SkyPilot Blog company blog

Grace Blackwell needs a new scheduler: 27% more usable GPUs from the same racks

NVL72 puts all 72 GPUs in a rack on one NVLink fabric, but only rack-local jobs get the speedup. How SkyPilot keeps multi-node training and inference on NVLink with topology-aware placement, in-place recovery from GPU failures, and automatic defragmentation.
Points
147
HN comments
22
Submissions
1
Cited by
0 commenters

Infrastructure & DevOpsAI & Machine Learning2026

On Hacker News

SubmittedTitle on HNPointsComments
Oct 29, 2026Grace Blackwell needs a new scheduler: 27% more usable GPUs from the same racks by wondersky14722

Related posts from the same era