Replies: 3 comments
-
Yes, it supports X=A\B where X and B are matrices (dense matrices to be precise, not sparse; only A is sparse). We exploit the parallelism inside the BLAS for this case, which is better than exploiting it per column or groups of columns for X and B. |
Beta Was this translation helpful? Give feedback.
-
That sounds exciting! Is there a dedicated method for this use case? I have some hard time trying to understand how |
Beta Was this translation helpful? Give feedback.
-
You use Paru_Solve (Sym, Num, nrhs, B, X, Control).
…On Wed, Nov 6, 2024 at 8:28 AM Zeyu Mao ***@***.***> wrote:
That sounds exciting! Is there a dedicated method for this use case? I
have some hard time trying to understand how ParU_Solve (Sym, Num, b, x,
&Control) would work with dense matrix B. How to provide necessary
information like dimension of B to the solver?
—
Reply to this email directly, view it on GitHub
<#878 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AEYIIOI54FCCECFLPQYI5T3Z7IRPNAVCNFSM6AAAAABQRDIBG6VHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTCMJWG42DEMY>
.
You are receiving this because you commented.Message ID:
<DrTimothyAldenDavis/SuiteSparse/repo-discussions/878/comments/11167423@
github.com>
|
Beta Was this translation helpful? Give feedback.
-
Congratulations on releasing the stable version of ParU!
Just wondering if ParU supports multiple rhs. Feels OpenMP would make ParU very powerful on solving multi-rhs efficiently.
Beta Was this translation helpful? Give feedback.
All reactions