Multi-GPU Development of a Neural Networks Based Reconstructor for Adaptive Optics
Aberrations introduced by the atmospheric turbulence in large telescopes are compensated using adaptive optics systems, where the use of deformable mirrors and multiple sensors relies on complex control systems. Recently, the development of larger scales of telescopes as the E-ELT or TMT has created...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2018-01-01
|
| Series: | Complexity |
| Online Access: | http://dx.doi.org/10.1155/2018/5348265 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832548541041475584 |
|---|---|
| author | Carlos González-Gutiérrez María Luisa Sánchez-Rodríguez José Luis Calvo-Rolle Francisco Javier de Cos Juez |
| author_facet | Carlos González-Gutiérrez María Luisa Sánchez-Rodríguez José Luis Calvo-Rolle Francisco Javier de Cos Juez |
| author_sort | Carlos González-Gutiérrez |
| collection | DOAJ |
| description | Aberrations introduced by the atmospheric turbulence in large telescopes are compensated using adaptive optics systems, where the use of deformable mirrors and multiple sensors relies on complex control systems. Recently, the development of larger scales of telescopes as the E-ELT or TMT has created a computational challenge due to the increasing complexity of the new adaptive optics systems. The Complex Atmospheric Reconstructor based on Machine Learning (CARMEN) is an algorithm based on artificial neural networks, designed to compensate the atmospheric turbulence. During recent years, the use of GPUs has been proved to be a great solution to speed up the learning process of neural networks, and different frameworks have been created to ease their development. The implementation of CARMEN in different Multi-GPU frameworks is presented in this paper, along with its development in a language originally developed for GPU, like CUDA. This implementation offers the best response for all the presented cases, although its advantage of using more than one GPU occurs only in large networks. |
| format | Article |
| id | doaj-art-606296336be64c7e8407da8864bfd55d |
| institution | Kabale University |
| issn | 1076-2787 1099-0526 |
| language | English |
| publishDate | 2018-01-01 |
| publisher | Wiley |
| record_format | Article |
| series | Complexity |
| spelling | doaj-art-606296336be64c7e8407da8864bfd55d2025-02-03T06:13:49ZengWileyComplexity1076-27871099-05262018-01-01201810.1155/2018/53482655348265Multi-GPU Development of a Neural Networks Based Reconstructor for Adaptive OpticsCarlos González-Gutiérrez0María Luisa Sánchez-Rodríguez1José Luis Calvo-Rolle2Francisco Javier de Cos Juez3Department of Exploitation and Exploration of Mines, University of Oviedo, Oviedo, SpainDepartment of Physics, University of Oviedo, Oviedo, SpainDepartment of Industrial Engineering, University of A Coruña, Ferrol, A Coruña, SpainDepartment of Exploitation and Exploration of Mines, University of Oviedo, Oviedo, SpainAberrations introduced by the atmospheric turbulence in large telescopes are compensated using adaptive optics systems, where the use of deformable mirrors and multiple sensors relies on complex control systems. Recently, the development of larger scales of telescopes as the E-ELT or TMT has created a computational challenge due to the increasing complexity of the new adaptive optics systems. The Complex Atmospheric Reconstructor based on Machine Learning (CARMEN) is an algorithm based on artificial neural networks, designed to compensate the atmospheric turbulence. During recent years, the use of GPUs has been proved to be a great solution to speed up the learning process of neural networks, and different frameworks have been created to ease their development. The implementation of CARMEN in different Multi-GPU frameworks is presented in this paper, along with its development in a language originally developed for GPU, like CUDA. This implementation offers the best response for all the presented cases, although its advantage of using more than one GPU occurs only in large networks.http://dx.doi.org/10.1155/2018/5348265 |
| spellingShingle | Carlos González-Gutiérrez María Luisa Sánchez-Rodríguez José Luis Calvo-Rolle Francisco Javier de Cos Juez Multi-GPU Development of a Neural Networks Based Reconstructor for Adaptive Optics Complexity |
| title | Multi-GPU Development of a Neural Networks Based Reconstructor for Adaptive Optics |
| title_full | Multi-GPU Development of a Neural Networks Based Reconstructor for Adaptive Optics |
| title_fullStr | Multi-GPU Development of a Neural Networks Based Reconstructor for Adaptive Optics |
| title_full_unstemmed | Multi-GPU Development of a Neural Networks Based Reconstructor for Adaptive Optics |
| title_short | Multi-GPU Development of a Neural Networks Based Reconstructor for Adaptive Optics |
| title_sort | multi gpu development of a neural networks based reconstructor for adaptive optics |
| url | http://dx.doi.org/10.1155/2018/5348265 |
| work_keys_str_mv | AT carlosgonzalezgutierrez multigpudevelopmentofaneuralnetworksbasedreconstructorforadaptiveoptics AT marialuisasanchezrodriguez multigpudevelopmentofaneuralnetworksbasedreconstructorforadaptiveoptics AT joseluiscalvorolle multigpudevelopmentofaneuralnetworksbasedreconstructorforadaptiveoptics AT franciscojavierdecosjuez multigpudevelopmentofaneuralnetworksbasedreconstructorforadaptiveoptics |