MI300X: FP8 vs BF16 Precision Comparison
How FP8 and BF16 precision affect Qwen 3.5 397B-A17B inference on MI300X (AMD CDNA 3). Throughput, latency, and cost across LLM workloads. Use the chart controls below to switch sequences and metrics — same interactions as the main inference chart.
At 45 tok/s/user on Qwen 3.5 397B-A17B (MI300X), FP8 delivers 312 tok/s/chip at $0.84 per million tokens; BF16 delivers 414 tok/s/chip at $0.64. BF16 is 32% cheaper per token; BF16 delivers 32% more tok/s/chip. Lower-precision quantization trades model accuracy for throughput — check the evaluation page for quality impact.
FP8 posts 179 tok/s/chip for $1.47 per million tokens at 54 tok/s/user on Qwen 3.5 397B-A17B (MI300X); BF16 posts 315 tok/s/chip for $0.84. BF16 is 76% cheaper per token; BF16 delivers 76% more tok/s/chip. Quantization-level accuracy differences are tracked on the evaluation tab.
Throughput at 63 tok/s/user on Qwen 3.5 397B-A17B (MI300X): FP8 hits 116 tok/s/chip, BF16 hits 258. Per-million costs land at $2.27 and $1.02 respectively. BF16 is 121% cheaper per token; BF16 delivers 121% more tok/s/chip. The cost-throughput tradeoff from lower precision is only part of the picture — see the evaluation page for accuracy data. (Numbers reflect the default 1k/1k selection for this URL — table and chart below update if you change sequence or model in the controls. Each side uses the best available serving configuration for that precision, which may include speculative decoding such as MTP where recipes exist — the same convention as the other comparison pages.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | FP8:312.4BF16:413.6 | FP8:179.1BF16:315.4 | FP8:116.3BF16:257.6 |
| Cost ($/M tok) | FP8:$0.845BF16:$0.638 | FP8:$1.474BF16:$0.837 | FP8:$2.268BF16:$1.024 |
| tok/s/MW | FP8:224732BF16:297572 | FP8:128820BF16:226906 | FP8:83694BF16:185318 |
| Concurrency | FP8:~29BF16:~39 | FP8:~14BF16:~23 | FP8:~8BF16:~17 |
Inference Performance
Agentic inference metrics from the AgentX scenario and fixed-sequence inference metrics across models, hardware configurations, and serving parameters.