Qwen 3.5 397B-A17B — MI300X vs MI325X Performance per Dollar
Cost per million tokens of MI300X (AMD CDNA 3) versus MI325X (AMD CDNA 3) 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.
On Qwen 3.5 397B-A17B at 55 tok/s/user, the per-million math comes out to $0.86 for MI300X and $0.76 for MI325X; MI325X delivers 12% more output per dollar.
At 68 tok/s/user on Qwen 3.5 397B-A17B, MI300X costs $1.17 per million tokens; MI325X costs $1.07. MI325X is 9% more cost-efficient at this operating point.
MI325X edges MI300X at 82 tok/s/user on Qwen 3.5 397B-A17B — $1.65 per million tokens versus $1.74, a 6% cost-per-token gap. (Numbers reflect the default 1k/1k · bf16 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
Chip pricing (owning hyperscaler): MI300X $0.95/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 | MI300X:$0.857MI325X:$0.765 | MI300X:$1.169MI325X:$1.074 | MI300X:$1.744MI325X:$1.653 |
| Concurrency | MI300X:~23MI325X:~32 | MI300X:~13MI325X:~17 | MI300X:~8MI325X:~10 |
Cost per Million Total Tokens (Owning - Hyperscaler) vs. Interactivity
Qwen3.5 397B • FP4 • 8K / 1K • Source: SemiAnalysis InferenceX™
Qwen3.5 397B • FP4 • 8K / 1K • Source: SemiAnalysis InferenceX™TCO $/chip/hr: VR200: 3.61H100: 1.17H200: 1.22B200: 1.73B300: 2.26GB200: 1.86GB300: 2.31MI300X: 0.95MI325X: 1.1MI355X: 1.5RTX6000PRO: 0.68Jalapeño (Teacup): 1.47
Source: SemiAnalysis Market July 2026 Pricing Surveys & AI Cloud TCO Model
Note: Disaggregated inference configurations (e.g., MoRI SGLang, Dynamo TRTLLM) calculate cost per million tokens per decode chip or per prefill chip, rather than per total chip count. This makes direct token cost comparison with aggregated configs not an apples-to-apples comparison.
Note: Disaggregated inference configurations (e.g., MoRI SGLang, Dynamo TRTLLM) calculate input throughput per decode chip or per prefill chip, rather than per total chip count. This makes direct input throughput comparison with aggregated configs not an apples-to-apples comparison.
Note: Disaggregated inference configurations (e.g., MoRI SGLang, Dynamo TRTLLM) calculate output throughput per decode chip or per prefill chip, rather than per total chip count. This makes direct output throughput comparison with aggregated configs not an apples-to-apples comparison.
Note: Disaggregated inference configurations (e.g., MoRI SGLang, Dynamo TRTLLM) calculate power per decode chip or per prefill chip, rather than per total chip count. This makes direct power comparison with aggregated configs not an apples-to-apples comparison.
Note: Disaggregated inference configurations (e.g., MoRI SGLang, Dynamo TRTLLM) calculate Joules per decode chip or per prefill chip, rather than per total chip count. This makes direct Joules per token comparison with aggregated configs not an apples-to-apples comparison.
No data available
Please change the model, sequence, precision, date range or chip selection.
Shift+Scroll to zoom • Drag to pan • Double-click to reset • Click a point to pin tooltip
Interactivity is the rate at which a single user receives generated tokens while the model streams its answer — how quickly new words appear on screen. Higher values feel snappier; operators trade it against batch throughput.
Estimated infrastructure cost of producing one million total tokens (input + output) at this operating point, priced with the all-in hourly ownership cost of a hyperscaler operator. Lower is cheaper.
Formula: $/Mtok = all-in cost per chip-hour ($) × 1,000,000 ÷ (3,600 × total tok/s/chip)