DeepSeek R1 — B200 vs GB200 NVL72
Head-to-head AI inference benchmark comparison of B200 (NVIDIA Blackwell) and GB200 NVL72 (NVIDIA Blackwell) 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.
Throughput at 76 tok/s/user on DeepSeek R1: B200 hits 2423 tok/s/chip, GB200 NVL72 hits 5749. Per-million costs land at $0.20 and $0.09 respectively. GB200 NVL72 is 121% cheaper per token; GB200 NVL72 delivers 137% more tok/s/chip.
B200 / GB200 NVL72 on DeepSeek R1 at 120 tok/s/user: 1111 / 3988 tok/s/chip, $0.43 / $0.13 per million tokens. GB200 NVL72 is 234% cheaper per token; GB200 NVL72 delivers 259% more tok/s/chip.
Toward the upper edge of the 32–208 tok/s/user interactivity band, at 164 tok/s/user on DeepSeek R1: B200 runs 751 tok/s/chip at $0.64/M tokens, GB200 NVL72 runs 818 at $0.63/M. GB200 NVL72 is 1% cheaper per token; GB200 NVL72 delivers 9% more tok/s/chip. (Numbers reflect the default 8k/1k · fp8 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | B200:2423.5GB200 NVL72:5749.2 | B200:1110.7GB200 NVL72:3987.5 | B200:751.2GB200 NVL72:817.8 |
| Cost ($/M tok) | B200:$0.198GB200 NVL72:$0.090 | B200:$0.433GB200 NVL72:$0.130 | B200:$0.640GB200 NVL72:$0.632 |
| tok/s/MW | B200:1417227GB200 NVL72:3074446 | B200:649556GB200 NVL72:2132361 | B200:439300GB200 NVL72:437312 |
| Concurrency | B200:~142GB200 NVL72:~618 | B200:~119GB200 NVL72:~333 | B200:~21GB200 NVL72:~27 |
Inference Performance
Agentic inference metrics from the AgentX scenario and fixed-sequence inference metrics across models, hardware configurations, and serving parameters.
Total Tokens per $1 USD (Owning - Hyperscaler) vs. Interactivity
DeepSeek R1 0528 671B • FP4 • 8K / 1K • Source: SemiAnalysis InferenceX™
TCO $/chip/hr: H100: 1.17H200: 1.22B200: 1.73B300: 2.26GB200: 1.86GB300: 2.31MI300X: 0.95MI325X: 1.1MI355X: 1.5RTX6000PRO: 0.68
Source: SemiAnalysis Market July 2026 Pricing Surveys & AI Cloud TCO Model
Note: Disaggregated inference configurations (e.g., MoRI SGLang, Dynamo TRTLLM) calculate cost per million tokens per decode chip or per prefill chip, rather than per total chip count. This makes direct token cost comparison with aggregated configs not an apples-to-apples comparison.
Note: Disaggregated inference configurations (e.g., MoRI SGLang, Dynamo TRTLLM) calculate input throughput per decode chip or per prefill chip, rather than per total chip count. This makes direct input throughput comparison with aggregated configs not an apples-to-apples comparison.
Note: Disaggregated inference configurations (e.g., MoRI SGLang, Dynamo TRTLLM) calculate output throughput per decode chip or per prefill chip, rather than per total chip count. This makes direct output throughput comparison with aggregated configs not an apples-to-apples comparison.
Note: Disaggregated inference configurations (e.g., MoRI SGLang, Dynamo TRTLLM) calculate power per decode chip or per prefill chip, rather than per total chip count. This makes direct power comparison with aggregated configs not an apples-to-apples comparison.
Note: Disaggregated inference configurations (e.g., MoRI SGLang, Dynamo TRTLLM) calculate Joules per decode chip or per prefill chip, rather than per total chip count. This makes direct Joules per token comparison with aggregated configs not an apples-to-apples comparison.
No data available
Please change the model, sequence, precision, date range or chip selection.
Shift+Scroll to zoom • Drag to pan • Double-click to reset • Click a point to pin tooltip