Head-to-head AI inference benchmark comparison of H100 (NVIDIA Hopper) and MI325X (AMD CDNA 3) on DeepSeek V4.1 Flash 552B. 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 →
H100 posts 9216 tok/s/chip for $0.04 per million tokens at 107 tok/s/user on DeepSeek V4.1 Flash 552B; MI325X posts 5134 tok/s/chip for $0.06. H100 is 69% cheaper per token; H100 delivers 79% more tok/s/chip.
Throughput at 139 tok/s/user on DeepSeek V4.1 Flash 552B: H100 hits 7462 tok/s/chip, MI325X hits 4270. Per-million costs land at $0.04 and $0.07 respectively. H100 is 64% cheaper per token; H100 delivers 75% more tok/s/chip.
H100 / MI325X on DeepSeek V4.1 Flash 552B at 172 tok/s/user: 6012 / 2143 tok/s/chip, $0.05 / $0.14 per million tokens. H100 is 164% cheaper per token; H100 delivers 181% more tok/s/chip. (Numbers reflect the default agentic-traces · fp4 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | H100:9215.7MI325X:5134.4 | H100:7461.7MI325X:4269.9 | H100:6012.0MI325X:2142.8 |
| Cost ($/M tok) | H100:$0.035MI325X:$0.060 | H100:$0.044MI325X:$0.072 | H100:$0.054MI325X:$0.143 |
| tok/s/MW | H100:6726805MI325X:3038100 | H100:5446488MI325X:2526560 | H100:4388357MI325X:1267920 |
| Concurrency | H100:~12MI325X:~7 | H100:~9MI325X:~6 | H100:~7MI325X:~2 |
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.
