Snowflake's Semi-Persistence Speeds Up Multi-Model GPU Swapping Up to 19.9x
Snowflake AI Research introduced Semi-Persistence, a CRIU-based approach for serving multiple models on shared GPUs, achieving up to 19.9x faster model swapping while avoiding wasted capacity.
2026-09-01 ~ 2026-09-01 · 2 related posts
- Snowflake's Semi-Persistence Cuts Multi-Model GPU Sleep/Wake Cycles by Up to 19.9x — StasBekman · 2026-09-01
- Snowflake's Semi-Persistence uses CRIU to serve many models on shared GPUs — bookwormengr · 2026-09-01