Head-to-head AI inference benchmark comparison of MI300X (AMD CDNA 3) and MI325X (AMD CDNA 3) on MiniMax M3 428B. Latency, throughput, and cost across LLM workloads. Use the chart controls below to switch sequences, precisions, and metrics β same interactions as the main inference chart.
AgentX replays real coding-agent sessions rather than fixed-length prompts, so context grows turn over turn and most of each request is served from cache instead of being recomputed. That turns the comparison into a systems question: KV transfer between nodes, prefix-aware routing, and cache capacity all move the curve alongside raw chip throughput. Fixed-sequence workloads stay the clean baseline for kernel and silicon performance, so the two scenarios answer different questions about the same hardware. Learn more about AgentX β
MI300X / MI325X on MiniMax M3 428B at 28 tok/s/user: 5025 / 940 tok/s/chip, $0.05 / $0.33 per million tokens. MI300X is 519% cheaper per token; MI300X delivers 435% more tok/s/chip.
Around the middle of the 28β30 tok/s/user interactivity band, at 29 tok/s/user on MiniMax M3 428B: MI300X runs 5025 tok/s/chip at $0.05/M tokens, MI325X runs 912 at $0.34/M. MI300X is 538% cheaper per token; MI300X delivers 451% more tok/s/chip.
Setting 30 tok/s/user as the target on MiniMax M3 428B, MI300X produces 5025 tok/s/chip ($0.05 per million tokens) and MI325X produces 882 ($0.35). MI300X is 560% cheaper per token; MI300X delivers 470% more tok/s/chip. (Numbers reflect the default agentic-traces Β· fp8 selection for this URL β table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | MI300X:5024.9MI325X:940.0 | MI300X:5024.9MI325X:911.6 | MI300X:5024.9MI325X:881.7 |
| Cost ($/M tok) | MI300X:$0.053MI325X:$0.325 | MI300X:$0.053MI325X:$0.335 | MI300X:$0.053MI325X:$0.347 |
| tok/s/MW | MI300X:3615012MI325X:556192 | MI300X:3615012MI325X:539380 | MI300X:3615012MI325X:521728 |
| Concurrency | MI300X:~10MI325X:~2 | MI300X:~10MI325X:~2 | MI300X:~10MI325X:~1 |
Agentic inference metrics from the AgentX scenario and fixed-sequence inference metrics across models, hardware configurations, and serving parameters.
