General Compute secured the first-ever $400 million loan backed by inference-chip collateral (not GPUs), signaling market acceptance of specialized AI accelerators as bankable assets.

General Compute, a Boston-based AI inference cloud startup, has secured a $400 million loan from investment firm Upper90—marking what is believed to be the first major financing deal backed by inference-specific chips as collateral, rather than Nvidia GPUs, which have dominated the AI infrastructure lending market.

Traditional AI infrastructure financing has centered on Nvidia GPUs, used to train large language models and widely accepted as collateral by lenders. Upper90 co-founder and CEO Billy Libby, a former Goldman Sachs quantitative trader, pioneered this financing approach in 2021 when Upper90 financed GPU purchases for data center startup Crusoe. Since then, companies like CoreWeave have built billion-dollar businesses around chip-backed loans.

General Compute's loan instead uses SambaNova chips, an Intel-backed chipmaker's purpose-built inference processors. The SN50 chips are more power-efficient than Nvidia GPUs and require no expensive water-cooling systems, enabling faster deployment across a wider range of data centers. General Compute claims the chips deliver 16 times faster inference than GPU-based clouds.

The deal reflects a broader shift toward cost-efficient AI infrastructure. As concerns about model inference costs intensified in late 2024, enterprises and developers increasingly turned to open-source models running on cheaper hardware. Companies like OpenRouter and Fireworks, providing access to open models, raised funding at high valuations. New models such as Kimi's K3 achieved coding performance benchmarks comparable to releases from Anthropic and OpenAI.

"Everyone doesn't need a supercomputer, but they do need inference and AI," Libby told *Bitcoin World*. "We think open source models are going to be important."

General Compute CEO Finn Puklowski framed the deal as more than a financing event. "This is the first signal of capital organizing itself and the fragmenting of Nvidia's monopolistic dominance," he said. The company's access to chips outside the Nvidia ecosystem provides flexibility in pricing and deployment. Other infrastructure providers, such as TensorWave, are pursuing similar strategies with AMD chips.

As alternative chipmakers like Groq and Cerebras gain traction, compute providers not locked into Nvidia contracts may achieve cost advantages in inference workloads. "There are a bunch of chips that are starting to scale that have amazing total cost of ownership, but there's not too many buyers for them," Puklowski said. The $400 million loan signals a maturing financing market, moving beyond Nvidia GPUs to broader hardware assets and potentially reshaping competitive dynamics in AI cloud computing.