Eclipse Ventures put $6.5 million into Cerebras Systems a few years ago. That stake is now worth roughly $2.5 billion — a return that underscores just how hungry investors have become for AI infrastructure plays.
From a $6.5M Seed to a $2.5B Payday
The venture firm originally backed Cerebras when the chipmaker was still in its early stages. Cerebras builds massive wafer-scale processors designed specifically for AI training and inference, a niche that was considered risky not long ago. Eclipse's bet paid off as demand for specialized hardware exploded alongside the generative AI boom.
People familiar with the matter say the $2.5 billion figure represents the current value of Eclipse's holdings, based on recent secondary transactions and the company's last funding round. The firm has already distributed some shares to its limited partners, but it still holds a significant stake.
Why Hardware Became the New AI Darling
The windfall signals a broader shift in venture capital. For years, most AI investment flowed into software — apps, models, platforms. But as training costs ballooned and inference workloads multiplied, investors began chasing the physical infrastructure underneath. Companies like Cerebras, which compete with Nvidia's GPUs, have become the darlings of a hardware-hungry market.
Eclipse's success is not an isolated case. Other venture firms have also seen outsized returns from AI chip and data center bets. The pattern suggests that the next wave of AI value creation may be in the silicon itself, not just the code running on it.
What the Return Means for the Broader Market
A $2.5 billion return on a $6.5 million check is the kind of math that makes fund managers pay attention. It could push more capital toward early-stage hardware startups, which typically require larger investments and longer timelines than software plays. But it also raises the bar — not every chipmaker will become a Cerebras.
The question now is whether this kind of payoff will pull mainstream venture money into AI hardware before the window tightens. Eclipse's timing was lucky, but it was also deliberate. The firm had been tracking the semiconductor space for years before placing that initial bet.
