Head-to-head AI inference benchmark comparison of H200 (NVIDIA Hopper) and MI325X (AMD CDNA 3) on GLM 5.3 744B. 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.
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 →
Near the low end of the 1–159 tok/s/user interactivity band, at 40 tok/s/user on GLM 5.3 744B: H200 runs 2769 tok/s/chip at $0.12/M tokens, MI325X runs 1218 at $0.25/M. H200 is 105% cheaper per token; H200 delivers 127% more tok/s/chip.
Setting 80 tok/s/user as the target on GLM 5.3 744B, H200 produces 2769 tok/s/chip ($0.12 per million tokens) and MI325X produces 1206 ($0.25). H200 is 107% cheaper per token; H200 delivers 130% more tok/s/chip.
At 120 tok/s/user interactivity on GLM 5.3 744B, H200 delivers 2769 tok/s/chip at $0.12 per million tokens; MI325X delivers 1206 tok/s/chip at $0.25. H200 is 107% cheaper per token; H200 delivers 130% more tok/s/chip at this point. (Numbers reflect the default agentic-traces · fp8 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | H200:2768.5MI325X:1217.9 | H200:2768.5MI325X:1205.8 | H200:2768.5MI325X:1205.8 |
| Cost ($/M tok) | H200:$0.122MI325X:$0.251 | H200:$0.122MI325X:$0.253 | H200:$0.122MI325X:$0.253 |
| tok/s/MW | H200:2020825MI325X:720648 | H200:2020825MI325X:713473 | H200:2020825MI325X:713473 |
| Concurrency | H200:~12MI325X:~2 | H200:~12MI325X:~1 | H200:~12MI325X:~1 |
Agentic inference metrics from the AgentX scenario and fixed-sequence inference metrics across models, hardware configurations, and serving parameters.
No measurements to plot for this selection. Review the benchmark controls above or adjust quick filters.
