Abstracting Sparse DNN Acceleration via Structured Sparse Tensor Decomposition

arXiv:2403.07953v1 Announce Type: new
Abstract: Exploiting sparsity in deep neural networks (DNNs) has been a promising area to meet the growing computation need of modern DNNs. However, in practice, sparse DNN acceleration still faces a key challenge. To minimize the overhead of sparse acceleration, hardware designers have proposed structured sparse hardware support recently, which provides limited flexibility and requires extra model fine-tuning. Moreover, any sparse model fine-tuned for certain structured sparse hardware cannot be accelerated by other structured hardware. To …

abstract arxiv challenge computation cs.ai cs.ar cs.lg designers dnn hardware however key modern networks neural networks practice sparsity tensor type via