Qwen 3.5 397B-A17B · Chip comparison
Qwen 3.5 397B-A17B — B200 vs MI300X
Head-to-head AI inference benchmark comparison of B200 (NVIDIA Blackwell) and MI300X (AMD CDNA 3) 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 18–198 tok/s/user interactivity band, at 63 tok/s/user on Qwen 3.5 397B-A17B: B200 runs 37,618 tok/s/chip at $0.013 per million tokens, MI300X runs 10,852 at $0.024. B200 is 47% cheaper per token; B200 delivers 3.5× the tok/s/chip.
Setting 108 tok/s/user as the target on Qwen 3.5 397B-A17B, B200 produces 35,473 tok/s/chip ($0.014 per million tokens) and MI300X produces 6,568 ($0.040). B200 is 66% cheaper per token; B200 delivers 5.4× the tok/s/chip.
At 153 tok/s/user interactivity on Qwen 3.5 397B-A17B, B200 delivers 30,355 tok/s/chip at $0.016 per million tokens; MI300X delivers 4,114 tok/s/chip at $0.064. B200 is 75% cheaper per token; B200 delivers 7.4× the tok/s/chip at this point. (Numbers reflect the default agentic-traces · fp8 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
Interpolated from real benchmark data. Edit target interactivity values below to compare at different operating points.
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | B20037,618MI300X10,852 | B20035,473MI300X6,568 | B20030,355MI300X4,114 |
| Cost ($/M tok) | B200$0.013MI300X$0.024 | B200$0.014MI300X$0.040 | B200$0.016MI300X$0.064 |
| tok/s/MW | B20021,998,821MI300X7,807,218 | B20020,744,416MI300X4,725,518 | B20017,751,291MI300X2,960,037 |
| Concurrency | B200~32MI300X~20 | B200~54MI300X~18 | B200~37MI300X~12 |