GLM 5/5.1 — B300 vs MI355X Performance per Dollar
Cost per million tokens of B300 (NVIDIA Blackwell) versus MI355X (AMD CDNA 4) 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.
Push GLM 5/5.1 to 38 tok/s/user and B300 lands at $0.74 per million tokens against MI355X's $0.49 — MI355X pulls ahead by 51%.
At 60 tok/s/user on GLM 5/5.1, B300 and MI355X land within ~1% on cost per million tokens ($1.10 vs $1.09) — call it a tie at this operating point.
Toward the upper edge of the 17–102 tok/s/user interactivity band — at 81 tok/s/user — B300 runs $1.54 per million tokens on GLM 5/5.1 while MI355X runs $1.73. B300 is the cheaper choice by 12%. (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): B300 $2.34/GPU/hr · MI355X $1.48/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 | B300:$0.739MI355X:$0.491 | B300:$1.102MI355X:$1.092 | B300:$1.543MI355X:$1.729 |
| Concurrency | B300:~93MI355X:~121 | B300:~40MI355X:~13 | B300:~21MI355X:~8 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.