Image credit: Pear VC.
Phinity announced $5.2M in seed funding, alongside its public launch. Uncork Capital led the round, with participation from Moxxie and angel investors including Jeff Dean. It follows a pre-seed round led by Pear through PearX.
Phinity describes itself as an applied research lab for autonomous chip design and discovery. Its stated mission is fully autonomous chip design, so that AI can search past the limits of human engineering and find computational systems people would not reach on their own.
The founders argue that hardware experimentation is held back by cost and time. Silicon development is a one-shot bet that takes 9 to 24 months, with separate teams handling design, verification, and physical implementation. By the time a chip ships, the workloads it was built for may have changed. With hundreds of millions of dollars at stake, teams stay cautious, which limits how much of the design space gets explored.
Phinity is building infrastructure for a closed loop in which agents design, evaluate, and revise computational systems using real physical feedback. The company claims its system produces verified designs an order of magnitude faster than experienced hardware engineers, with better power, performance, and area. It says the comparison uses the same specifications and frontier coding agents on the human side. That is a company claim, and no independent benchmark was provided.
The founders also lean on Jevons Paradox. If custom chips become faster and cheaper to build, they expect demand to grow sharply, moving from a world where a few companies design chips to one where many products get silicon built for their exact needs. They point to examples like cell-scale simulation hardware, spacecraft that reason at the edge of the solar system, and low-heat brain implants. They also point to AI models running on hardware designed for their own architecture.
Sonya and Aadi came to chip design from post-training. Sonya built synthetic data pipelines for NVIDIA’s chip design and code reasoning LLMs and did graph ML research at the AWS AI Lab. Aadi is a previously exited founder who has worked on synthetic data generation and post-training.
The team also includes a former leader of Intel’s custom silicon group, a former ASIC lead at Kepler Computing, and former NVIDIA architects who have led multiple tapeouts on advanced nodes. It also has AI researchers who have trained LLMs for chip design and formal verification. Collectively, they have shipped AI accelerators ranging from automotive NPUs to 3D-stacked datacenter silicon with HBM.
Phinity says it is already working with some of the largest foundation model labs and has reached 8-figure annualized run rate. Both figures are self-reported.
Source:
Phinity: Introducing Phinity
