netrasemi, is upbeat about the company’s growth prospects.
He believes “this is the time” to shine for semiconductor startups in India, which are witnessing increased investor interest and unprecedented government backing. Buoyed by recent developments, Indirabhai is eager to steer Netrasemi forward with cutting edge in-house innovation.
The Thiruvananthapuram-based startup’s ambition is bold and clear: to unleash AI’s full potential by embedding it directly into edge devices through a “unique set of chips” built with a sharp focus on cost, energy efficiency, and high performance.
An edge device is a computing device located at the ‘edge’ of a network, close to data sources or users. Data processing and analysis occurs directly on the device. Edge AI, also known as AI on the edge, refers to deployment of AI algorithms and models directly on edge devices, rather than relying on centralised cloud servers.
In a conversation with Shradha Sharma, Founder and CEO of YourStory, Indirabhai explains, “Edge means you are bringing intelligence on devices and on‑premises. Instead of sending sensor data or key information to the cloud or servers, you are bringing that intelligence in the device itself.”
Many chip companies are entering the edge AI space. Netrasemi too is in that race, says Indirabhai, and hopes to create a real “edge” in the market.
In an era when milliseconds matter, on-device processing slashes latency, bolsters privacy as data never leaves the unit, and delivers impressive energy efficiency. Whether it’s a drone deftly navigating a crowded warehouse, or a security camera instantly recognising a threat, edge AI has become the frontier of innovation where real‑time response, data sovereignty, and sustainability converge.
Chips from the ground up
Behind this vision lies an in‑house strategy. Rather than purchasing off‑the‑shelf cores, Netrasemi has chosen to build its own suite of domain‑specific accelerators from the ground up.
“We are building a unique set of chips, not one chip, multiple chips with one unique feature which is domain‑specific hardware acceleration for AI,” says Indirabhai.
This patented architecture spans use cases such as surveillance, robotics, autonomous drones, and industrial automation, with each chip optimised for the stringent power, cost and performance constraints endemic to edge environments.
However, this approach did not materialise overnight. Netrasemi, founded in 2020, deliberately delayed its first tape‑out until its IP portfolio and the accompanying business plan were rock‑solid.
In chip design, ‘tape-out’ refers to the final stage wherein the finalised design data is sent to a fabrication facility for manufacturing.
“We created an IP portfolio first… made a good business plan ready and convinced ourselves before starting this (company),” recalls Indirabhai.
Indirabhai—along with co‑founders and semiconductor veterans Sreejith Varma and Deepa Geetha—spent years refining their work and hiring the right talent. This patient R&D phase laid the bedrock for Netrasemi’s success.
R&D to real-world impact
Funding has arrived at a pivotal moment as Netrasemi prepares to pivot from R&D to full‑scale commercialisation. The latest funding round is part of a larger, national undertaking.
“This (funding) came at the right time where the company is planning to scale up… We are in a juncture where we are switching from R&D phase to commercialisation phase, and that needs a lot of money, especially since it is a semiconductor journey,” observes Indirabhai.
In an industry where a single mask set can cost crores, capital is the oxygen that fuels progress. The latest infusion will underwrite Netrasemi’s scaling of R&D, bolster manufacturing and marketing capabilities, and accelerate the rollout of four advanced SoC (system on chip) variants, featuring AI and video analytics, into production.
SoC is a single chip that packs everything a device needs—such as the processor, memory, and other key components—into one compact unit.
Netrasemi has completed development of two advanced video-enabled edge-AI chips, which are now in the tapeout stage. Over the next 12–18 months, the company plans to bring its SoC families into production and kick off R&D on next-gen, ultra-high-performance chips for edge servers. Launches are expected by early 2026.
As Indian chipmakers step onto the global stage, they find themselves in head‑to‑head competition with entrenched powerhouses in the United States, Israel and China.
“You are competing against the US, Israel, then China where heavy investment is going on this,” admits Indirabhai.
However, he stresses that India cannot simply replicate the US playbook. “They have enough successful people around, enough stories in the semiconductor space for them to grow. Because today, if you look at it, the largest company in the world is a fabulous semiconductor company called NVIDIA.”
Indian challengers, he says, must craft lean, focused roadmaps centred around edge‑AI niches, leverage homegrown IP, and forge alliances that maximise speed without sacrificing long‑term viability. Only by playing to these unique strengths can they disrupt markets long dominated by multinationals, he adds.
Recognising these challenges, the Indian government has enacted a series of design‑linked incentives (DLI) and support schemes, such as the Chip‑to‑System (C2S) programme, to underwrite the staggering upfront costs of chip development.
“Once the government put that focus, they enabled startups with different kinds of programmes like DLI, C2S,” explains Indirabhai.
These initiatives reimburse a proportion of design expenditure, subsidise prototyping, and unlock matching grants, thereby easing the burden on nascent ventures. By de-risking the earliest stages of R&D, policymakers hope to cultivate a thriving semiconductor ecosystem that can sustain long‑term innovation, rather than driving talent and capital to gravitate exclusively towards software or services.
For a nation that currently imports nearly 100% of its semiconductor chips, this support could be transformative.
A miniscule part—maybe about “0.01%” according to Indirabhai—is made in India, in very strategic domains such as military, space, and similar areas, but the remaining is imported. This dependency not only imposes significant import costs, but it also stifles domestic aspirations to become a genuine product‑centric economy.
India as a country is not able to produce, manufacture, or do R&D, primarily because “we don’t have key components,” says Indirabhai, adding that one of the most important components for an electronic product is the chip. And that is missing here. Electronics design and manufacturing—areas of high margins—are tied to foreign suppliers.
If Netrasemi can harness its latest funding to move from prototype to mass‑market ASICs (application-specific integrated circuits), it may mark a major moment in India’s “Make in India” dream.
But this leap is fraught with hurdles: each new generation of chips demands fresh mask costs, fresh engineering cycles, and fresh validation budgets, all against relentless market timetables.
Amid these high‑stakes dynamics, Indirabhai extends a note of caution for aspiring semiconductor entrepreneurs: resist the allure of premature capital, which can lock a startup into an untested concept. “The worst thing that can happen to you is premature investment… Once you get a premature investment, then I think you are locked to a premature idea,” he warns.
Unlike software, which can pivot in weeks or months, semiconductor products often require seven to eight years of iterative R&D before they reach commercial maturity, he says. Founders must, therefore, remain steadfastly R&D‑focused, cultivate deep domain expertise, and seek mentors who understand the marathon‑like nature of chip development.
With its in‑house IP, freshly bolstered coffers, and a clear view of competitive global factors, Netrasemi, Indirabhai believes, is well-positioned to ride the edge AI wave.
“You should expect really cool chips from India… (built) for India, and for the globe.”
Edited by Swetha Kannan
