SparSP: Exploiting Communication Sparsity for Sequence-Parallel Video DiTs
This paper argues that attention sparsity should be treated as a communication primitive and presents SparSP, an efficient sparse sequence parallel communication system that co-designs token distribution, communication routing, and asynchronous execution for sparse video diffusion models.
De-Sen Sun, Xin-Rui Zhong, Yu-Ke Wang et al.
· 0 citations