Systolic optimization on GPU platforms

作者: Enrique Alba , Pablo Vidal

DOI: 10.1007/978-3-642-27549-4_48

关键词:

摘要: The article presents a systolic algorithm implemented using NVIDIA's Compute Unified Device Architecture (CUDA). works as general disposition of the elements in mesh by sinchronously computing basic solutions among processing elements. We have used instances Subset Sum Problem for evaluating to study behavior proposed model. experimental results show that approach is very efficient, especially large problem and consumes shorter times compared other algorithms like parallel Genetic Algorithms Random Search.

参考文章(11)
Charles E Leiserson, H T Kung, Systolic Arrays for (VLSI). ,(1978)
Vipin Kumar, Introduction to Parallel Computing ipc. ,(2002)
Kung, Why systolic architectures IEEE Computer. ,vol. 15, pp. 300- 309 ,(1982) , 10.1109/MC.1982.1653825
H. T. Kung, Philip L. Lehman, Systolic (VLSI) arrays for relational database operations Proceedings of the 1980 ACM SIGMOD international conference on Management of data - SIGMOD '80. pp. 105- 116 ,(1980) , 10.1145/582250.582267
G.M. Megson, I.M. Bland, Synthesis of a systolic array genetic algorithm merged international parallel processing symposium and symposium on parallel and distributed processing. pp. 316- 320 ,(1998) , 10.1109/IPPS.1998.669933
Pinaki Mazumder, Heming Chan, A Systolic Architecture for High Speed Hypergraph Partitioning Using a Genetic Algorithm. Evo Workshops. pp. 109- 126 ,(1994)