GLM 5/5.1 — GB200 NVL72 vs MI325X
Head-to-head AI inference benchmark comparison of GB200 NVL72 (NVIDIA Blackwell) and MI325X (AMD CDNA 3) on GLM 5/5.1. 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.
GB200 NVL72: 6049 tok/s/GPU, $0.10 per million tokens at 68 tok/s/user on GLM 5/5.1. MI325X is unmeasured here.
At 98 tok/s/user on GLM 5/5.1, GB200 NVL72 delivers 1772 tok/s/GPU at $0.35 per million tokens; MI325X hasn't been benchmarked at this target.
GB200 NVL72 hits 963 tok/s/GPU for $0.63 per million tokens at 127 tok/s/user on GLM 5/5.1. No MI325X data at this operating point. (Numbers reflect the default 8k/1k · fp4 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/gpu) | GB200 NVL72:6049.3MI325X:— | GB200 NVL72:1772.3MI325X:— | GB200 NVL72:962.6MI325X:— |
| Cost ($/M tok) | GB200 NVL72:$0.101MI325X:— | GB200 NVL72:$0.346MI325X:— | GB200 NVL72:$0.635MI325X:— |
| tok/s/MW | GB200 NVL72:3234942MI325X:— | GB200 NVL72:947766MI325X:— | GB200 NVL72:514742MI325X:— |
| Concurrency | GB200 NVL72:~725MI325X:— | GB200 NVL72:~84MI325X:— | GB200 NVL72:~49MI325X:— |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.