Auto-Arch Tournament vs Euphony

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.

Euphony is an open-source browser-based visualization tool released by OpenAI for inspecting Harmony chat data and Codex agent session logs. It renders structured conversation timelines from JSON/JSONL files, clipboard data, or public URLs, making multi-step agentic sessions navigable instead of a wall of nested JSON. An optional FastAPI backend enables loading logs from remote sources. Licensed Apache 2.0. The debugging problem Euphony solves is real and growing: as AI agents execute increasingly long horizon tasks — dozens of tool calls, branching decision trees, nested sub-agent invocations — understanding what actually happened during a session becomes genuinely hard. Standard log formats are machine-readable but not human-comprehensible. Euphony renders them as interactive conversation timelines that preserve the temporal structure of the agent's reasoning. OpenAI releasing this as open-source is slightly surprising — it signals genuine investment in developer tooling transparency rather than keeping all agent debugging inside a proprietary platform. The timing aligns with broader industry pressure to make agentic systems more auditable and interpretable. For teams running Codex in production or building on OpenAI's agent APIs, Euphony is immediately useful as a debugging and post-session review tool.