DeepSeek R1 · Chip comparison
DeepSeek R1 — B300 vs H100
Head-to-head AI inference benchmark comparison of B300 (NVIDIA Blackwell) and H100 (NVIDIA Hopper) 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.
Near the low end of the 20–122 tok/s/user interactivity band, at 45 tok/s/user on DeepSeek R1: B300 runs 6,086 tok/s/chip at $0.10 per million tokens, H100 runs 739 at $0.44. B300 is 77% cheaper per token; B300 delivers 8.2× the tok/s/chip.
Setting 71 tok/s/user as the target on DeepSeek R1, B300 produces 4,027 tok/s/chip ($0.16 per million tokens) and H100 produces 281 ($1.16). B300 is 87% cheaper per token; B300 delivers 14× the tok/s/chip.
At 97 tok/s/user interactivity on DeepSeek R1, B300 delivers 1,914 tok/s/chip at $0.33 per million tokens; H100 delivers 154 tok/s/chip at $2.11. B300 is 84% cheaper per token; B300 delivers 12× the tok/s/chip at this point. (Numbers reflect the default 8k/1k · fp8 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
Interpolated from real benchmark data. Edit target interactivity values below to compare at different operating points.
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | B3006,086H100739 | B3004,027H100281 | B3001,914H100154 |
| Cost ($/M tok) | B300$0.10H100$0.44 | B300$0.16H100$1.16 | B300$0.33H100$2.11 |
| tok/s/MW | B3003,203,218H100539,778 | B3002,119,595H100204,852 | B3001,007,140H100112,644 |
| Concurrency | B300~358H100~132 | B300~103H100~41 | B300~51H100~17 |