Kimi K2.5/K2.6/K2.7-Code 1T · MI355X · Precision comparison

MI355X: FP4 vs INT4 precision comparison

How FP4 and INT4 precision affect Kimi K2.5/K2.6/K2.7-Code 1T 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.

Throughput at 27 tok/s/user on Kimi K2.5/K2.6/K2.7-Code 1T (MI355X): FP4 hits 4,934 tok/s/chip, INT4 hits 1,791. Costs per million tokens land at $0.084 and $0.23 respectively. FP4 delivers 176% more tok/s/chip on the same chip, which makes it 64% cheaper per token. The cost-throughput tradeoff from lower precision is only part of the picture — see the evaluation page for accuracy data.

Around the middle of the 11–78 tok/s/user interactivity band, at 44 tok/s/user on Kimi K2.5/K2.6/K2.7-Code 1T (MI355X): FP4 runs 3,522 tok/s/chip at $0.12 per million tokens, INT4 runs 1,224 at $0.34. FP4 delivers 188% more tok/s/chip on the same chip, which makes it 65% cheaper per token. Precision changes affect both inference speed and model quality — consult the evaluation tab for accuracy benchmarks.

At 61 tok/s/user on Kimi K2.5/K2.6/K2.7-Code 1T (MI355X), FP4 delivers 2,692 tok/s/chip at $0.15 per million tokens; INT4 delivers 863 tok/s/chip at $0.48. FP4 delivers 3.1× the tok/s/chip on the same chip, which makes it 68% cheaper per token. Lower-precision quantization trades model accuracy for throughput — check the evaluation page for quality impact. (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)
FP44,934INT41,791
FP43,522INT41,224
FP42,692INT4863
Cost ($/M tok)
FP4$0.084INT4$0.23
FP4$0.12INT4$0.34
FP4$0.15INT4$0.48
tok/s/MW
FP42,360,720INT4856,871
FP41,685,225INT4585,711
FP41,287,933INT4412,954
Concurrency
FP4~250INT4~31
FP4~49INT4~13
FP4~20INT4~6
Kimi K2.5/K2.6/K2.7-Code 1T: FP4 versus INT4 throughput and cost at matched interactivity levels on MI355X
FP4 versus INT4 throughput and cost per million tokens on MI355X for this comparison's canonical default workload
Open full-size image ↗