Qwen 3.5 397B-A17B — H100 vs MI355X Performance per Dollar
Cost per million tokens of H100 (NVIDIA Hopper) versus MI355X (AMD CDNA 4) on Qwen 3.5 397B-A17B. 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.
Near the low end of the 45–110 tok/s/user interactivity band — at 61 tok/s/user — H100 runs $1.43 per million tokens on Qwen 3.5 397B-A17B while MI355X runs $0.18. MI355X is the cheaper choice by 675%.
On Qwen 3.5 397B-A17B at 78 tok/s/user, the per-million math comes out to $1.78 for H100 and $0.24 for MI355X; MI355X delivers 629% more output per dollar.
At 94 tok/s/user on Qwen 3.5 397B-A17B, H100 costs $2.38 per million tokens; MI355X costs $0.31. MI355X is 673% more cost-efficient at this operating point. (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): H100 $1.30/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 | H100:$1.429MI355X:$0.184 | H100:$1.779MI355X:$0.244 | H100:$2.381MI355X:$0.308 |
| Concurrency | H100:~17MI355X:~73 | H100:~10MI355X:~44 | H100:~7MI355X:~29 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.