Meeting the Real-Time Challenges of Ground-Based Telescopes Using Low-Rank Matrix Computations

Hatem Ltaief; Jesse Cranney; Damien Gratadour; Yuxi Hong; Laurent Gatineau; David E. Keyes

doi:10.1145/3458817.3476225

Meeting the Real-Time Challenges of Ground-Based Telescopes Using Low-Rank Matrix Computations

Hatem Ltaief, Jesse Cranney, Damien Gratadour, Yuxi Hong, Laurent Gatineau, David E. Keyes

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

11 Scopus citations

Abstract

Adaptive Optics (AO) is a technology that permits to measure and mitigate the distortion effects of atmospheric turbulence on optical beams. AO must operate in real-Time by controlling thousands of actuators to shape the surface of deformable mirrors deployed on ground-based telescopes to compensate for these distortions. The command vectors that trigger how each individual actuator should act to bend a portion of the mirror are obtained from Matrix-Vector Multiplications (MVM). We identify and leverage the data sparsity structure of these control matrices coming from the MAVIS instruments for the European Southern Observatory s Very Large Telescope. We provide performance evaluation on x86 and acceleratorbased systems.We present the impact of tile low-rank (TLR) matrix approximations on time-To-solution for the MVM and assess the produced image quality. We achieve performance improvement up to two orders of magnitude for TLR-MVM compared to regular dense MVM, while maintaining the image quality.

Original language	English (US)
Title of host publication	Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis
Publisher	ACM
ISBN (Print)	9781450384421
DOIs	https://doi.org/10.1145/3458817.3476225
State	Published - Nov 13 2021

Bibliographical note

KAUST Repository Item: Exported on 2022-02-11
Acknowledgements: ENTS The authors would like thank Fujitsu limited for their support on the evaluation environment and AMD/NVIDIA for the remote accesses steto their respective GPU-based systems.

Access to Document

10.1145/3458817.3476225

https://repository.kaust.edu.sa/bitstream/10754/669813/3/3458817.3476225.pdf

Cite this

@inproceedings{7192c0989a09470bbfaf70f1a2aa1846,

title = "Meeting the Real-Time Challenges of Ground-Based Telescopes Using Low-Rank Matrix Computations",

abstract = "Adaptive Optics (AO) is a technology that permits to measure and mitigate the distortion effects of atmospheric turbulence on optical beams. AO must operate in real-Time by controlling thousands of actuators to shape the surface of deformable mirrors deployed on ground-based telescopes to compensate for these distortions. The command vectors that trigger how each individual actuator should act to bend a portion of the mirror are obtained from Matrix-Vector Multiplications (MVM). We identify and leverage the data sparsity structure of these control matrices coming from the MAVIS instruments for the European Southern Observatory s Very Large Telescope. We provide performance evaluation on x86 and acceleratorbased systems.We present the impact of tile low-rank (TLR) matrix approximations on time-To-solution for the MVM and assess the produced image quality. We achieve performance improvement up to two orders of magnitude for TLR-MVM compared to regular dense MVM, while maintaining the image quality.",

author = "Hatem Ltaief and Jesse Cranney and Damien Gratadour and Yuxi Hong and Laurent Gatineau and Keyes, {David E.}",

note = "KAUST Repository Item: Exported on 2022-02-11 Acknowledgements: ENTS The authors would like thank Fujitsu limited for their support on the evaluation environment and AMD/NVIDIA for the remote accesses steto their respective GPU-based systems.",

year = "2021",

month = nov,

day = "13",

doi = "10.1145/3458817.3476225",

language = "English (US)",

isbn = "9781450384421",

booktitle = "Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis",

publisher = "ACM",

}

TY - GEN

T1 - Meeting the Real-Time Challenges of Ground-Based Telescopes Using Low-Rank Matrix Computations

AU - Ltaief, Hatem

AU - Cranney, Jesse

AU - Gratadour, Damien

AU - Hong, Yuxi

AU - Gatineau, Laurent

AU - Keyes, David E.

N1 - KAUST Repository Item: Exported on 2022-02-11 Acknowledgements: ENTS The authors would like thank Fujitsu limited for their support on the evaluation environment and AMD/NVIDIA for the remote accesses steto their respective GPU-based systems.

PY - 2021/11/13

Y1 - 2021/11/13

N2 - Adaptive Optics (AO) is a technology that permits to measure and mitigate the distortion effects of atmospheric turbulence on optical beams. AO must operate in real-Time by controlling thousands of actuators to shape the surface of deformable mirrors deployed on ground-based telescopes to compensate for these distortions. The command vectors that trigger how each individual actuator should act to bend a portion of the mirror are obtained from Matrix-Vector Multiplications (MVM). We identify and leverage the data sparsity structure of these control matrices coming from the MAVIS instruments for the European Southern Observatory s Very Large Telescope. We provide performance evaluation on x86 and acceleratorbased systems.We present the impact of tile low-rank (TLR) matrix approximations on time-To-solution for the MVM and assess the produced image quality. We achieve performance improvement up to two orders of magnitude for TLR-MVM compared to regular dense MVM, while maintaining the image quality.

AB - Adaptive Optics (AO) is a technology that permits to measure and mitigate the distortion effects of atmospheric turbulence on optical beams. AO must operate in real-Time by controlling thousands of actuators to shape the surface of deformable mirrors deployed on ground-based telescopes to compensate for these distortions. The command vectors that trigger how each individual actuator should act to bend a portion of the mirror are obtained from Matrix-Vector Multiplications (MVM). We identify and leverage the data sparsity structure of these control matrices coming from the MAVIS instruments for the European Southern Observatory s Very Large Telescope. We provide performance evaluation on x86 and acceleratorbased systems.We present the impact of tile low-rank (TLR) matrix approximations on time-To-solution for the MVM and assess the produced image quality. We achieve performance improvement up to two orders of magnitude for TLR-MVM compared to regular dense MVM, while maintaining the image quality.

UR - http://hdl.handle.net/10754/669813

UR - https://dl.acm.org/doi/10.1145/3458817.3476225

UR - http://www.scopus.com/inward/record.url?scp=85119993532&partnerID=8YFLogxK

U2 - 10.1145/3458817.3476225

DO - 10.1145/3458817.3476225

M3 - Conference contribution

SN - 9781450384421

BT - Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis

PB - ACM

ER -

Meeting the Real-Time Challenges of Ground-Based Telescopes Using Low-Rank Matrix Computations

Abstract

Bibliographical note

Access to Document

Other files and links

Fingerprint

Cite this