Llama 3.3 70B — B200 vs MI325X Performance per Dollar
Cost per million tokens of B200 (NVIDIA Blackwell) versus MI325X (AMD CDNA 3) on Llama 3.3 70B. 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 Llama 3.3 70B to 56 tok/s/user and B200 lands at $0.09 per million tokens against MI325X's $0.26 — B200 pulls ahead by 201%.
B200: $0.13 per million tokens. MI325X: $0.45. Both at 79 tok/s/user on Llama 3.3 70B, with B200 257% cheaper.
Toward the upper edge of the 34–124 tok/s/user interactivity band — at 102 tok/s/user — B200 runs $0.18 per million tokens on Llama 3.3 70B while MI325X runs $0.87. B200 is the cheaper choice by 384%. (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): B200 $1.95/GPU/hr · MI325X $1.28/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 | B200:$0.087MI325X:$0.263 | B200:$0.127MI325X:$0.454 | B200:$0.179MI325X:$0.866 |
| Concurrency | B200:~128MI325X:~64 | B200:~124MI325X:~44 | B200:~63MI325X:~16 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.