gpt-oss 120B — MI300X vs MI355X
Head-to-head AI inference benchmark comparison of MI300X (AMD CDNA 3) and MI355X (AMD CDNA 4) on gpt-oss 120B. 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.
Setting 109 tok/s/user as the target on gpt-oss 120B, MI300X produces 1673 tok/s/GPU ($0.19 per million tokens) and MI355X produces 9193 ($0.04). MI355X is 314% cheaper per token; MI355X delivers 450% more tok/s/GPU.
At 156 tok/s/user interactivity on gpt-oss 120B, MI300X delivers 822 tok/s/GPU at $0.36 per million tokens; MI355X delivers 4578 tok/s/GPU at $0.09. MI355X is 304% cheaper per token; MI355X delivers 457% more tok/s/GPU at this point.
MI300X posts 316 tok/s/GPU for $0.96 per million tokens at 202 tok/s/user on gpt-oss 120B; MI355X posts 2110 tok/s/GPU for $0.20. MI355X is 390% cheaper per token; MI355X delivers 568% more tok/s/GPU. (Numbers reflect the default 1k/1k · fp4 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/gpu) | MI300X:1672.5MI355X:9193.0 | MI300X:822.3MI355X:4577.6 | MI300X:316.0MI355X:2110.0 |
| Cost ($/M tok) | MI300X:$0.186MI355X:$0.045 | MI300X:$0.364MI355X:$0.090 | MI300X:$0.958MI355X:$0.195 |
| tok/s/MW | MI300X:934375MI355X:3469061 | MI300X:459381MI355X:1727403 | MI300X:176517MI355X:796242 |
| Concurrency | MI300X:~16MI355X:~43 | MI300X:~5MI355X:~17 | MI300X:~7MI355X:~22 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.