GLM 5/5.1 · Chip comparison

GLM 5/5.1 — GB300 NVL72 vs MI325X

Head-to-head AI inference benchmark comparison of GB300 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.

GB300 NVL72: 10,192 tok/s/chip, $0.063 per million tokens at 67 tok/s/user on GLM 5/5.1. MI325X is unmeasured here.

At 105 tok/s/user on GLM 5/5.1, GB300 NVL72 delivers 4,238 tok/s/chip at $0.15 per million tokens; MI325X hasn't been benchmarked at this target.

GB300 NVL72 hits 802 tok/s/chip for $0.80 per million tokens at 143 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.)

View performance per dollar →

Interpolated from real benchmark data. Edit target interactivity values below to compare at different operating points.

Metric
Interactivity (tok/s/user)
Interactivity (tok/s/user)
Interactivity (tok/s/user)
Throughput (tok/s/chip)
GB300 NVL7210,192MI325X—
GB300 NVL724,238MI325X—
GB300 NVL72802MI325X—
Cost ($/M tok)
GB300 NVL72$0.063MI325X—
GB300 NVL72$0.15MI325X—
GB300 NVL72$0.80MI325X—
tok/s/MW
GB300 NVL724,807,740MI325X—
GB300 NVL721,999,045MI325X—
GB300 NVL72378,127MI325X—
Concurrency
GB300 NVL72~749MI325X—
GB300 NVL72~263MI325X—
GB300 NVL72~22MI325X—