DeepSeek V4 Pro 0813 1.6T · MI355X · Precision comparison

MI355X: FP4 vs FP8 precision comparison

How FP4 and FP8 precision affect DeepSeek V4 Pro 0813 1.6T inference on MI355X (AMD CDNA 4). Throughput, latency, and cost across LLM workloads. Use the chart controls below to switch sequences and metrics — same interactions as the main inference chart.

FP4 posts 4,294 tok/s/chip for $0.097 per million tokens at 34 tok/s/user on DeepSeek V4 Pro 0813 1.6T (MI355X); FP8 posts 20 tok/s/chip for $20.4. FP4 delivers 210× the tok/s/chip on the same chip, which makes it 99.5% cheaper per token. Quantization-level accuracy differences are tracked on the evaluation tab.

Throughput at 67 tok/s/user on DeepSeek V4 Pro 0813 1.6T (MI355X): FP4 hits 1,733 tok/s/chip, FP8 hits 20. Costs per million tokens land at $0.24 and $20.4 respectively. FP4 delivers 85× the tok/s/chip on the same chip, which makes it 99% cheaper per token. The cost-throughput tradeoff from lower precision is only part of the picture — see the evaluation page for accuracy data.

Toward the upper edge of the 1–133 tok/s/user interactivity band, at 100 tok/s/user on DeepSeek V4 Pro 0813 1.6T (MI355X): FP4 runs 851 tok/s/chip at $0.49 per million tokens, FP8 runs 20 at $20.4. FP4 delivers 42× the tok/s/chip on the same chip, which makes it 98% cheaper per token. Precision changes affect both inference speed and model quality — consult the evaluation tab for accuracy benchmarks. (Numbers reflect the default 8k/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.)

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)
FP44,294FP820
FP41,733FP820
FP4851FP820
Cost ($/M tok)
FP4$0.097FP8$20.4
FP4$0.24FP8$20.4
FP4$0.49FP8$20.4
tok/s/MW
FP42,054,308FP89,764
FP4829,088FP89,764
FP4407,322FP89,764
Concurrency
FP4~512FP8~8
FP4~32FP8~8
FP4~8FP8~8
DeepSeek V4 Pro 0813 1.6T: FP4 versus FP8 throughput and cost at matched interactivity levels on MI355X
FP4 versus FP8 throughput and cost per million tokens on MI355X for this comparison's canonical default workload
Open full-size image ↗