General Compute Raises $100M Series A to Build the Inference Cloud for Ultra-Fast Tokens

General Compute, the world’s first heterogeneous inference cloud, today announced $100M in Series A funding co-led by Blackbird and Abstract, along with Fuse Venture Partners, Village Global, and others. General Compute’s cloud provides AI companies access to both GPUs and frontier inference chips on-demand in a flexible contract. Its first major deployments will go live in Q1 2027, with more than 50 MW of capacity in the pipeline.

AI’s Infrastructure Problem: Skyrocketing Inference Demand

Frontier labs are moving their priorities from training to inference. The need for ultra-fast tokens is skyrocketing. Goldman Sachs reports demand will increase 24x by 2030. Google’s demand for tokens has surged 330x between 2024 and 2026. To solve this, a new generation of inference-specific chips have emerged, but the infrastructure to deploy them has lagged. This includes the physical datacenters, routing technology, financing, and software stack.

“The era of one-size-fits-all AI hardware is over. As inference becomes the largest and most complex workload in computing, the future belongs to specialized silicon,” said Finn Puklowski, co-founder and CEO of General Compute. “Our mission is to bridge the gap between the world’s most advanced alternative chips and the companies that desperately need them today.”

General Compute’s Solution: Inference Cloud for Ultra-Fast Tokens

General Compute operates as the first inference cloud for ultra-fast tokens, automating the operational complexities of operating heterogeneous hardware. By financing, deploying, and managing a diverse fleet of alternative silicon, including systems from Cerebras, SambaNova, and Positron AI, General Compute delivers high-performance compute as a simple, accessible utility. Customers can seamlessly route multi-stage workloads to the precise architecture that runs them best, unlocking unprecedented speed and scale without ever having to procure or manage physical infrastructure.

Key features include:

• Unmatched End-to-End Speed: Customers get access to frontier inference chips from Cerebras, SambaNova, and Positron AI, resulting in up to 20x faster results.
• Open-Source Compatible: Able to run nearly any model across major frontier labs & open-source models.
• Intelligent Workload Routing: Dynamically matches each agentic call to its ideal silicon profile, resulting in the most efficient token generation.
• Flexible Consumption Models: Procure compute either through a guaranteed tokens-per-second Service Level Agreement (SLA) or as dedicated bare metal.

“Hardware and power are everything in this market. A year or two ago, the bottleneck was silicon. Now it’s power, and both power and silicon supply chains are slow to scale. Finding creative ways through those constraints will decide who wins over the next several years. Finn and the team are the ultimate coalition builders, and they think in decades, not days,” said James Palmer, Principal at Blackbird Ventures.

“We’ve known the General Compute team for over a year and watched them execute incredibly well. They focused on a specialized idea, ultra-fast inference and the chips that support it, and they had a plan from day one for every infrastructure layer it takes to deliver it. When we first met, they were pitching an idea. Within months they had a big partnership with Cerebras and a debt facility of up to $400 million, and it has accelerated month by month since,” said David Kwon, Partner at Abstract Ventures.

General Compute’s Business Model, Funding, and Milestones

General Compute buys the hardware, deploys it into colocation space, and operates it: cooling, networking, energy and software. The company recently secured a $400 million debt facility from Upper90 to finance its neocloud buildout. In recent months, General Compute also announced major partnerships with Cerebras, Sambanova, and Positron AI to deploy their frontier hardware. The round also included Bold Capital, QED Fund, 645 Ventures, Carbide Ventures, Uncorrelated Fund, Endeavor Catalyst, Icehouse Ventures, MatterScale, Carya, & Plug and Play, angels and others.

Here’s what General Compute’s partners say:

• Sean Lie, CTO and co-founder, Cerebras: “General Compute understands that inference performance depends on how the full system is deployed and operated. We value their approach and commitment to fast execution at scale as a partner in delivering the fastest AI in the world to AI-first companies.”

• Rodrigo Liang, CEO and co-founder, SambaNova: “Everyone in this market is selling speed. The real question is whether you can make money doing it. General Compute bet early that inference wouldn’t run on GPUs alone, and SambaNova was there from day one, serving their first customers. As they grow, we’ll give them speed their users feel, savings that make the economics work, and scale in weeks, not years.”

• Mitesh Agrawal, CEO of Positron AI: “Our partnership with General Compute has been a phenomenal growth story spanning across two generations of Positron’s products. They’re bringing together debt financing for purpose-built inference silicon and a cloud offering that matches the hardware to the workload. That’s an important combination as customers look for faster inference and better economics. For Positron, this partnership is a path to get our architecture into more customers’ hands and scale with a team that shares that focus.”

AI labs and companies can learn more at https://www.generalcompute.com/.

About General Compute

General Compute is the neocloud for alternative chips. The company finances, deploys, and operates purpose-built inference silicon - like Cerebras - and sells the result as dedicated inference capacity under one contract and one set of SLAs. General Compute recently announced a $400M debt facility from Upper90 and is headquartered in San Francisco. Visit generalcompute.com for more.

Contact Information

General Compute

Press Contact

General Compute

Contact via E-mail