DeepSeek R1 — H200 vs MI300X
Head-to-head AI inference benchmark comparison of H200 (NVIDIA Hopper) and MI300X (AMD CDNA 3) 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.
H200 posts 2362 tok/s/chip for $0.14 per million tokens at 23 tok/s/user on DeepSeek R1; MI300X posts 725 tok/s/chip for $0.36. H200 is 154% cheaper per token; H200 delivers 226% more tok/s/chip.
Throughput at 32 tok/s/user on DeepSeek R1: H200 hits 1844 tok/s/chip, MI300X hits 601. Per-million costs land at $0.18 and $0.44 respectively. H200 is 139% cheaper per token; H200 delivers 207% more tok/s/chip.
H200 / MI300X on DeepSeek R1 at 42 tok/s/user: 1824 / 406 tok/s/chip, $0.19 / $0.65 per million tokens. H200 is 250% cheaper per token; H200 delivers 349% more tok/s/chip. (Numbers reflect the default 8k/1k · fp8 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | H200:2361.8MI300X:725.2 | H200:1844.1MI300X:601.1 | H200:1823.8MI300X:406.3 |
| Cost ($/M tok) | H200:$0.143MI300X:$0.364 | H200:$0.184MI300X:$0.439 | H200:$0.186MI300X:$0.650 |
| tok/s/MW | H200:1723946MI300X:521723 | H200:1346056MI300X:432443 | H200:1331257MI300X:292288 |
| Concurrency | H200:~409MI300X:~30 | H200:~128MI300X:~18 | H200:~128MI300X:~10 |
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