Intel® oneAPI Math Kernel Library
Ask questions and share information with other developers who use Intel® Math Kernel Library.
6677 Discussions

How to use sparse LDLT decomposition in MKL?

NEROKAI
Beginner
321 Views

Hi!

I want to use LDLT to decompose a large sparse matrix of mine. I read the manual of MKL and found that there is Bunch-Kaufman decomposition in LAPCAK, which can achieve the same effect as LDLT decomposition, but LAPCK in MKL seems to only support the decomposition of dense matrices .
Is there any way in MKL to decompose a large sparse matrix using LDLT?
Thank you, your help is invaluable to me!

0 Kudos
1 Solution
VidyalathaB_Intel
Moderator
273 Views

Hi Nero,


Thanks for reaching out to us.

Could you please let us know if using "Direct Sparse Solver (DSS) Interface Routines" matches your use case (dss_factor_real, dss_factor_complex - which also uses LDLT factorization)?

For more details please refer to the below link

https://www.intel.com/content/www/us/en/develop/documentation/onemkl-developer-reference-c/top/spars...


Regards,

Vidya.


View solution in original post

3 Replies
VidyalathaB_Intel
Moderator
274 Views

Hi Nero,


Thanks for reaching out to us.

Could you please let us know if using "Direct Sparse Solver (DSS) Interface Routines" matches your use case (dss_factor_real, dss_factor_complex - which also uses LDLT factorization)?

For more details please refer to the below link

https://www.intel.com/content/www/us/en/develop/documentation/onemkl-developer-reference-c/top/spars...


Regards,

Vidya.


NEROKAI
Beginner
254 Views

Thank you, I found this function in DSS, and I found that PARDISO can also be decomposed. May I ask you, is the efficiency of DSS and PARDISO the same?

VidyalathaB_Intel
Moderator
234 Views

Hi Nero,


>>I found this function in DSS, and I found that PARDISO can also be decomposed.is the efficiency of DSS and PARDISO the same?


Speaking in terms of efficiency, the MKL manual says Pardiso is efficient.


"The Intel® oneAPI Math Kernel Library PARDISO package is a high-performance, robust, memory efficient, and easy to use software package for solving large sparse linear systems of equations on shared-memory multiprocessors. "


Reference Link:

https://www.intel.com/content/www/us/en/develop/documentation/onemkl-developer-reference-c/top/spars...

I hope it answers your question.

Thanks for accepting the solution.

As the issue is resolved we are closing this thread. Please post a new question if you need any additional assistance from Intel as this thread will no longer be monitored.


Have a Nice Day!


Regards,

Vidya.


Reply