Kimi K2.5/K2.6/K2.7-Code 1T — GB300 NVL72 vs MI355X Performance per Dollar
Cost per million tokens of GB300 NVL72 (NVIDIA Blackwell) versus MI355X (AMD CDNA 4) on Kimi K2.5/K2.6/K2.7-Code 1T. 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.
At 50 tok/s/user on Kimi K2.5/K2.6/K2.7-Code 1T, GB300 NVL72 costs $0.07 per million tokens; MI355X costs $0.37. GB300 NVL72 is 427% more cost-efficient at this operating point.
GB300 NVL72 edges MI355X at 76 tok/s/user on Kimi K2.5/K2.6/K2.7-Code 1T — $0.18 per million tokens versus $0.59, a 226% cost-per-token gap.
Push Kimi K2.5/K2.6/K2.7-Code 1T to 102 tok/s/user and GB300 NVL72 lands at $1.20 per million tokens against MI355X's $0.83 — MI355X pulls ahead by 44%. (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.)
GPU pricing (owning hyperscaler): GB300 NVL72 $2.65/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 | GB300 NVL72:$0.071MI355X:$0.375 | GB300 NVL72:$0.181MI355X:$0.589 | GB300 NVL72:$1.199MI355X:$0.832 |
| Concurrency | GB300 NVL72:~4301MI355X:~44 | GB300 NVL72:~1242MI355X:~19 | GB300 NVL72:~135MI355X:~10 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.