GLM 5/5.1 — B300 vs MI325X Performance per Dollar
Cost per million tokens of B300 (NVIDIA Blackwell) versus MI325X (AMD CDNA 3) 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.
On GLM 5/5.1 at 21 tok/s/user, the per-million math comes out to $0.48 for B300 and $1.79 for MI325X; B300 delivers 274% more output per dollar.
At 26 tok/s/user on GLM 5/5.1, B300 costs $0.54 per million tokens; MI325X costs $3.04. B300 is 462% more cost-efficient at this operating point.
B300 edges MI325X at 30 tok/s/user on GLM 5/5.1 — $0.59 per million tokens versus $4.91, a 729% cost-per-token gap. (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.)
Chip pricing (owning hyperscaler): B300 $2.26/chip/hr · MI325X $1.10/chip/hr. Source: SemiAnalysis Market July 2026 Pricing Surveys & AI Cloud TCO Model.
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Dollar per Million Tokens | B300:$0.479MI325X:$1.789 | B300:$0.540MI325X:$3.037 | B300:$0.592MI325X:$4.907 |
| Concurrency | B300:~256MI325X:~34 | B300:~194MI325X:~16 | B300:~149MI325X:~9 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.