NVIDIA open-sources Snapshot, which saves an initialized GPU inference replica as a checkpoint and restores it, and design partner Crusoe's own test cut time-to-ready from about 8 minutes to 31 seconds.

Concept illustration: at the lower left stands a dark grey server cabinet with horizontal vents, an orange diamond mark and the lettering CRUSOE, and to its right is a large dark-framed board whose top tab reads GLM-5.3. Inside the board, two rows start beside a thin vertical line on the left: the upper row shows, from left to right and linked by grey bars, a terminal with a blank screen, a stack of cards lettered MODEL, a long bar lettered WARM-UP under a row of five gears, and a terminal showing a speech bubble; the much shorter lower row shows a blank-screen terminal partly covered by a brick-red card lettered NVIDIA SNAPSHOT, which also carries a small green emblem and a picture of a terminal with a speech bubble, and a grey bar links this row to a terminal showing a speech bubble. Printed on the image: ~8 min→31 s GLM-5.3 replica time-to-ready without vs with Snapshot, in NVIDIA design partner Crusoe's own test.

Inference services may be able to keep fewer warm replicas standing by for peaks, but the figures come from NVIDIA's design partners' own tests and no third party has repeated them. Compared with caching only model weights, Snapshot restores a replica that has already finished initializing, and NVIDIA says the model doesn't have to be reloaded or warmed up. Crusoe is a cloud provider that rents out GPUs, and it tested GLM-5.3. Another partner, Saturn Cloud, tested scaling the open-source inference engine vLLM from zero, and startup time fell from 2 to 8.5 minutes to 6 to 34 seconds (NVIDIA, 2026-11-10). How we checked →

Source and checks

This item ran in the brief of Nov 11, 2026, alongside 15 other items. How we checked it, the quote checks and what was still missing that day are on that issue's How we checked page.

Cite this item

SecondSource (2026). "NVIDIA open-sources Snapshot, which saves an initialized GPU inference replica as a checkpoint and restores it, and design partner Crusoe's own test cut time-to-ready from about 8 minutes to 31 seconds.." https://secondsource.io/en/product/2026-11-11-1/. Accessed access date.