Head-to-head AI inference benchmark comparison of GB300 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 β same interactions as the main inference chart.
Throughput at 95 tok/s/user on DeepSeek R1: GB300 NVL72 hits 12350 tok/s/chip, MI355X hits 7592. Per-million costs land at $0.05 and $0.05 respectively. GB300 NVL72 is 6% cheaper per token; GB300 NVL72 delivers 63% more tok/s/chip.
GB300 NVL72 / MI355X on DeepSeek R1 at 167 tok/s/user: 4744 / 1909 tok/s/chip, $0.14 / $0.22 per million tokens. GB300 NVL72 is 61% cheaper per token; GB300 NVL72 delivers 148% more tok/s/chip.
Toward the upper edge of the 23β312 tok/s/user interactivity band, at 240 tok/s/user on DeepSeek R1: GB300 NVL72 runs 961 tok/s/chip at $0.67/M tokens, MI355X runs 538 at $0.77/M. GB300 NVL72 is 16% cheaper per token; GB300 NVL72 delivers 79% more tok/s/chip. (Numbers reflect the default 8k/1k Β· fp4 selection for this URL β table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | GB300 NVL72:12350.5MI355X:7592.3 | GB300 NVL72:4743.7MI355X:1909.0 | GB300 NVL72:961.3MI355X:538.1 |
| Cost ($/M tok) | GB300 NVL72:$0.052MI355X:$0.055 | GB300 NVL72:$0.135MI355X:$0.218 | GB300 NVL72:$0.667MI355X:$0.774 |
| tok/s/MW | GB300 NVL72:5825707MI355X:3632663 | GB300 NVL72:2237582MI355X:913385 | GB300 NVL72:453449MI355X:257481 |
| Concurrency | GB300 NVL72:~690MI355X:~204 | GB300 NVL72:~181MI355X:~55 | GB300 NVL72:~23MI355X:~8 |
Agentic inference metrics from the AgentX scenario and fixed-sequence inference metrics across models, hardware configurations, and serving parameters.
TCO $/chip/hr:
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 β’ Drag to pan β’ Double-click to reset β’ 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