AI agents narrow formally verified compiler CompCert's gap to gcc -O2 from 28% to 6%
swarat · x · 2026-11-17
- CompCert, a formally verified C compiler with machine-checked proofs, ran 28% slower than gcc -O2. Aananth used AI agents to close the gap.
- Key insight: disabling several GCC optimizations that CompCert's semantics rule out barely changed the gap, pointing to optimization opportunities within existing semantics — loop-invariant code motion and better x86-64 instruction selection.
- Agents implemented the passes and wrote machine-checked proofs in Rocq. On CompCert's own benchmark suite, the default configuration narrowed the runtime gap to gcc -O2 from 28% to 6%, with smaller gains on five larger held-out programs.
- The post also highlights where human judgment mattered most: diagnosing interactions between passes and scrutinizing changes to the trusted model.
Related event: Closing CompCert's 28% performance gap with gcc using AI agents(2 posts)→
More from coding & agent
- Developer uses his GitHub Radar project as an agent entry point to decide what to work on next — dSebastien · 2026-11-17
- Lectern CLI turns agent-written Markdown into narrated explainer videos in under 30s — gabe_ragland · 2026-11-17
- J.Doe AI assistant goes viral: personal context from recordings, screen, and Gmail — HeyZoyaKhan · 2026-11-17
- Expert: Orgs should monitor customer channels for AI incidents like they monitor production logs — simpsoka · 2026-11-17
- IBM's Tejas Kumar on Building Agent Harnesses From Scratch: Guardrails and Daiku — TejasKumar_ · 2026-11-17
- Cohere Labs-backed FeatureGraph standardizes scientific time-series analysis into a single source of truth — Cohere_Labs · 2026-11-17