vLLM P/D prototype hits 91.24% prefix cache hit rate on 12× GB300 with zero request errors
vllm_project · x · 2026-10-15
vLLM shared pre-release P/D disaggregation results for Step 5 Preview: DEP4 prefill connected to DEP8 decode on 12× GB300 using FP8, NIXL transfer and a 384 GiB MooncakeStore cache. The run recorded 514,784 prefix-hit tokens (91.24% hit rate), and two 1,000-request OCR runs scored 90.4% and 89.4% with zero request errors. The P/D PR remains open; final-build validation is separate.
More from Infra
- How Microsoft optimized Windows for Nvidia's Arm-based RTX Spark AI laptops — tomwarren · 2026-10-15
- How Microsoft optimized Windows for Nvidia's Arm-based RTX Spark laptops — The Verge AI · 2026-10-15
- Global smartphone shipments plunge 22% in worst quarterly decline on record as AI demand fuels memory costs — Polymarket · 2026-10-15
- SK Hynix drops 11% again, now below 3x FY2027 earnings as AI memory selloff deepens — JOBhakdi · 2026-10-15
- ASML Sells Zero Lithography Machines into Europe as US Hosts 75%+ of AI Compute — NinaDSchick · 2026-10-15
- Nvidia quietly reclassified SPCX as 'Hyperscale', signaling deeper neocloud reliance — RihardJarc · 2026-10-15