Auto-Arch Tournament vs stagewise

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.

stagewise is an open-source AI coding agent built specifically for frontend work on existing codebases. Unlike agents that only read source files, stagewise runs in its own browser environment — it can see the live DOM, observe console errors, and interact with the actual rendered UI before making code edits. This closes the loop between "here's the code" and "here's what the user actually sees." It's BYOK (bring your own key) with support for any major LLM, and is explicitly designed for established projects rather than greenfield apps — the agent understands how to navigate a real codebase and propose minimal, surgical edits. Launched April 16, 2026 and hit #6 on Product Hunt with 181 votes. The core insight is that frontend bugs are often invisible to agents working from source alone: a CSS cascade issue, a hydration mismatch, a console error — none of these appear in static file reads. stagewise makes these visible. For teams maintaining large frontend codebases, this is the agent setup that actually matches how human developers debug: look at the thing, then fix the code.