method · first seen 5 Oct 2026
GPU memory sharing
Also: VRAM management
Letting inference and training workloads share and reclaim GPU memory at fine granularity on the same device.
Mentions per week
PapersCodeDiscussion
Earliest sources
- MOLT: A Fine-Grained GPU Memory Sharing System for LLM Serving with Opportunistic Fine-TuningPapers · 5 Oct 2026 · Jaehoon Yang, Yongbeom Kim, Hojoon Kim, Seung Yul Lee, Jae W. Lee
- xiaoxiao150/vram-switcherCode · 26 Oct 2026 · xiaoxiao150 · 71 stars
Latest sources
- xiaoxiao150/vram-switcherCode · 26 Oct 2026 · xiaoxiao150 · 71 stars
- MOLT: A Fine-Grained GPU Memory Sharing System for LLM Serving with Opportunistic Fine-TuningPapers · 5 Oct 2026 · Jaehoon Yang, Yongbeom Kim, Hojoon Kim, Seung Yul Lee, Jae W. Lee