DeepSeek R1 — H200 vs Vera Rubin NVL72
Head-to-head AI inference benchmark comparison of H200 (NVIDIA Hopper) and Vera Rubin NVL72 (NVIDIA Vera Rubin) on DeepSeek R1. Latency, throughput, and cost across LLM workloads. Use the chart controls below to switch sequences, precisions, and metrics — same interactions as the main inference chart.
At 51 tok/s/user on DeepSeek R1, H200 delivers 1499 tok/s/chip at $0.23 per million tokens; Vera Rubin NVL72 hasn't been benchmarked at this target.
H200 hits 785 tok/s/chip for $0.43 per million tokens at 90 tok/s/user on DeepSeek R1. No Vera Rubin NVL72 data at this operating point.
H200: 523 tok/s/chip, $0.65 per million tokens at 129 tok/s/user on DeepSeek R1. Vera Rubin NVL72 is unmeasured here. (Numbers reflect the default 8k/1k · fp8 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | H200:1499.5Vera Rubin NVL72:— | H200:785.4Vera Rubin NVL72:— | H200:523.1Vera Rubin NVL72:— |
| Cost ($/M tok) | H200:$0.226Vera Rubin NVL72:— | H200:$0.431Vera Rubin NVL72:— | H200:$0.648Vera Rubin NVL72:— |
| tok/s/MW | H200:1094512Vera Rubin NVL72:— | H200:573273Vera Rubin NVL72:— | H200:381794Vera Rubin NVL72:— |
| Concurrency | H200:~54Vera Rubin NVL72:— | H200:~8Vera Rubin NVL72:— | H200:~4Vera Rubin NVL72:— |