Page 1 of 1
will ?gees (Schur decomposition) be supported?
Posted: Wed Aug 21, 2013 8:35 am
Is by any chance ?gees in the pipeline for development as part of MAGMA?
Re: will ?gees (Schur decomposition) be supported?
Posted: Wed Aug 21, 2013 2:26 pm
Not in the near-term. The eigensolvers themselves (as opposed to the Hessenberg or tridiagonal reduction) are fairly complicated codes that employ BLAS-1 and BLAS-2 functions, which benefit less from the GPU's parallelism. Possibly accumulation of the Schur vectors Z could be done on the GPU. However, the total geev time is now actually dominated by computing the eigenvectors in trevc, which we are looking at.