Where the $3k GPU budget went: parallelizing bounds in an AI-agent math study
DimitrisPapail · x · 2026-09-15
Dimitris Papail explains how the $3k GPU budget was spent in his multi-agent math research project: parallelizing lower- and upper-bound computations, with agents given plenty to try—different input sources, different bounding strategies, and going deeper into the complexity of the relaxation.
More from coding & agent
- Translation MCP team: the hard part wasn't the server, it was getting the model to call it — Pure-Ad7786 · 2026-09-15
- Developer runs Claude inside Siri on macOS 27 via Apple's hidden model-provider support, open-sourced — Teknium · 2026-09-15
- OpenResearch tops GitHub trending, turns Claude Code and Codex into research agents — TheMoonMidas · 2026-09-15
- CoreSpeed Launches: One MCP Endpoint for Tools, Memory and Agent Budgets — Kind-Atmosphere9655 · 2026-09-15
- Track anything cheaply: Astra + SAM 3 pipeline cuts tokens 26x — deepakns · 2026-09-15
- Subagents make coding slower and dumber: dev's first-hand vibe check — _AustinCalvert_ · 2026-09-15