First DeepSeek V4.1 Flash results on Vera Rubin NVL72: up to 93.3x per-chip throughput vs GB300
vllm_project · x · 2026-11-17
The vLLM project published the first DeepSeek V4.1 Flash inference results on NVIDIA Vera Rubin NVL72, powered by Dynamo + vLLM on SemiAnalysis's AgentX benchmark.
Compared with GB300 NVL72 running the same Dynamo + vLLM stack, InferenceX's perf ruler shows per-chip total token throughput of:
- 63.4x at 400 tok/s/user
- 93.3x at 600 tok/s/user
The chart shows the gap widening sharply with interactivity requirements: GB300 throughput collapses past 400 tok/s/user while Vera Rubin sustains usable throughput beyond 1800 tok/s/user.
More from Infra
- Why tensor parallelism breaks RNG reproducibility in model init — ezyang · 2026-11-17
- Bittensor's Engy tops Venice's AI inference volume despite 28x smaller valuation — markjeffrey · 2026-11-17
- Fractile raises $720M for AI inference chips, betting inference cost breaks AI economics first — YvesMulkers · 2026-11-17
- DeepSeek V4.1 Flash on Vera Rubin NVL72: 63-93x per-chip throughput vs GB300 in vLLM benchmarks — vllm_project · 2026-11-17
- Put an open-source AI on your phone in 10 minutes: free, offline, no account — alex_verem · 2026-11-17
- Cloudflare open-sources Kitesurf browser engine for agents, ships DO and spend-limit updates — ritakozlov · 2026-11-17