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 12484 tok/s/chip for $0.03 per million tokens at 61 tok/s/user on DeepSeek V4.1 Flash 552B; MI325X posts 15608 tok/s/chip for $0.02. MI325X is 33% cheaper per token; MI325X delivers 25% more tok/s/chip.
Throughput at 112 tok/s/user on DeepSeek V4.1 Flash 552B: H100 hits 9028 tok/s/chip, MI325X hits 9279. Per-million costs land at $0.04 and $0.03 respectively. MI325X is 9% cheaper per token; MI325X delivers 3% more tok/s/chip.
H100 / MI325X on DeepSeek V4.1 Flash 552B at 164 tok/s/user: 6522 / 4124 tok/s/chip, $0.05 / $0.07 per million tokens. H100 is 49% cheaper per token; H100 delivers 58% more tok/s/chip. (Numbers reflect the default agentic-traces · fp4 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | H100:12484.1MI325X:15608.2 | H100:9027.9MI325X:9279.5 | H100:6521.6MI325X:4124.2 |
| Cost ($/M tok) | H100:$0.026MI325X:$0.020 | H100:$0.036MI325X:$0.033 | H100:$0.050MI325X:$0.074 |
| tok/s/MW | H100:9112460MI325X:9235634 | H100:6589682MI325X:5490810 | H100:4760271MI325X:2440359 |
| Concurrency | H100:~19MI325X:~6 | H100:~11MI325X:~2 | H100:~7MI325X:~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.
