Sorry, you need to enable JavaScript to visit this website.

facebooktwittermailshare

CUTENSOR-TUBAL: OPTIMIZED GPU LIBRARY FOR LOW-TUBAL-RANK TENSORS

Abstract: 

In this paper, we optimize the computations of third-order low-tubal-rank tensor operations on many-core GPUs. Tensor operations are compute-intensive and existing studies optimize such operations in a case-by-case manner, which can be inefficient and error-prone. We develop and optimize a BLAS-like library for the low-tubal-rank tensor model called cuTensor-tubal, which includes efficient GPU primitives for tensor operations and key processes. We compute tensor operations in the frequency domain and fully exploit tube-wise and slice-wise parallelisms. We design, implement, and optimize four key tensor operations namely t-FFT, inverse t-FFT, t-product, and t-SVD. For t-product and t-SVD, cuTensor-tubal demonstrates significant speedups: maximum 29.16X, 6.72X speedups over the non-optimized GPU counterparts, and maximum 16.91X and 27.03X speedups over the CPU implementations running on dual 10-core Xeon CPUs.

Paper link on IEEE Xplore:https://ieeexplore.ieee.org/document/8682323

up
0 users have voted:

Paper Details

Authors:
Tao Zhang, Xiao-Yang Liu
Submitted On:
8 May 2019 - 8:21am
Short Link:
Type:
Poster
Event:
Presenter's Name:
HAI LI
Paper Code:
2071
Document Year:
2019
Cite

Document Files

The poster for paper entitled "CUTENSOR-TUBAL: OPTIMIZED GPU LIBRARY FOR LOW-TUBAL-RANK TENSORS"

(67)

Subscribe

[1] Tao Zhang, Xiao-Yang Liu, "CUTENSOR-TUBAL: OPTIMIZED GPU LIBRARY FOR LOW-TUBAL-RANK TENSORS", IEEE SigPort, 2019. [Online]. Available: http://sigport.org/4082. Accessed: Dec. 06, 2019.
@article{4082-19,
url = {http://sigport.org/4082},
author = {Tao Zhang; Xiao-Yang Liu },
publisher = {IEEE SigPort},
title = {CUTENSOR-TUBAL: OPTIMIZED GPU LIBRARY FOR LOW-TUBAL-RANK TENSORS},
year = {2019} }
TY - EJOUR
T1 - CUTENSOR-TUBAL: OPTIMIZED GPU LIBRARY FOR LOW-TUBAL-RANK TENSORS
AU - Tao Zhang; Xiao-Yang Liu
PY - 2019
PB - IEEE SigPort
UR - http://sigport.org/4082
ER -
Tao Zhang, Xiao-Yang Liu. (2019). CUTENSOR-TUBAL: OPTIMIZED GPU LIBRARY FOR LOW-TUBAL-RANK TENSORS. IEEE SigPort. http://sigport.org/4082
Tao Zhang, Xiao-Yang Liu, 2019. CUTENSOR-TUBAL: OPTIMIZED GPU LIBRARY FOR LOW-TUBAL-RANK TENSORS. Available at: http://sigport.org/4082.
Tao Zhang, Xiao-Yang Liu. (2019). "CUTENSOR-TUBAL: OPTIMIZED GPU LIBRARY FOR LOW-TUBAL-RANK TENSORS." Web.
1. Tao Zhang, Xiao-Yang Liu. CUTENSOR-TUBAL: OPTIMIZED GPU LIBRARY FOR LOW-TUBAL-RANK TENSORS [Internet]. IEEE SigPort; 2019. Available from : http://sigport.org/4082