%0 Conference Proceedings
%T Segmented Merge: A New Primitive for Parallel Sparse Matrix Computations
%+ China University of Petroleum Beijing (CUP)
%+ Virginia Tech [Blacksburg]
%+ Department of Computer Science and Engineering [Colombus]
%+ Aarhus University [Aarhus]
%A Ji, Haonan
%A Lu, Shibo
%A Hou, Kaixi
%A Wang, Hao
%A Liu, Weifeng
%A Vinter, Brian
%Z Part 3: Algorithm
%< avec comité de lecture
%( Lecture Notes in Computer Science
%B 17th IFIP International Conference on Network and Parallel Computing (NPC)
%C Zhengzhou, China
%Y Xin He
%Y En Shao
%Y Guangming Tan
%I Springer International Publishing
%3 Network and Parallel Computing
%V LNCS-12639
%P 170-181
%8 2020-09-28
%D 2020
%R 10.1007/978-3-030-79478-1_15
%K Parallel computing
%K Segmented merge
%K Sparse matrix
%K GPU
%Z Computer Science [cs]Conference papers
%X Segmented operations, such as segmented sum, segmented scan and segmented sort, are important building blocks for parallel irregular algorithms. We in this work propose a new parallel primitive called segmented merge. Its function is in parallel merging q sub-segments to p segments, both of nonuniform lengths. We implement the segmented merge primitive on GPUs and demonstrate its efficiency on parallel sparse matrix transposition (SpTRANS) and sparse matrix-matrix multiplication (SpGEMM) operations.
%G English
%Z TC 10
%Z WG 10.3
%2 https://inria.hal.science/hal-03768762/document
%2 https://inria.hal.science/hal-03768762/file/511910_1_En_15_Chapter.pdf
%L hal-03768762
%U https://inria.hal.science/hal-03768762
%~ IFIP-LNCS
%~ IFIP
%~ IFIP-TC
%~ IFIP-TC10
%~ IFIP-NPC
%~ IFIP-WG10-3
%~ IFIP-LNCS-12639