{"slug": "vllm-lets-compressed-tensors-block-fp8-weights-survive-an-engine-restart", "title": "vLLM lets compressed-tensors block-FP8 weights survive an engine restart", "summary": "VLLM merged CUDA IPC weight-cache restart support on November 15th, 2026 for serialized compressed-tensors models using 128-by-128 block FP8 linear weights, letting a restarted engine map processed GPU weights and scales held through CUDA inter-process communication instead of rebuilding that prepared storage. The merged path restores the metadata required by the validated CUTLASS block-FP8 configuration, and exact greedy output was checked on a B300.", "body_md": "# vLLM lets compressed-tensors block-FP8 weights survive an engine restart\n\n**The merged change lets a restarted engine map processed CUTLASS block-FP8 weights through CUDA IPC, with exact greedy output checked on a B300.**\n\n        By [RuntimeWire Staff](https://runtimewire.com/author/runtimewire-staff)\n        · Published \n\nvLLM merged CUDA IPC weight-cache restart support on November 15th, 2026 for serialized compressed-tensors models that use 128-by-128 block FP8 linear weights. The change lets a restarted engine map processed GPU weights and scales held through CUDA inter-process communication instead of rebuilding that prepared storage from scratch.\n\nThe merged path reuses processed weights and scales through CUDA IPC while restoring the metadata required by the validated CUTLASS block-FP8 configuration - explanatory diagram, not documentary evidence. RuntimeWire / AI-generated diagram.…", "url": "https://wpnews.pro/news/vllm-lets-compressed-tensors-block-fp8-weights-survive-an-engine-restart", "canonical_source": "https://runtimewire.com/article/vllm-compressed-tensors-block-fp8-cuda-ipc-cache-restarts", "published_at": "2026-11-16 03:37:04+00:00", "updated_at": "2026-11-16 03:51:17.847558+00:00", "lang": "en", "topics": ["ai-infrastructure", "large-language-models", "mlops", "ai-tools"], "entities": ["vLLM", "CUDA IPC", "CUTLASS", "compressed-tensors", "B300"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/vllm-lets-compressed-tensors-block-fp8-weights-survive-an-engine-restart", "markdown": "https://wpnews.pro/news/vllm-lets-compressed-tensors-block-fp8-weights-survive-an-engine-restart.md", "text": "https://wpnews.pro/news/vllm-lets-compressed-tensors-block-fp8-weights-survive-an-engine-restart.txt", "jsonld": "https://wpnews.pro/news/vllm-lets-compressed-tensors-block-fp8-weights-survive-an-engine-restart.jsonld"}}