Hareesh Chandrasekar's AGNIT Semiconductors: The 18-Year Patient Capital Play Behind India's Chip Dreams
How a Bangalore startup turned academic R&D into strategic advantage, raising just $4.87M to compete in the $1.5B defense semiconductor market
The billion-dollar question in deep tech isn’t “how fast can you move?” It’s “how long can you wait?”
On the Indian Institute of Science campus in Bangalore, an 18-person team is attempting something that typically requires $20 billion fabs and thousands of employees. AGNIT Semiconductors, founded by Hareesh Chandrasekar, has spent nearly two decades developing gallium nitride (GaN) semiconductor technology. They’ve raised only $4.87 million in total funding. Their nearest competitors have raised 60 times more capital.
The company’s next milestone, 100,000 chips sold in 12 months, represents a 1,763-times increase from current production levels. If they succeed, it validates an entirely different playbook for deep tech commercialization.
Check out the video of the conversation here or read on for insights.
The Institutional Subsidy Hack Meets Geopolitical Timing
Most semiconductor startups burn through Series A funding just to prove basic science. AGNIT took a different path. The company emerged from 18 years of research at IISc’s Centre for Nanoscience & Engineering, research funded by India’s Ministry of Electronics and IT, which invested approximately ₹300 crores to establish a gallium nitride pilot foundry on campus.
This created unusual capital efficiency. The expensive, decade-long technical risk had been absorbed by institutional grants before private investors were approached. The $4.87 million raised, including a $3.5 million seed round in September 2024, is deployment capital, not invention capital. The company holds 20+ patents covering the entire GaN value chain before venture capital ever entered the picture.
Hareesh’s journey mirrors this patient approach. After earning his PhD from IISc in 2016, winning the Best Thesis Prize for Applied Research on GaN devices, he completed postdoctoral research at Bristol and Ohio State. He also spent time as a chip designer at IBM India, an experience he now calls “what feels like a past life,” but one that gave him essential commercial exposure.
Then geopolitics intervened, not as disruption, but as validation.
In July 2023, China announced licensing requirements for gallium exports. By December 2024, this escalated to a full U.S. export embargo. China controls 87% to 90% of global gallium reserves, and gallium is irreplaceable for GaN semiconductors in defense and telecommunications. The GaN defense sector alone is projected to reach $1.5 billion by 2026.
For AGNIT, which had already focused on strategic sectors, this wasn’t a pivot. It was confirmation.
There’s nobody working in the strategic sector whom we wanted to get a meeting with that we did not get a meeting. That just goes to show that the problem we are trying to solve is a very immediate problem for them. They really need a solution.
China’s move to weaponize critical materials inadvertently created the market conditions that make indigenous suppliers commercially viable. When export controls prevent access to foreign technology, companies don’t compete on price. They compete on availability and trust.
Defense First, Consumer Later
AGNIT’s product pipeline includes three chips in field trials with defense contractors: jammer systems that disrupt drone communications, radio link amplifiers for secure field communications, and video transmission chips for long-range surveillance drones.
If you use a conventional silicon based solution today, maybe an equivalent sized silicon part will give you about 20-25 kilometers of range. But what do you do if you want 40 kilometers, 70 kilometers, 90? In that case, using a gallium nitride power amplifier makes a lot of sense.
This strategic sector focus is the opposite of what most hardware startups pursue. Consumer electronics offer 10 to 100 times the volume potential. But AGNIT chose the hardest customers first, those with the most rigorous specifications and longest qualification cycles.
This follows a classic military-to-commercial playbook. Defense contracts provide high margins that subsidize achieving manufacturing quality at scale. They force early resolution of reliability problems that would be catastrophic in consumer markets. Following validation in defense and telecom, AGNIT plans to enter the electric two-wheeler market, where India’s electrification push creates enormous volume opportunities and GaN chips offer 3% to 4% higher efficiency than silicon alternatives.
But that consumer pivot depends entirely on first proving the technology and supply chain work at defense volumes.
