DeepSeek R1 — B300 vs MI325X
Head-to-head AI inference benchmark comparison of B300 (NVIDIA Blackwell) and MI325X (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.
B300 / MI325X on DeepSeek R1 at 25 tok/s/user: 7091 / 854 tok/s/chip, $0.09 / $0.36 per million tokens. B300 is 304% cheaper per token; B300 delivers 730% more tok/s/chip.
Around the middle of the 15–54 tok/s/user interactivity band, at 35 tok/s/user on DeepSeek R1: B300 runs 6590 tok/s/chip at $0.10/M tokens, MI325X runs 677 at $0.45/M. B300 is 373% cheaper per token; B300 delivers 873% more tok/s/chip.
Setting 44 tok/s/user as the target on DeepSeek R1, B300 produces 6132 tok/s/chip ($0.10 per million tokens) and MI325X produces 457 ($0.67). B300 is 553% cheaper per token; B300 delivers 1242% more tok/s/chip. (Numbers reflect the default 8k/1k · fp8 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | B300:7091.1MI325X:854.2 | B300:6589.6MI325X:677.5 | B300:6132.0MI325X:456.8 |
| Cost ($/M tok) | B300:$0.089MI325X:$0.358 | B300:$0.095MI325X:$0.451 | B300:$0.102MI325X:$0.669 |
| tok/s/MW | B300:3732157MI325X:505466 | B300:3468197MI325X:400884 | B300:3227361MI325X:270289 |
| Concurrency | B300:~1361MI325X:~33 | B300:~599MI325X:~19 | B300:~374MI325X:~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