Skip to content

Hybrid Cache Acceleration Benchmark

Baseline SCM S S* SCM F D* SCM U D* +TS +compile +FP8*
24.85s 15.4s 11.4s 8.2s 8.2s πŸŽ‰7.1s πŸŽ‰4.5s

Scheme: DBCache + SCM(steps_computation_mask) + TS(TaylorSeer) + FP8*, L20x1, S*: static cache, D*: dynamic cache, S: Slow, F: Fast, U: Ultra Fast, TS: TaylorSeer, FP8*: FP8 DQ + Sage, FLUX.1-Dev

clip-score-bench

cache-dit will support more mainstream Cache acceleration algorithms in the future. More benchmarks will be released, please stay tuned for update. Here, only the results of some precision and performance benchmarks are presented. The test dataset is DrawBench. For a complete benchmark, please refer to πŸ“šBenchmarks.

Text2Image DrawBench

Comparisons between different FnBn compute block configurations show that more compute blocks result in higher precision. For example, the F8B0_W8MC0 configuration achieves the best Clip Score (33.007) and ImageReward (1.0333). Device: NVIDIA L20. F: Fn_compute_blocks, B: Bn_compute_blocks, 50 steps.

Config Clip Score(↑) ImageReward(↑) PSNR(↑) TFLOPs(↓) SpeedUp(↑)
[FLUX.1-dev]: 50 steps 32.9217 1.0412 INF 3726.87 1.00x
F8B0_W4MC0_R0.08 32.9871 1.0370 33.8317 2064.81 1.80x
F8B0_W4MC2_R0.12 32.9535 1.0185 32.7346 1935.73 1.93x
F8B0_W4MC3_R0.12 32.9234 1.0085 32.5385 1816.58 2.05x
F4B0_W4MC3_R0.12 32.8981 1.0130 31.8031 1507.83 2.47x
F4B0_W4MC4_R0.12 32.8384 1.0065 31.5292 1400.08 2.66x

SOTA Performance

The comparison between cache-dit: DBCache and algorithms such as Ξ”-DiT, Chipmunk, FORA, DuCa, TaylorSeer and FoCa is as follows. Now, in the comparison with a speedup ratio less than 4x, cache-dit achieved the best accuracy. Surprisingly, cache-dit: DBCache still works in the extremely few-step distill model. For a complete benchmark, please refer to πŸ“šBenchmarks. NOTE: Except for DBCache, other performance data are referenced from the paper FoCa, arxiv.2508.16211.

Method TFLOPs(↓) SpeedUp(↑) ImageReward(↑) Clip Score(↑)
[FLUX.1-dev]: 50 steps 3726.87 1.00Γ— 0.9898 32.404
[FLUX.1-dev]: 60% steps 2231.70 1.67Γ— 0.9663 32.312
Ξ”-DiT(N=2) 2480.01 1.50Γ— 0.9444 32.273
Ξ”-DiT(N=3) 1686.76 2.21Γ— 0.8721 32.102
[FLUX.1-dev]: 34% steps 1264.63 3.13Γ— 0.9453 32.114
Chipmunk 1505.87 2.47Γ— 0.9936 32.776
FORA(N=3) 1320.07 2.82Γ— 0.9776 32.266
DBCache(S) 1400.08 2.66Γ— 1.0065 32.838
DuCa(N=5) 978.76 3.80Γ— 0.9955 32.241
TaylorSeer(N=4,O=2) 1042.27 3.57Γ— 0.9857 32.413
DBCache(S)+TS 1153.05 3.23Γ— 1.0221 32.819
DBCache(M) 944.75 3.94Γ— 0.9997 32.849
DBCache(M)+TS 944.75 3.94Γ— 1.0107 32.865
FoCa(N=5): arxiv.2508.16211 893.54 4.16Γ— 1.0029 32.948
[FLUX.1-dev]: 22% steps 818.29 4.55Γ— 0.8183 31.772
FORA(N=7) 670.14 5.55Γ— 0.7418 31.519
ToCa(N=12) 644.70 5.77Γ— 0.7155 31.808
DuCa(N=10) 606.91 6.13Γ— 0.8382 31.759
TeaCache(l=1.2) 669.27 5.56Γ— 0.7394 31.704
TaylorSeer(N=7,O=2) 670.44 5.54Γ— 0.9128 32.128
DBCache(F) 651.90 5.72x 0.9271 32.552
FoCa(N=8): arxiv.2508.16211 596.07 6.24Γ— 0.9502 32.706
DBCache(F)+TS 651.90 5.72x 0.9526 32.568
DBCache(U)+TS 505.47 7.37x 0.8645 32.719

Text2Image Distillation DrawBench

Surprisingly, cache-dit: DBCache still works in the extremely few-step distill model. For example, Qwen-Image-Lightning w/ 4 steps, with the F16B16 configuration, the PSNR is 34.8163, the Clip Score is 35.6109, and the ImageReward is 1.2614. It maintained a relatively high precision.

Config PSNR(↑) Clip Score(↑) ImageReward(↑) TFLOPs(↓) SpeedUp(↑)
[Lightning]: 4 steps INF 35.5797 1.2630 274.33 1.00x
F24B24_W2MC1_R0.8 36.3242 35.6224 1.2630 264.74 1.04x
F16B16_W2MC1_R0.8 34.8163 35.6109 1.2614 244.25 1.12x
F12B12_W2MC1_R0.8 33.8953 35.6535 1.2549 234.63 1.17x
F8B8_W2MC1_R0.8 33.1374 35.7284 1.2517 224.29 1.22x
F1B0_W2MC1_R0.8 31.8317 35.6651 1.2397 206.90 1.33x