Groups

SLATE User

1–30 of 131

Welcome to the SLATE User Forum.

This forum is for general discussion of the SLATE library and its associated BLAS++ and LAPACK++ libraries.

http://icl.utk.edu/slate

For bug reports, please file an issue on Bitbucket. This helps us track the status.

https://github.com/icl-utk-edu/slate/issues/

https://github.com/icl-utk-edu/blaspp/issues/

https://github.com/icl-utk-edu/lapackpp/issues/

Members can post messages online or by emailing slate-user@icl.utk.edu.

See the My Settings menu, above right, to adjust Membership and email settings, including receiving no email updates, combined updates, or every new email message.

0 selected

Kevin Ingles, Mark Gates2

Jun 21

QR Decompositions

LAPACK, hence LAPACK++, is column-major. In most cases, the only way we could support row-major is by

unread,

QR Decompositions

LAPACK, hence LAPACK++, is column-major. In most cases, the only way we could support row-major is by

Jun 21

seon yeung he, Mark Gates2

Mar 25

slate::lu_solve GPU Issue and CPU Deadlock with p=1, q=2

The LU solver should work for an arbitrary MPI grid. One problem in your code is the MPI_Init. You

unread,

slate::lu_solve GPU Issue and CPU Deadlock with p=1, q=2

The LU solver should work for an arbitrary MPI grid. One problem in your code is the MPI_Init. You

Mar 25

Nakul Iyer, Paul Lin7

Feb 28

Projects builds and runs but not with cuda aware MPI

This has been working for us. Thank you very much for your help! On Thursday, February 13, 2025 at 12

unread,

Projects builds and runs but not with cuda aware MPI

This has been working for us. Thank you very much for your help! On Thursday, February 13, 2025 at 12

Feb 28

Aaron Altman, Mark Gates10

Jan 16

scalapack_api not using GPUs

On Sun, Dec 22, 2024 at 4:18 PM Mark Gates <mga...@icl.utk.edu> wrote: The implementation

unread,

scalapack_api not using GPUs

On Sun, Dec 22, 2024 at 4:18 PM Mark Gates <mga...@icl.utk.edu> wrote: The implementation

Jan 16

12/6/24

Issue with Multi-process Calculation Correctness for the heev Interface

Hi,Mark, When testing the heev interface using the test program, with 4 processes, dim=1000, and nb=

unread,

Issue with Multi-process Calculation Correctness for the heev Interface

Hi,Mark, When testing the heev interface using the test program, with 4 processes, dim=1000, and nb=

12/6/24

韩梅, Mark Gates4

11/27/24

he2hb Multi-Process Crash Issue

Sorry, we've been away at a conference and haven't had time to investigate. We'll keep

unread,

he2hb Multi-Process Crash Issue

Sorry, we've been away at a conference and haven't had time to investigate. We'll keep

11/27/24

bcsj, Mark Gates3

9/29/24

tileTick deprecation

Thank you very much for the reply. It helped a lot. - bcsj On Tuesday, September 24, 2024 at 8:44:35

unread,

tileTick deprecation

Thank you very much for the reply. It helped a lot. - bcsj On Tuesday, September 24, 2024 at 8:44:35

9/29/24

Robert Langefeld, Mark Gates3

7/18/24

Multinode Eigendecomposition

As a follow up, 1) Is there a timeline for that kind of implementation? Even ballpark like within a

unread,

Multinode Eigendecomposition

As a follow up, 1) Is there a timeline for that kind of implementation? Even ballpark like within a

7/18/24

Robert Langefeld, Mark Gates6

7/16/24

Multiply gives different answers for different block sizes

Thus far, the two issues I ran into were reading things into memory and writing from memory. I spent

unread,

Multiply gives different answers for different block sizes

Thus far, the two issues I ran into were reading things into memory and writing from memory. I spent

7/16/24

bcsj, Mark Gates10

6/21/24

Assertion `array_dev != nullptr' failed. in tileLayoutConvert

Hey, thank you for the reply! :) Good to know that row-major is perhaps something to keep an eye on

unread,

Assertion `array_dev != nullptr' failed. in tileLayoutConvert

Hey, thank you for the reply! :) Good to know that row-major is perhaps something to keep an eye on

