Head-to-head AI inference benchmark comparison of H200 (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 →
At 61 tok/s/user interactivity on DeepSeek V4.1 Flash 552B, H200 delivers 25325 tok/s/chip at $0.01 per million tokens; MI325X delivers 15608 tok/s/chip at $0.02. H200 is 46% cheaper per token; H200 delivers 62% more tok/s/chip at this point.
H200 posts 15186 tok/s/chip for $0.02 per million tokens at 112 tok/s/user on DeepSeek V4.1 Flash 552B; MI325X posts 9279 tok/s/chip for $0.03. H200 is 48% cheaper per token; H200 delivers 64% more tok/s/chip.
Throughput at 164 tok/s/user on DeepSeek V4.1 Flash 552B: H200 hits 10457 tok/s/chip, MI325X hits 4124. Per-million costs land at $0.03 and $0.07 respectively. H200 is 129% cheaper per token; H200 delivers 154% more tok/s/chip. (Numbers reflect the default agentic-traces · fp4 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | H200:25324.6MI325X:15608.2 | H200:15186.3MI325X:9279.5 | H200:10457.2MI325X:4124.2 |
| Cost ($/M tok) | H200:$0.013MI325X:$0.020 | H200:$0.022MI325X:$0.033 | H200:$0.032MI325X:$0.074 |
| tok/s/MW | H200:18485105MI325X:9235634 | H200:11084915MI325X:5490810 | H200:7632978MI325X:2440359 |
| Concurrency | H200:~20MI325X:~6 | H200:~10MI325X:~2 | H200:~6MI325X:~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.
