GLM 5/5.1 · GPU comparison

GLM 5/5.1 — B200 vs MI355X

Head-to-head AI inference benchmark comparison of B200 (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.

Throughput at 34 tok/s/user on GLM 5/5.1: B200 hits 641 tok/s/GPU, MI355X hits 913. Per-million costs land at $0.86 and $0.45 respectively. MI355X is 89% cheaper per token; MI355X delivers 42% more tok/s/GPU.

B200 / MI355X on GLM 5/5.1 at 57 tok/s/user: 369 / 397 tok/s/GPU, $1.47 / $1.04 per million tokens. MI355X is 42% cheaper per token; MI355X delivers 7% more tok/s/GPU.

Toward the upper edge of the 12–102 tok/s/user interactivity band, at 80 tok/s/user on GLM 5/5.1: B200 runs 221 tok/s/GPU at $2.48/M tokens, MI355X runs 251 at $1.65/M. MI355X is 50% cheaper per token; MI355X delivers 14% more tok/s/GPU. (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.)

View performance-per-dollar view →

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/gpu)
B200:640.8MI355X:913.2
B200:369.0MI355X:396.5
B200:220.9MI355X:250.9
Cost ($/M tok)
B200:$0.856MI355X:$0.453
B200:$1.466MI355X:$1.036
B200:$2.481MI355X:$1.650
tok/s/MW
B200:295317MI355X:344591
B200:170064MI355X:149636
B200:101810MI355X:94693
Concurrency
B200:~198MI355X:~128
B200:~25MI355X:~14
B200:~11MI355X:~8

Inference Performance

Inference performance metrics across different models, hardware configurations, and serving parameters.