Auto-Arch Tournament vs CatDoes v4

Auto-Arch Tournament is an autonomous research system where an AI agent iteratively proposes, implements, and validates microarchitectural improvements to a RISC-V CPU. Starting from a standard 5-stage pipeline, the loop runs hypotheses in parallel, each going through formal verification (53 symbolic checks), cycle-accurate simulation, multi-seed FPGA place-and-route, and CoreMark CRC validation. Only hypotheses that beat the current champion get merged; everything else gets discarded. Starting from 301 iterations/second, the system hit 577 iter/s (+92%) across 73 attempts in 9.8 hours — producing a design 26% faster and 40% smaller in LUTs than the baseline. The insight the author drives home is that the real innovation isn't the AI agent — it's the verifier. The orchestrator is hardcoded to prevent agents from manipulating their own evaluation gates, a simple but critical design constraint that turns a creative process into a trustworthy one. Without a rigorous verification harness, agent-driven optimization becomes a confidence trick. This is early but fascinating proof that AI-driven hardware design loops can produce commercially meaningful gains. The repo uses Claude Code or Codex as the coding agent, SystemVerilog for the RTL, and standard open-source EDA tooling (Yosys, nextpnr, Verilator). It's a compelling template for anyone building agentic optimization loops where correctness matters.

CatDoes v4 ships with Compose — an autonomous AI agent that runs on its own cloud computer to build mobile apps, websites, and internal tools from plain text descriptions. You describe what you want, Compose plans the work, writes code, runs tests, fixes its own errors, and deploys — even after you close the browser tab. Every project comes pre-wired with a full backend stack: database, authentication, storage, edge functions, and real-time events. The v4 release focuses on higher reliability and GitHub integration for developers who want to export and own their codebase. Free plans start at 25 credits; paid plans begin at $20/month with more projects and higher cloud limits. What distinguishes CatDoes from the crowded AI app builder space is the "own computer" framing. The agent doesn't just generate code for you to paste — it has an execution environment where it can actually run and debug the app, catching errors before you see them. Whether that closed-loop debugging holds up in practice for complex apps is the open question.