hn.today

Show HN: Neoma P-CAM – Sub-millisecond physical equilibrium and 64-spin Ising

github.com5 points0 comments
Screenshot of Show HN: Neoma P-CAM – Sub-millisecond physical equilibrium and 64-spin Ising

Neoma P-CAM is presented as a room-temperature analog-silicon accelerator for fast equilibrium calculations and optimization. Its repository combines a Rust/Axum service, hosted MCP server for AI agents, Python SDK, API examples, and integrations for areas such as battery analysis, robotics, chemistry, and financial execution. The central argument is that physical silicon can solve nonlinear equilibrium problems and 64-spin Ising models with deterministic, sub-millisecond responses, while avoiding pseudorandom-number generators. It also emphasizes a hardware-provenance system: execution results can carry cryptographic seals tied to a silicon root of trust.

The repository reports several performance and validation results. It cites a 64-spin Ising solver, typical response times in the tens of microseconds, and a GPU-offload test said to reduce power draw by 83.08% while lowering GPU utilization. For battery analysis, it describes processing all 168 discharge cycles of NASA’s B0005 lithium-ion dataset, reporting a mean latency of 16.3 microseconds and individual provenance seals. These figures are presented alongside audit files and reproducible scripts. The project also offers tools for market-order slicing, robot-joint control, and AI-inference middleware, positioning P-CAM as a co-processor for cloud and edge workloads rather than a replacement for conventional systems.

Read on github.com0 comments on Hacker News

Summary generated by AI from the linked article. hn.today is not affiliated with Hacker News or Y Combinator.

More in Hardware

The daily digest

Today's best Hacker News stories, summarized and screenshotted, one email a day.