Accelerate Dense Matrix Multiplication on Heterogeneous-GPUs

	Accelerate Dense Matrix Multiplication on Heterogeneous-GPUs
	Sun, Jianan1,2 ; Liao, Mingxue1 ; Chao, Yongyue1,2 ; Lv, Pin1
	2023-12
会议名称	2023 IEEE 29th International Conference on Parallel and Distributed Systems (ICPADS)
会议日期	2023-12
会议地点	Ocean Flower Island, Hainan, China
摘要	Matrix multiplication is crucial in scientific computing, but it demands substantial resources. We propose a framework for effectively utilizing heterogeneous GPUs to large matrix multiplication. By splitting matrices into small blocks and using Douglas’s variant of Strassen’s algorithm, we enable concurrent tasks on heterogeneous systems. Our framework improves speed by 89.5% on homogeneous GPU servers and by 108% in multi-server heterogeneous GPU setups.
七大方向——子方向分类	其他
国重实验室规划方向分类	其他
是否有论文关联数据集需要存交	否
文献类型	会议论文
条目标识符	http://ir.ia.ac.cn/handle/173211/56571
专题	复杂系统认知与决策实验室
通讯作者	Liao, Mingxue
作者单位	1.Institute of Automation, Chinese Academy of Sciences. 2.School of Artificial Intelligence, University of Chinese Academy of Sciences
第一作者单位	中国科学院自动化研究所
通讯作者单位	中国科学院自动化研究所
推荐引用方式 GB/T 7714	Sun, Jianan,Liao, Mingxue,Chao, Yongyue,et al. Accelerate Dense Matrix Multiplication on Heterogeneous-GPUs[C],2023.