Auto-research on idle GPUs: a CoreWeave pipeline that keeps agents busy
_ScottCondron · x · 2026-10-20
The author proposes an auto-research pipeline: queue up research ideas and let agents work on them whenever GPUs aren't busy with higher-priority jobs, so idle compute keeps producing.
On CoreWeave this can be assembled from:
- SUNK (Slurm on Kubernetes) to share nodes between priority jobs and agents, with preemption
- GPU sandboxes to run agent-written code in isolation
- Spot capacity for cheap extra compute
- Dedicated inference plus Dynamo agenthints to deprioritize agents' own model calls below production traffic
Traces, evals, and training runs are visible in one place to monitor what the agents are doing.
More from coding & agent
- Decisions API integration with Live API showcased as fast GPT backend in demo — gabrielchua · 2026-10-20
- DoorDash runs 130K automated tasks a month: inside 8 AI software factories — Pavan_Belagatti · 2026-10-20
- DeepSeek-V4.1-Flash-1017 Hits Qoder Day One, Free at 0.0x Credits Until Oct 24 — alifcoder · 2026-10-20
- Agents Ask for Approval 20 Times a Day—Are You Still Reviewing or Just Clicking Yes? — Glittering812 · 2026-10-20
- Skyrim AI mod lets bards compose songs about your adventures — and got flamed off Reddit — ReignMocap · 2026-10-20
- LangChain's Harrison Chase: how vertical agents differentiate via context, harness and model — hwchase17 · 2026-10-20