GLM 5/5.1 — H200 vs MI325X Performance per Dollar
Cost per million tokens of H200 (NVIDIA Hopper) versus MI325X (AMD CDNA 3) on GLM 5/5.1. Owning-hyperscaler TCO normalized by output tokens — performance per dollar across LLM workloads. Pick the more cost-efficient SKU at every target interactivity level. Use the chart controls below to switch sequences, precisions, and metrics — same interactions as the main inference chart.
At 24 tok/s/user on GLM 5/5.1, H200 costs $0.99 per million tokens; MI325X costs $2.46. H200 is 149% more cost-efficient at this operating point.
H200 edges MI325X at 28 tok/s/user on GLM 5/5.1 — $0.99 per million tokens versus $4.82, a 387% cost-per-token gap.
Push GLM 5/5.1 to 31 tok/s/user and H200 lands at $0.99 per million tokens against MI325X's $6.55 — H200 pulls ahead by 562%. (Numbers reflect the default 1k/1k · fp8 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
GPU pricing (owning hyperscaler): H200 $1.41/GPU/hr · MI325X $1.28/GPU/hr. Source: SemiAnalysis Market August 2025 Pricing Surveys & AI Cloud TCO Model.
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Dollar per Million Tokens | H200:$0.990MI325X:$2.465 | H200:$0.990MI325X:$4.823 | H200:$0.990MI325X:$6.553 |
| Concurrency | H200:~64MI325X:~25 | H200:~64MI325X:~14 | H200:~64MI325X:~8 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.