Positron AI Raises $875 Million To Fund Asimov Inference Chip Ramp
Positron AI’s $875 million Series C will fund Asimov tapeout, a 2 MW-plus engineering data centre and production of its Titan inference system.
Positron AI has raised $875 million at a $5 billion post-money valuation, giving the inference-chip company funding for its next silicon generation and production systems, Frontier Enterprise reported.
Jim Clark, known for founding Silicon Graphics and Netscape, participated in the round with NEA, Atreides Management, Andra Capital, Valor Equity Partners and SemiAnalysis Capital, the vehicle associated with Dylan Patel.
The financing gives Positron a larger balance sheet for an AI hardware market where inference economics increasingly depend on lower power consumption, lower serving cost and systems that can run continuously after models leave training.
The company’s current product is Atlas, a first-generation inference system already deployed at hyperscaler scale.
More than 50 Atlas racks are being installed at Oracle Cloud Infrastructure, with partner Parasail using that capacity for its own inference service.
Jump Trading and i3d.net are also listed as Atlas production customers.
The new capital is directed at Asimov, Positron’s next-generation custom silicon, and Titan, the multi-terabyte-memory inference system that Asimov will power.
Asimov is scheduled to tape out in late 2026, with production planned for the second half of 2027.
Titan is being positioned for long-context and next-generation AI workloads, a segment where memory capacity and deployment economics are becoming central design constraints.
Positron’s architecture is designed to run in air-cooled or liquid-cooled data centres at different rack densities.
The company also frames the design around avoiding dependence on constrained HBM or CoWoS supply chains, two bottlenecks that have shaped AI accelerator availability across the market.
The financing will fund Asimov’s tapeout, the bring-up of a 2 MW-plus engineering data centre and emulation platform, and Titan’s production ramp.
CEO Mitesh Agrawal said Atlas deployments had shaped what inference customers need, and the next step is to tape out Asimov, move Titan into production and scale manufacturing against visible demand.
