HOSVD prototype based on modular SW libraries running on a high-performance CPU+GPU platform
作者： R.I. Acosta-Quiñonez , D. Torres-Roman , R. Rodriguez-Avila
DOI: 10.1016/J.SYSARC.2020.101897
关键词: Software 、 Symmetric multiprocessor system 、 Speedup 、 Computer science 、 Modularity 、 Throughput (business) 、 Modular design 、 Overhead (computing) 、 Computer architecture 、 Reusability
摘要: Abstract Efficient prototyping is an invaluable resource for modern enterprises and research centers. An efficient tool exhibits high throughput while maintaining flexibility, reduces design validation efforts, resulting in low time-to-market competitiveness. This paper presents a modular implementation of high-performance software (SW) libraries running on Heterogeneous Computing Platform (HCP) based CPU+GPU. The proposed SW enable fast easy comparison prototype under different criteria maintain reusability due to their definition. These features accelerate the task by removing overhead designing validating ad-hoc implementations. novelty benefits this proposal are presented analysis multilinear SVD or Higher-Order (HOSVD), important, widely-used, computationally demanding tensor decomposition. mean square error (MSE), processing time, speedup case study show its performance, modularity maintains flexibility. HOSVD reaches maximum 17 × that one most important implementations state art.
sci-hub.st 参考文章(51)
Stanley I Grossman, Elementary Linear Algebra ,(1980)
R. Sepulchre, R. Mahony, P.-A. Absil, Optimization Algorithms on Matrix Manifolds ,(2007)
Woody Austin, Grey Ballard, Tamara G. Kolda, Parallel Tensor Compression for Large-Scale Scientific Data international parallel and distributed processing symposium. pp. 912- 922 ,(2016) , 10.1109/IPDPS.2016.67
Tyng-Yeu Liang, Hung-Fu Li, Yu-Jie Lin, Bi-Shing Chen, A Distributed PTX Virtual Machine on Hybrid CPU/GPU Clusters Journal of Systems Architecture. ,vol. 62, pp. 63- 77 ,(2016) , 10.1016/J.SYSARC.2015.10.003
Franeois Quitin, Muhammad Mahboob Ur Rahman, Raghuraman Mudumbai, Upamanyu Madhow, A Scalable Architecture for Distributed Transmit Beamforming with Commodity Radios: Design and Proof of Concept IEEE Transactions on Wireless Communications. ,vol. 12, pp. 1418- 1428 ,(2013) , 10.1109/TWC.2013.012513.121029
J.I. Park, Y. Lee, S. Yoon, Partition-Based Hybrid Decoding (PHD): A Class of ML Decoding Schemes for MIMO Signals Based on Tree Partitioning and Combined Depth- and Breadth-First Search Journal of Applied Research and Technology. ,vol. 11, pp. 213- 224 ,(2013) , 10.1016/S1665-6423(13)71531-1
Lieven De Lathauwer, Bart De Moor, Joos Vandewalle, A Multilinear Singular Value Decomposition SIAM Journal on Matrix Analysis and Applications. ,vol. 21, pp. 1253- 1278 ,(2000) , 10.1137/S0895479896305696
Lieven De Lathauwer, Bart De Moor, Joos Vandewalle, On the Best Rank-1 and Rank-(R1 ,R2 ,. . .,RN) Approximation of Higher-Order Tensors SIAM Journal on Matrix Analysis and Applications. ,vol. 21, pp. 1324- 1342 ,(2000) , 10.1137/S0895479898346995
Tamara G. Kolda, Brett W. Bader, Tensor Decompositions and Applications Siam Review. ,vol. 51, pp. 455- 500 ,(2009) , 10.1137/07070111X