Llama 3.3 70B — MI300X vs MI355X Performance per Dollar
Cost per million tokens of MI300X (AMD CDNA 3) versus MI355X (AMD CDNA 4) 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 53 tok/s/user and MI300X lands at $0.26 per million tokens against MI355X's $0.18 — MI355X pulls ahead by 43%.
MI300X: $0.48 per million tokens. MI355X: $0.25. Both at 73 tok/s/user on Llama 3.3 70B, with MI355X 91% cheaper.
Toward the upper edge of the 34–112 tok/s/user interactivity band — at 93 tok/s/user — MI300X runs $0.83 per million tokens on Llama 3.3 70B while MI355X runs $0.46. MI355X is the cheaper choice by 79%. (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): MI300X $1.12/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 | MI300X:$0.257MI355X:$0.179 | MI300X:$0.479MI355X:$0.251 | MI300X:$0.832MI355X:$0.464 |
| Concurrency | MI300X:~50MI355X:~47 | MI300X:~32MI355X:~51 | MI300X:~17MI355X:~19 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.