General Compute, an AI inference cloud startup, has secured a $400 million loan from tech investment firm Upper90, in what may be the first financing deal to use inference-specific chips as collateral. The deal was announced in July 2026.
Inference chips are built to run already-trained AI models quickly and efficiently, and are less expensive than the chips used to train models in the first place. General Compute’s infrastructure is built around the SN50 chip from SambaNova, an Intel-backed chipmaker. The company says those chips deliver 16 times faster inference than GPU-based clouds, are power-efficient, and do not require expensive water-cooling systems — allowing deployment across a wider range of data centers.
General Compute was founded by CEO Finn Puklowski and raised a $15 million seed round in May 2026 to build what is known as an inference neocloud — purpose-built AI infrastructure distinct from general-purpose offerings by hyperscalers like AWS or Azure.
Upper90 co-founder and CEO Billy Libby, a former Goldman Sachs quantitative trader, has used a similar approach before. In 2021, his firm financed GPU purchases by data center startup Crusoe, which he describes as the first loan made against the value of advanced chips. That model later gained mainstream acceptance as CoreWeave built chip-backed loans into its business and went public. “When we financed Nvidia GPUs as the first group to do that, the market was inefficient,” Libby told TechCrunch. “We could really put together something as an early participant, and kind of get compensated for the risk.”
With GPUs now comparatively well understood, Upper90 is turning toward inference-focused companies. “Everyone doesn’t need a supercomputer, but they do need inference and AI,” Libby said.
The deal may signal a broader shift in AI infrastructure financing. Puklowski framed it as more than a funding milestone: “This is the first signal of capital organizing itself and the fragmenting of Nvidia’s monopolistic dominance,” he said. General Compute’s use of non-Nvidia silicon places it alongside other infrastructure providers, such as TensorWave, which is pursuing a similar strategy through a partnership with AMD.
Source: TechCrunch