Fab-Lite Strategy, Unicorn Competition
AGNIT describes itself as “fab-lite,” controlling design and IP while relying on external foundries for volume manufacturing. Certain specialized process steps, particularly the epitaxy that deposits GaN layers at atomic-level precision, remain at the IISc pilot facility where the company can iterate quickly for custom applications.
For the strategic and telecom sectors, the foundry has enough capacity to actually manufacture here. But for consumer use cases, fast chargers, solar inverters, you’ll need tens of millions of units a year.
Setting up a GaN foundry costs approximately $500 million, compared to $20 billion for cutting-edge silicon facilities, because GaN operates effectively at 50-nanometer to 500-nanometer scales rather than the 3-nanometer processes that silicon has reached.
The company’s $13.2 million valuation exists in stark contrast to competitors: Cambridge GaN Devices has raised $59.4 million, while Diamond Foundry has raised $315 million and achieved a $1.8 billion valuation. AGNIT can’t compete on capital deployment. It competes on scarcity and necessity, its 20+ patent portfolio and indigenous supplier status creating technical and regulatory moats.
We started out in a world where these systems also have a lot of control circuitry, logic and memory, all of that stuff is still on silicon. That’s what’s telling the gallium nitride, you turn on, you turn off. But gallium nitride is what’s doing all the amplification.
The company isn’t replacing silicon entirely. It’s targeting specific applications where GaN’s advantages in power density and efficiency justify higher component costs.
The 100,000-Chip Test
As of March 2024, AGNIT reported revenue of $56,700. The 100,000-chip target isn’t just a revenue goal. It’s validation of the entire business model: that foundry partnerships work at scale, supply chains deliver consistent quality, and customer qualification cycles can be navigated successfully.
The execution challenge is severe with 18 employees, roughly 5,555 chips per team member per year. Each chip passes through hundreds of manufacturing process steps. Quality control in defense applications allows essentially zero margin for field failures.
Once you have that, then essentially think about it as the foundry shipping you back your chips. Now they’re all in physical form. Earlier it was in design, a bunch of zeros and ones. And now you need to start the process of assembling it, packaging it, testing it.
Success likely leads to Series A funding at significantly higher valuation and validates the military-to-commercial strategy. Failure would leave the company as an IP holder with unproven commercial execution.
AGNIT’s trajectory intersects with India’s broader semiconductor ambitions. The India Semiconductor Mission targets 5% of global semiconductor production by 2030, with a $1 billion Deep Tech Alliance focused on semiconductors. Hareesh envisions AGNIT becoming the indigenous technology provider for future Indian GaN fabs, not just a chip designer.
A person with a new idea is a crank until the idea succeeds. We’ve spent nearly two decades being that crank. Now we find out if the idea succeeds.
The patient capital approach that allowed AGNIT to build technical moats without dilution now confronts commercial execution reality. The next 12 months determine whether 18 years of preparation was enough.
Listen now!
Other ways to listen:
Your Feedback matters
As always, I’d love to hear your thoughts! Whether it’s about this episode or ideas you’ve been playing around with, shoot me an email at ad@thepodium.in. Your feedback keeps these conversations going, and I’m always up for chatting about your startup ideas too.
Until next time,
Your Host,
Akshay Datt
Exceptional piece on how institutional capital de-risks deep tech before private markets enter. The ₹300 crore government subsidy essentially converted foundry capacity from a capital expense into patient R&D infrastructure that's the inversion most seminconductor startups miss. What's fascinating is how the fab-lite strategy maps perfectly to the 50-500nm GaN process window. You're not competing on node shrinkage like silicon foundries, so the $500M foundry capex becomes manageable once volume justifies it. The 100k chip milestone isn't just revenue validation; it's proof that your epitaxy layer consistency can scale beyond pilot batches. One underexplored aspect: does the IISc facility's specialization in custom epitaxy give you an effective moat against Asian foundries that optimize for high-volume standarized processes?