AivexaNewsSearch
AI news for builders and product teamsChecked every hour

General Compute raises $100 million Series A, countering the neocloud disaster narrative

Collected Oct 27, 2026

General Compute, a neocloud startup that builds its cloud on inference chips from outside the Nvidia ecosystem, has raised a $100 million Series A led by Blackbird Ventures and Abstract Ventures. The round values the company at $800 million post-money, up sharply from a $60 million valuation after a May seed round. The company says a large customer deal worth $100 million in annual recurring revenue paved the way for the raise.

The timing is pointed. CoreWeave, a publicly traded neocloud that rents out AI computing power, has seen its stock fall to a 12-month low, and Elon Musk has predicted that more than 200 neoclouds will face what he called guppy-level death and consolidation while he flexes $40 billion in debt to buy compute capacity. Co-founder and CEO Finn Puklowski is unbothered by that backdrop. He said he is bullish on CoreWeave's chances and suggested Musk is talking his book to push down GPU prices, describing fears of a chip-led collapse as massively overstated. He added that Anthropic's bullish S-1 is settling fears for today.

General Compute's differentiation is its chip supply. It struck a $400 million deal to buy new chips from SambaNova, which the company describes as the first debt financing of inference chips, and has since added Positron and Cerebras as suppliers. That is a deliberate zig: while most neoclouds chased GPUs for training, General Compute built around silicon tuned for inference, the serving side of the workload where models answer user requests. For developers, inference economics often dominate the bill once a model is in production, so cheaper or more efficient serving hardware can matter more than peak training throughput. Relying on non-Nvidia chips also means taking on the software-porting and availability risk that comes with smaller vendors, a likely trade-off.

The physical buildout follows the same contrarian logic. Rather than constructing large data centers from scratch, General Compute uses smaller power sources, rents space inside other companies' data centers and converts former bitcoin-mining sites into facilities that run AI models. The payoff is speed and flexibility: when chips arrive, there is somewhere to plug them in. Puklowski expects the company to deploy 75 megawatts of computing capacity in 2027 and believes it will have access to 200 megawatts of power, more than it can likely fund and supply with hardware.

Its customers are cloud companies that offer developers high-speed inference. General Compute supplies bare metal, raw servers with no software layer, and those partners bring their own offerings to market on top. James Palmer, a Blackbird principal who co-led the round, said his firm's confidence partly stems from an investment in Baseten, an inference cloud company. He described observing Baseten's growth over the last 10 months, along with demand for open-weight inference and for companies owning their own intelligence. He argued that Cerebras, Positron and SambaNova all need distribution partners who can take them to scale and navigate the physical world.

Executives at all three chipmakers praised General Compute's focus on making AI cloud economics work. SambaNova CEO and co-founder Rodrigo Liang framed the industry's real question as whether anyone can make money selling speed. Financing remains the open work: Puklowski said capital will be a full-time job for him across both debt and equity.

Why it matters: Product and platform teams choosing inference providers may soon see more capacity and price competition from clouds running on non-Nvidia silicon, though availability of those chips and the maturity of their software stacks will shape how usable that capacity is. For founders, the round suggests investors are still funding narrow neocloud plays built around rented space and specialized hardware rather than large-scale construction, which is my inference from the deal terms. The bet depends on inference demand persisting and on the smaller chip vendors scaling alongside General Compute.

Read at TechCrunch · AI

Based on reporting from the original publisher. Visit the source for full context and later updates.

Publisher excerpt

General Compute is betting on inference chips from outside the Nvidia ecosystem.