Head-to-head AI inference benchmark comparison of GB300 NVL72 (NVIDIA Blackwell) and MI355X (AMD CDNA 4) on Qwen 3.8 Flash Next 176B-A6B. 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.
AgentX replays real coding-agent sessions rather than fixed-length prompts, so context grows turn over turn and most of each request is served from cache instead of being recomputed. That turns the comparison into a systems question: KV transfer between nodes, prefix-aware routing, and cache capacity all move the curve alongside raw chip throughput. Fixed-sequence workloads stay the clean baseline for kernel and silicon performance, so the two scenarios answer different questions about the same hardware. Learn more about AgentX β
Near the low end of the 67β145 tok/s/user interactivity band, at 86 tok/s/user on Qwen 3.8 Flash Next 176B-A6B: GB300 NVL72 runs 68791 tok/s/chip at $0.01/M tokens, MI355X runs 58839 at $0.01/M. MI355X is 32% cheaper per token; GB300 NVL72 delivers 17% more tok/s/chip.
Setting 106 tok/s/user as the target on Qwen 3.8 Flash Next 176B-A6B, GB300 NVL72 produces 60866 tok/s/chip ($0.01 per million tokens) and MI355X produces 36602 ($0.01). GB300 NVL72 is 8% cheaper per token; GB300 NVL72 delivers 66% more tok/s/chip.
At 125 tok/s/user interactivity on Qwen 3.8 Flash Next 176B-A6B, GB300 NVL72 delivers 54621 tok/s/chip at $0.01 per million tokens; MI355X delivers 19807 tok/s/chip at $0.02. GB300 NVL72 is 79% cheaper per token; GB300 NVL72 delivers 176% more tok/s/chip at this point. (Numbers reflect the default agentic-traces Β· fp4 selection for this URL β table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | GB300 NVL72:68790.6MI355X:58838.9 | GB300 NVL72:60865.6MI355X:36602.3 | GB300 NVL72:54621.5MI355X:19807.3 |
| Cost ($/M tok) | GB300 NVL72:$0.009MI355X:$0.007 | GB300 NVL72:$0.011MI355X:$0.011 | GB300 NVL72:$0.012MI355X:$0.021 |
| tok/s/MW | GB300 NVL72:32448398MI355X:28152565 | GB300 NVL72:28710169MI355X:17513060 | GB300 NVL72:25764843MI355X:9477154 |
| Concurrency | GB300 NVL72:~14MI355X:~13 | GB300 NVL72:~12MI355X:~8 | GB300 NVL72:~11MI355X:~5 |
Agentic inference metrics from the AgentX scenario and fixed-sequence inference metrics across models, hardware configurations, and serving parameters.
