Qwen 3.5 397B-A17B · Chip comparison
Qwen 3.5 397B-A17B — MI325X vs MI355X
Head-to-head AI inference benchmark comparison of MI325X (AMD CDNA 3) and MI355X (AMD CDNA 4) on Qwen 3.5 397B-A17B. 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.
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 →
Near the low end of the 32–58 tok/s/user interactivity band, at 38 tok/s/user on Qwen 3.5 397B-A17B: MI325X runs 9,668 tok/s/chip at $0.032 per million tokens, MI355X runs 57,481 at $0.007. MI355X is 77% cheaper per token; MI355X delivers 5.9× the tok/s/chip.
Setting 45 tok/s/user as the target on Qwen 3.5 397B-A17B, MI325X produces 8,066 tok/s/chip ($0.038 per million tokens) and MI355X produces 56,717 ($0.007). MI355X is 81% cheaper per token; MI355X delivers 7.0× the tok/s/chip.
At 52 tok/s/user interactivity on Qwen 3.5 397B-A17B, MI325X delivers 6,762 tok/s/chip at $0.045 per million tokens; MI355X delivers 52,689 tok/s/chip at $0.008. MI355X is 82% cheaper per token; MI355X delivers 7.8× the tok/s/chip at this point. (Numbers reflect the default agentic-traces · fp8 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
Interpolated from real benchmark data. Edit target interactivity values below to compare at different operating points.
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | MI325X9,668MI355X57,481 | MI325X8,066MI355X56,717 | MI325X6,762MI355X52,689 |
| Cost ($/M tok) | MI325X$0.032MI355X$0.007 | MI325X$0.038MI355X$0.007 | MI325X$0.045MI355X$0.008 |
| tok/s/MW | MI325X5,720,808MI355X27,503,088 | MI325X4,772,527MI355X27,137,324 | MI325X4,001,198MI355X25,209,846 |
| Concurrency | MI325X~20MI355X~70 | MI325X~9MI355X~67 | MI325X~9MI355X~57 |