Researcher begs for spare compute as university cluster stalls, ~18 GPU-hours left to finish
vxnuaj · x · 2026-11-19
A researcher shared a screenshot of their university CARC shared cluster stuck for 3 hours: the S7 confirmation test frozen at 127/500 (all slices waiting on priority) and L1 replication 0/3000 queued, with their CARC priority very low after three days of light use. About 8 GPU-hours for S7 and 10 for L1-R remain (18 total). Two options on the table: top up $25 on Prime Intellect for roughly 12-15 spot-A100 hours via a parallel-worker trick to finish today, or stay on CARC and wait possibly until Nov 22-25. A vivid look at the compute struggles of academic researchers on shared queues.
More from Infra
- One Amazon EBS gp3 command scales a 10M-vector Weaviate index 26×, from 9.8 to 260 QPS — philipvollet · 2026-11-19
- Local MiniMax H3 on 6GB VRAM: 50 minutes tuned vs 6.5 hours on defaults — No_Cow26 · 2026-11-19
- OVH launches Olala mini 7A1B: open-source European 7B LLM that runs on CPU — jedisct1 · 2026-11-19
- Compress Wikipedia to ~27GB locally: neural compression via tufazip — Terminator857 · 2026-11-19
- One GPU transcribes 8.7 hours of mock GP calls in 18 minutes, 27.8x realtime — MaziyarPanahi · 2026-11-19
- Ramp AI Index: token prices down 21.2% in six weeks, Anthropic holds 58.7% of spend — RebeccaBellan · 2026-11-19