Indian chipmaking does not need more capital. It needs more patient capital.

The standard complaint about semiconductor startups in India is that there is not enough capital. Having spent time on both sides of that conversation, we think the complaint is slightly wrong, and the correction matters.

There is capital. What there is not much of is capital with a holding period that matches a tape-out.

Why chipmaking breaks the normal venture clock

A software company can show you a curve in six months. A fabless semiconductor company has a development cycle measured in years, a capital requirement that lands in large discrete lumps rather than a smooth burn, and a first revenue event that sits on the far side of a customer's own qualification process — which the founder does not control.

None of that is a flaw in the business. It is the physics of the sector, and it is precisely why the resulting companies are so hard to displace once they work. But it maps badly onto a fund structure that needs a markup before the next fundraise.

The consequence is a specific, visible failure mode: a strong team raises on the promise of a first working chip, then spends the eighteen months before tape-out managing investors instead of engineering risk.

What is actually changing

Two things have genuinely shifted, and neither is hype:

• The talent stopped leaving. Senior design engineers who would once have spent a career at an overseas parent company are now startable. That is the single biggest input, and it has changed quietly over about five years.
• The policy support became money. Design-linked incentives and the surrounding programmes are real disbursements against real milestones, not an announcement. For a fabless team, that materially changes the amount of equity that has to be raised against the riskiest stretch.

What has not changed is the composition of the cap table at the earliest stage. That is where the gap still is.

Where a family office fits

We are not going to out-bid a fund on a competitive round, and we have no illusions about that. What we can do is something a fund structurally struggles with: hold a position through a development cycle without needing an interim markup to justify it to anyone.

That is the whole argument for family office capital in this sector. Not that we are cleverer — that our clock is different. When we backed a fabless team building edge-AI chips, the relevant question was never whether there would be a markup inside eighteen months. It was whether the chip would work and whether anyone would want it. Those are the right questions, and you can only ask them if your capital lets you.

If you are building chips and the hardest part of your last raise was explaining the timeline, we are probably a useful conversation.