GLM 5/5.1 — GB300 NVL72 vs MI355X
Head-to-head AI inference benchmark comparison of GB300 NVL72 (NVIDIA Blackwell) and MI355X (AMD CDNA 4) 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.
Setting 41 tok/s/user as the target on GLM 5/5.1, GB300 NVL72 produces 3128 tok/s/GPU ($0.23 per million tokens) and MI355X produces 474 ($0.85). GB300 NVL72 is 265% cheaper per token; GB300 NVL72 delivers 560% more tok/s/GPU.
At 50 tok/s/user interactivity on GLM 5/5.1, GB300 NVL72 delivers 696 tok/s/GPU at $1.04 per million tokens; MI355X delivers 301 tok/s/GPU at $1.38. GB300 NVL72 is 33% cheaper per token; GB300 NVL72 delivers 131% more tok/s/GPU at this point.
GB300 NVL72 posts 434 tok/s/GPU for $1.68 per million tokens at 58 tok/s/user on GLM 5/5.1; MI355X posts 194 tok/s/GPU for $2.11. GB300 NVL72 is 26% cheaper per token; GB300 NVL72 delivers 124% more tok/s/GPU. (Numbers reflect the default 1k/1k · fp4 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/gpu) | GB300 NVL72:3128.5MI355X:473.8 | GB300 NVL72:696.0MI355X:300.8 | GB300 NVL72:433.8MI355X:193.7 |
| Cost ($/M tok) | GB300 NVL72:$0.234MI355X:$0.854 | GB300 NVL72:$1.035MI355X:$1.376 | GB300 NVL72:$1.683MI355X:$2.114 |
| tok/s/MW | GB300 NVL72:1489755MI355X:178809 | GB300 NVL72:331442MI355X:113501 | GB300 NVL72:206554MI355X:73097 |
| Concurrency | GB300 NVL72:~3055MI355X:~23 | GB300 NVL72:~512MI355X:~12 | GB300 NVL72:~343MI355X:~7 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.