Has the hunt for AI compute uncovered the next Cerebras?
General Compute raises $15 million to offer specialized AI processing power using SambaNova chips, aiming to solve AI compute obstacles.

The demand for computers to run AI models has accelerated, but two major obstacles remain: getting the right chips and getting them into data centers. General Compute, a new inference neocloud, thinks it has the answers. General Compute has raised a $15 million seed round at a $60 million post-money valuation, led by FUSE VC with participation from Carya Venture Partners and Village Global Ventures.
The company specializes in renting out AI processing power, particularly during the inference phase when models are running and responding to users. The right chip for AI inference is a key question. While GPUs are in high demand, they're not the best-suited for running trained AI models.
A new class of chips is being designed specifically for inference. Nvidia's $20 billion Groq transaction and Cerebras' $57 billion IPO point to this trend. General Compute's co-founders, CEO Finn Puklowski and CTO Jason Goodison, have turned to specialized chips built by SambaNova, an Intel-backed chipmaker focused on inference.
SambaNova's new chips, set to be released this year, promise to outperform GPUs and other specialized chips. They offer a more flexible architecture and use more memory to store context during inference calculations. SambaNova claims that its chips will generate 600 to 700 tokens per second, compared to 250 tokens per second for GPUs.
General Compute has $300 million of SambaNova's SN50 chips on order and will be the first neocloud to deploy them. These chips also solve the problem of where to put them. They're air-cooled, consume less power, and can be installed in existing data center facilities without new infrastructure investments.
General Compute is pursuing colocation deals with data center providers and crypto miners looking to repurpose their infrastructure. General Compute launched its cloud offering last week, claiming it's already the fastest at running MiniMax 2.7, a powerful open-source LLM. Joe Hasselmann, a venture investor who backed Groq in 2021, sees parallels between SambaNova's partnership with General Compute and CoreWeave's relationship with Nvidia.
The question remains what kind of computer architecture will capture the most value in the AI future.
Source: TechCrunch