How to parallelize many small SVD's on a GPU

8 vues (au cours des 30 derniers jours)
Tim
Tim le 8 Déc 2021
Commenté : Tim le 18 Déc 2021
I'm computing SVD's of millions of small (e.g. 5 x 10) matrices. It is surprisingly fast as a for-loop on the CPU, however I need the computation to be even faster and would like to know if parallelization can be leveraged on a GPU. SVD doesn't currently appear to be a function supported by pagefun for gpuArray objects. I'm using R2020a and the Parallel Computing toolbox. Thanks

Réponse acceptée

Christine Tobler
Christine Tobler le 9 Déc 2021
As of R2021b, there is a function pagesvd which is supported on both CPU and distributed arrays. It's not supported for gpuArray as of now.
  7 commentaires
Joss Knight
Joss Knight le 18 Déc 2021
gpuArray/pagesvd coming next year, watch this space...
Tim
Tim le 18 Déc 2021
Great news, thank you Joss

Connectez-vous pour commenter.

Plus de réponses (1)

Matt J
Matt J le 9 Déc 2021
Modifié(e) : Matt J le 9 Déc 2021
SVD doesn't currently appear to be a function supported by pagefun for gpuArray objects
Perhaps not for gpuArrays, but it is supported for distributed arrays:

Catégories

En savoir plus sur Linear Algebra dans Help Center et File Exchange

Produits


Version

R2020a

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by