2

My goal is to invert a large dense matrix in a distributed environment.

I would like a library which would be "easily" integrated in a C++ application, thus it would be ideal that the library has a C/C++ interface or/and implementation.

A good candidate was ScalaPACK, but what I need is in Fortran (awaiting confirmation from the developers).

Another good candidate was SuperLU, but it is for sparse matrices.


PS - "distributed-computing" and "linear-algebra" would be nice to exist as tags, but I do not have the rep to create them.

2 Answers 2

2

The main contender for replacing ScaLAPACK seems to be SLATE (repository, manual). It targets distributed systems (including accelerators) for dense linear algebra. It is implemented in C++ and provides additional C and Fortran APIs.

1

So far I got

enter image description here enter image description here

I would happy to know that there is more.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.