DeepSeek R1 — MI300X vs MI355X
Head-to-head AI inference benchmark comparison of MI300X (AMD CDNA 3) and MI355X (AMD CDNA 4) 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.
Near the low end of the 25–68 tok/s/user interactivity band, at 36 tok/s/user on DeepSeek R1: MI300X runs 268 tok/s/GPU at $1.16/M tokens, MI355X runs 1836 at $0.23/M. MI355X is 413% cheaper per token; MI355X delivers 586% more tok/s/GPU.
Setting 47 tok/s/user as the target on DeepSeek R1, MI300X produces 186 tok/s/GPU ($1.66 per million tokens) and MI355X produces 1398 ($0.29). MI355X is 466% cheaper per token; MI355X delivers 650% more tok/s/GPU.
At 58 tok/s/user interactivity on DeepSeek R1, MI300X delivers 113 tok/s/GPU at $2.74 per million tokens; MI355X delivers 787 tok/s/GPU at $0.54. MI355X is 407% cheaper per token; MI355X delivers 594% more tok/s/GPU at this point. (Numbers reflect the default 1k/1k · fp8 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/gpu) | MI300X:267.6MI355X:1836.2 | MI300X:186.4MI355X:1398.0 | MI300X:113.3MI355X:786.9 |
| Cost ($/M tok) | MI300X:$1.163MI355X:$0.227 | MI300X:$1.663MI355X:$0.294 | MI300X:$2.737MI355X:$0.540 |
| tok/s/MW | MI300X:149525MI355X:692902 | MI300X:104145MI355X:527564 | MI300X:63313MI355X:296951 |
| Concurrency | MI300X:~31MI355X:~887 | MI300X:~17MI355X:~425 | MI300X:~8MI355X:~203 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.