Head-to-head AI inference benchmark comparison of GB200 NVL72 (NVIDIA Blackwell) and MI355X (AMD CDNA 4) 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.
Setting 93 tok/s/user as the target on DeepSeek R1, GB200 NVL72 produces 11532 tok/s/chip ($0.04 per million tokens) and MI355X produces 7706 ($0.05). GB200 NVL72 is 21% cheaper per token; GB200 NVL72 delivers 50% more tok/s/chip.
At 162 tok/s/user interactivity on DeepSeek R1, GB200 NVL72 delivers 4192 tok/s/chip at $0.12 per million tokens; MI355X delivers 2021 tok/s/chip at $0.21. GB200 NVL72 is 67% cheaper per token; GB200 NVL72 delivers 107% more tok/s/chip at this point.
GB200 NVL72 posts 1012 tok/s/chip for $0.51 per million tokens at 231 tok/s/user on DeepSeek R1; MI355X posts 581 tok/s/chip for $0.72. GB200 NVL72 is 40% cheaper per token; GB200 NVL72 delivers 74% more tok/s/chip. (Numbers reflect the default 8k/1k · fp4 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | GB200 NVL72:11531.8MI355X:7706.4 | GB200 NVL72:4192.5MI355X:2020.7 | GB200 NVL72:1012.2MI355X:581.3 |
| Cost ($/M tok) | GB200 NVL72:$0.045MI355X:$0.054 | GB200 NVL72:$0.123MI355X:$0.206 | GB200 NVL72:$0.510MI355X:$0.717 |
| tok/s/MW | GB200 NVL72:6166725MI355X:3687294 | GB200 NVL72:2241975MI355X:966838 | GB200 NVL72:541259MI355X:278113 |
| Concurrency | GB200 NVL72:~913MI355X:~215 | GB200 NVL72:~178MI355X:~61 | GB200 NVL72:~19MI355X:~8 |
Agentic inference metrics from the AgentX scenario and fixed-sequence inference metrics across models, hardware configurations, and serving parameters.
TCO $/chip/hr: VR200: 3.61H100: 1.17H200: 1.22B200: 1.73B300: 2.26GB200: 1.86GB300: 2.31MI300X: 0.95MI325X: 1.1MI355X: 1.5RTX6000PRO: 0.68Jalapeño (Teacup): 1.47
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.
Matching measurements exist, but their chip series are hidden.
Shift+Scroll to zoom • Drag to pan • Double-click to reset • Click a point to pin tooltip
1 The ATOM engine is promising, however it has yet to serve production tokens. It is still in its infant stage.
DeepSeek R1 0528 671B 8K / 1K Total Tokens per $1 TCO vs. Interactivity