DeepSeek R1 — Jalapeño vs MI325X
Head-to-head AI inference benchmark comparison of Jalapeño (OpenAI Jalapeño) and MI325X (AMD CDNA 3) on DeepSeek R1. Latency, throughput, and cost across LLM workloads. Use the chart controls below to switch sequences, precisions, and metrics — same interactions as the main inference chart.
MI325X: 966 tok/s/chip, $0.32 per million tokens at 35 tok/s/user on DeepSeek R1. Jalapeño is unmeasured here.
At 57 tok/s/user on DeepSeek R1, MI325X delivers 747 tok/s/chip at $0.41 per million tokens; Jalapeño hasn't been benchmarked at this target.
MI325X hits 607 tok/s/chip for $0.50 per million tokens at 78 tok/s/user on DeepSeek R1. No Jalapeño data at this operating point. (Numbers reflect the default 8k/1k · fp8 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | Jalapeño:—MI325X:966.0 | Jalapeño:—MI325X:747.0 | Jalapeño:—MI325X:606.7 |
| Cost ($/M tok) | Jalapeño:—MI325X:$0.316 | Jalapeño:—MI325X:$0.409 | Jalapeño:—MI325X:$0.504 |
| tok/s/MW | Jalapeño:—MI325X:571610 | Jalapeño:—MI325X:442003 | Jalapeño:—MI325X:358995 |
| Concurrency | Jalapeño:—MI325X:~24 | Jalapeño:—MI325X:~11 | Jalapeño:—MI325X:~7 |