TurboDiffusion: Accelerating Video Diffusion Models by 100-200 Times

Abstract

We introduce TurboDiffusion, a video generation acceleration framework that can speed up end-to-end diffusion generation by 100-200× while maintaining video quality. TurboDiffusion mainly relies on several components for acceleration: (1) Attention acceleration: TurboDiffusion uses low-bit SageAttention and trainable Sparse-Linear Attention (SLA) to speed up attention computation. (2) Step distillation: TurboDiffusion adopts rCM for efficient step distillation. (3) W8A8 quantization: TurboDiffusion quantizes model parameters and activations to 8 bits to accelerate linear layers and compress the model. We conduct experiments on the Wan2.2-I2V-A14B-720P, Wan2.1-T2V-1.3B-480P, Wan2.1-T2V-14B-720P, and Wan2.1-T2V-14B-480P models. Experimental results show that TurboDiffusion achieves 100-200× speedup for video generation on a single RTX 5090 GPU, while maintaining comparable video quality. The GitHub repository, which contains model checkpoints, training, and inference code, is available at https://github.com/thu-ml/TurboDiffusion.