Qwen 3.5 397B-A17B · Chip comparison

Qwen 3.5 397B-A17B — H100 vs MI355X

Head-to-head AI inference benchmark comparison of H100 (NVIDIA Hopper) 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 →

Setting 81 tok/s/user as the target on Qwen 3.5 397B-A17B, H100 produces 6,053 tok/s/chip ($0.054 per million tokens) and MI355X produces 45,394 ($0.009). MI355X is 83% cheaper per token; MI355X delivers 7.5× the tok/s/chip.

At 130 tok/s/user interactivity on Qwen 3.5 397B-A17B, H100 delivers 3,379 tok/s/chip at $0.096 per million tokens; MI355X delivers 31,665 tok/s/chip at $0.013. MI355X is 86% cheaper per token; MI355X delivers 9.4× the tok/s/chip at this point.

H100 posts 2,476 tok/s/chip for $0.13 per million tokens at 179 tok/s/user on Qwen 3.5 397B-A17B; MI355X posts 21,816 tok/s/chip for $0.019. MI355X is 85% cheaper per token; MI355X delivers 8.8× the tok/s/chip. (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.)

View performance per dollar →

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)
H1006,053MI355X45,394
H1003,379MI355X31,665
H1002,476MI355X21,816
Cost ($/M tok)
H100$0.054MI355X$0.009
H100$0.096MI355X$0.013
H100$0.13MI355X$0.019
tok/s/MW
H1004,418,126MI355X21,719,712
H1002,466,146MI355X15,150,628
H1001,807,548MI355X10,438,147
Concurrency
H100~12MI355X~44
H100~6MI355X~23
H100~4MI355X~14