AI Paper Atlas

Research map / Video / Video generation

BWCache: Accelerating Video Diffusion Transformers through Block-Wise Caching

Hanshuai Cui, Zhiqing Tang, Zhifei Xu, Zhi Yao, Wenyi Zeng, Weijia Jia

ICLR 2026 · Video › Video generation · approach: Efficient video diffusion models · 8 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use BWCache instead of TeaCacheTeaCache · Open-Soraspeedup: 1.50x -> 1.61xImproved
use BWCache instead of TeaCacheTeaCache · Open-Sora-Planspeedup: 4.41x -> 2.24xHurt
use BWCache instead of TeaCacheTeaCache · Open-SoraLPIPS: 0.1496 -> 0.0879Improved
apply BWCache to modelOpen-Sora · VBenchVBench: 80.33% -> 80.03%No clear effect
apply BWCache to modelLatte · VBenchVBench: 78.95% -> 78.28%Hurt
apply BWCache instead of Skip-DiTSkip-DiT · LatteVBench: 75.76% -> 78.28%Improved
raise indicator threshold from 0.15 to 0.25 (ablation)BWCache · Open-SoraVBench: 80.03% -> 79.03%Hurt
lengthen reuse interval from 5% to 25% (ablation)BWCache · Open-SoraVBench: 80.28% -> 79.31%Hurt
lengthen reuse interval from 5% to 25% (ablation)BWCache · Open-Soralatency (s): 34.70 -> 25.76Improved
use caching instead of reducing steps to 19 (ablation)Open-Sora · Open-SoraLPIPS: 0.3139 -> 0.0879Improved
scale to 8 GPUs with DSPBWCache · Open-Sora (204 frames)speedup: 2.4x -> 17.2xImproved
use BWCache instead of TeaCache with 8 GPUsTeaCache · Open-Sora-Plan (221 frames)speedup: 28.1x -> 13.3xHurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy