Google's DuoDiscover tops NVIDIA SOL-ExecBench, 2.68x faster than production TPU kernels
lateinteraction · x · 2026-10-15
Google researchers unveiled DuoDiscover, a full-stack code optimization agent whose key trick is co-discovering the optimization target together with a cheap proxy evaluator (a "yardstick"), attacking the cost-of-evaluation bottleneck in engineering.
Results: #1 on NVIDIA's SOL-ExecBench, 2× human experts on SWE-fficiency, 1.92× speedup on vLLM, 1.44× on FreeToken, 2.68× over production TPU kernels, and it beats the best human- and AI-designed CPU prefetchers.
Related event: DuoDiscover Tops NVIDIA SOL-ExecBench, Speeds Up TPU Kernels 2.68x(5 posts)→
More from coding & agent
- Open-Agent-Email: free open-source alternative to AgentMail for AI agents — matchaman11 · 2026-10-15
- AWS Gen AI Weekly Redesign: One-Page Exec Summary, AgentCore Memory Goes Usage-Based — DigitalColmer · 2026-10-15
- Nous Research Launches 50+ Connectors for Hermes Agent, No Account Required — NousResearch · 2026-10-15
- Dev group chat: skip runtime LLM web search, index products via API instead — vxnuaj · 2026-10-15
- Tev-Gym: 1,170 dense-reward RL environments push long-horizon agentic training — RulinShao · 2026-10-15
- Microsoft's AutoEnvScaling lets agents build their own RL environments, lifting Qwen3 Terminal-Bench from 41.1 to 53.3 — omarsar0 · 2026-10-15