Block Cascading: Training Free Acceleration of Block-Causal Video Models
Hmrishav Bandyopadhyay
Abstract
Block Cascading parallelizes video block generation, achieving significant speed improvements without compromising quality.
Block-causal video generation faces a stark speed-quality trade-off: small 1.3B models manage only 16 FPS while large 14B models crawl at 4.5 FPS, forcing users to choose between responsiveness and quality. Block Cascading significantly mitigates this trade-off through training-free parallelization. Our key insight: future video blocks do not need fully denoised current blocks to begin generation. By starting block generation with partially denoised context from predecessors, we transform sequential pipelines into parallel cascades where multiple blocks denoise simultaneously. With 5 GPUs exploiting temporal parallelism, we achieve ~2x acceleration across all model scales: 1.3B models accelerate from 16 to 30 FPS, 14B models from 4.5 to 12.5 FPS. Beyond inference speed, Block Cascading eliminates overhead from KV-recaching (of ~200ms) during context switches for interactive generation. Extensive evaluations validated against multiple block-causal pipelines demonstrate no significant loss in generation quality when switching from block-causal to Block Cascading pipelines for inference. Project Page: https://hmrishavbandy.github.io/block_cascading_page/
Community
Block-causal video generation faces a stark speed-quality trade-off: small 1.3B models manage only 16 FPS while large 14B models crawl at 4.5 FPS, forcing users to choose between responsiveness and quality. Block Cascading significantly mitigates this trade-off through training-free parallelization. Our key insight: future video blocks do not need fully denoised current blocks to begin generation. By starting block generation with partially denoised context from predecessors, we transform sequential pipelines into parallel cascades where multiple blocks denoise simultaneously. With 5 GPUs exploiting temporal parallelism, we achieve ~2x acceleration across all model scales: 1.3B models accelerate from 16 to 30 FPS, 14B models from 4.5 to 12.5 FPS. Beyond inference speed, Block Cascading eliminates overhead from KV-recaching (of ~200ms) during context switches for interactive generation. Extensive evaluations validated against multiple block-causal pipelines demonstrate no significant loss in generation quality when switching from block-causal to Block Cascading pipelines for inference.
Project page: https://hmrishavbandy.github.io/block_cascading_page/
In this work, you achieved a 2× speedup using 5 GPUs. What about Unified Sequence Parallelism? It usually yields near-linear speedup with the number of GPUs in non-causal scenarios. Have you tried its effect in the causal case?
This is an automated message from the Librarian Bot. I found the following papers similar to this paper.
The following papers were recommended by the Semantic Scholar API
- Rolling Forcing: Autoregressive Long Video Diffusion in Real Time (2025)
- MotionStream: Real-Time Video Generation with Interactive Motion Controls (2025)
- Self-Forcing++: Towards Minute-Scale High-Quality Video Generation (2025)
- Infinity-RoPE: Action-Controllable Infinite Video Generation Emerges From Autoregressive Self-Rollout (2025)
- Towards One-step Causal Video Generation via Adversarial Self-Distillation (2025)
- STARFlow-V: End-to-End Video Generative Modeling with Normalizing Flows (2025)
- FlashVSR: Towards Real-Time Diffusion-Based Streaming Video Super-Resolution (2025)
Please give a thumbs up to this comment if you found it helpful!
If you want recommendations for any Paper on Hugging Face checkout this Space
You can directly ask Librarian Bot for paper recommendations by tagging it in a comment:
@librarian-bot
recommend
Models citing this paper 0
No model linking this paper
Datasets citing this paper 0
No dataset linking this paper
Spaces citing this paper 0
No Space linking this paper
Collections including this paper 0
No Collection including this paper