Sep 2026· Proceedings of the International Conference on Parallel Processing· 0 citations· 49 references
TL;DR
This work proposes TileSpMM, which breaks the static-granularity bottleneck through a variable-size tiling algorithm that dynamically adapts to local sparsity patterns, and is equipped with an adaptive load-balancing strategy and customized granularity-specific kernels to improve hardware utilization and mitigate computation redundancy.
Abstract
Sparse Matrix-Matrix Multiplication (SpMM) is a fundamental primitive in scientific computing and artificial intelligence applications. Modern hardware, notably Tensor Core Units (TCUs), offers immense computational power, creating promising opportunities for SpMM acceleration. However, it is difficult to map sparse matrices with irregular structures onto TCUs due to their requirement for regular operands. Existing fixed-granularity tiling methods frequently face a trade-off between massive zero-padding in sparse regions and poor spatial data locality in dense regions. To bridge this gap, we propose TileSpMM, which breaks the static-granularity bottleneck through a variable-size tiling algorithm that dynamically adapts to local sparsity patterns. Furthermore, TileSpMM is equipped with an adaptive load-balancing strategy and customized granularity-specific kernels to improve hardware utilization and mitigate computation redundancy. Experiments on NVIDIA H100 and RTX 5090 GPUs with a diverse range of benchmark matrices show that TileSpMM delivers overall better performance than existing SpMM methods across the evaluated platforms and datasets. Compared with cuSPARSE, SSpMM, Acc-SpMM and FlashSparse, TileSpMM achieves geometric mean speedups of 4.76 × , 2.74 × , 2.38 × and 1.58 × , respectively.
AFH-SpMM is a novel Auto-Fit Heterogeneous SpMM framework designed for adaptively parallelizing sparse-dense matrix multiplication on Tensor Core-equipped GPUs that achieves average speedups of 1.33 ×, and often leads cuSPARSE, ASpT, Sputnik, RoDe, Acc-SpMM, and MP-SpMM, with especially strong gains on medium and large...
The proposed DistSpMM proposes DistSpMM, a co-design framework integrating data layout, pipelining, and communication strategies, which introduces HSDMA, a lightweight algorithm that reduces communication by optimizing the dense matrix allocation.
Junyu Gu, Jue Wang, Zhi-Kuang Xin et al.· ACM Transactions on Architec...· 0 citations
CoTC-SpMM is introduced, a cooperative Tensor–CUDA cores scheme for efficient SpMM on GPUs that first proposes the HTC format to partition sparse matrices into dense and sparse components, enabling specialized kernels to leverage the distinct advantages of different computing units and maximize hardware utilization.
Qi Du, Sheng-Le Lin, Yue-Dan Chen et al.· International Conference on...· 0 citations
This work proposes the segmented fiber tree (SFT) data structure, which extends the conventional fiber tree through further partitioning to better support the dataflow paradigm while enhancing data reuse, and decouples the multiplication and merging phases.
Sheng-Bai Luo, Sheng Ma, Bo Wang et al.· ACM Transactions on Architec...· 0 citations
Sparse Matrix-Vector Multiplication (SpMV) is a fundamental sparse kernel in both scientific computing and artificial intelligence. However, the performance of existing SpMV algorithms is highly sensitive to diverse sparsity patterns, necessitating an adaptive framework that can effectively extract matrix structures an...
Guoqing Xiao, Shi-Jie Zhao, Qin Du et al.· IEEE transactions on compute...· 0 citations
Sparse matrix multiplications—including SpMV, SpMM, and SpGEMM—are fundamental to scientific computing, graph analytics, and machine learning. Despite extensive GPU-focused optimizations such as custom sparse formats and load balance, CSR-style and block-based methods can still underexploit fine-grained cache locality...
Xing Cong, Fu-Kai Sun, Yi-Ding Liu et al.· ACM Transactions on Architec...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.