6/21/24

5/31/24

SLATE, BLAS++, LAPACK++ v2024.05.31 releases

SLATE, BLAS++, LAPACK++ have v2024.05.31 releases. See GitHub for release notes: https://github.com/

unread,

SLATE, BLAS++, LAPACK++ v2024.05.31 releases

SLATE, BLAS++, LAPACK++ have v2024.05.31 releases. See GitHub for release notes: https://github.com/

5/31/24

bcsj, Neil Lindquist4

5/16/24

GPU-to-GPU communication

Okay, that makes sense. I should be using the 2023.11.05-release version, so I guess I just missed

unread,

GPU-to-GPU communication

Okay, that makes sense. I should be using the 2023.11.05-release version, so I guess I just missed

5/16/24

Andrew Cunningham, Mark Gates3

4/30/24

Trying to build on Mac OS 14.4.1

Worked perfectly, thank you! On Tuesday, April 30, 2024 at 6:19:11 AM UTC-7 mga...@icl.utk.edu wrote:

unread,

Trying to build on Mac OS 14.4.1

Worked perfectly, thank you! On Tuesday, April 30, 2024 at 6:19:11 AM UTC-7 mga...@icl.utk.edu wrote:

4/30/24

Runfeng Jin, … Mark Gates5

3/20/24

Performance about eigenvalue solve

The DPLASMA eigensolver is not GPU accelerated. I inquired with the DPLASMA developers, who said GPU

unread,

Performance about eigenvalue solve

The DPLASMA eigensolver is not GPU accelerated. I inquired with the DPLASMA developers, who said GPU

3/20/24

Robert Knop, Mark Gates3

3/1/24

Is it possible to access matrix elemnts within Cuda?

Great, thanks. I'll have to look at the examples, and think about how to actually parallelize. In

unread,

Is it possible to access matrix elemnts within Cuda?

Great, thanks. I'll have to look at the examples, and think about how to actually parallelize. In

3/1/24

bcsj, Mark Gates7

2/9/24

gels precission?

In our tester, we keep copies of A and B to compute residuals. I will have to see whether we can

unread,

gels precission?

In our tester, we keep copies of A and B to compute residuals. I will have to see whether we can

2/9/24

Miguel Morales, Mark Gates2

12/12/23

lapack return values/error messages

Hi Miguel, Now SLATE LU returns info for exactly singular matrices, and Cholesky returns info for

unread,

lapack return values/error messages

Hi Miguel, Now SLATE LU returns info for exactly singular matrices, and Cholesky returns info for

12/12/23

12/12/23

Re: [SLATE user] Segmentation fault in slate::gesv when using CUDA-aware MPI

Thanks for the bug report. I don't have an immediate fix, it will need some investigation. We

unread,

Re: [SLATE user] Segmentation fault in slate::gesv when using CUDA-aware MPI

Thanks for the bug report. I don't have an immediate fix, it will need some investigation. We

12/12/23

andrea chiariello, Mark Gates2

10/25/23

Slate performance on Generalized eigenvalues decomposition

Hi Andrea, The functionality for a generalized eigensolver, either with or without eigenvectors,

unread,

Slate performance on Generalized eigenvalues decomposition

Hi Andrea, The functionality for a generalized eigensolver, either with or without eigenvectors,

10/25/23

bcsj, Neil Lindquist11

9/29/23

Segmentation fault in slate::gels

Seems there is a larger reservation on LUMI until Tuesday next week on the partition, so my trying it

unread,

Segmentation fault in slate::gels

Seems there is a larger reservation on LUMI until Tuesday next week on the partition, so my trying it

9/29/23

bcsj, Mark Gates3

8/31/23

Bitbucket documentation/function reference gone?

Hi Mark, Thank you for replying. I was trying to access: https://icl.bitbucket.io/slate/index.html It

unread,

Bitbucket documentation/function reference gone?

Hi Mark, Thank you for replying. I was trying to access: https://icl.bitbucket.io/slate/index.html It

8/31/23

Robert Knop, Mark Gates3

8/22/23

Guidance on block size, mpi ranks per node, etc.

OK, thanks. I'll take a look at those wiki pages. What I'm doing is a cholesky decomposition,

unread,

Guidance on block size, mpi ranks per node, etc.

OK, thanks. I'll take a look at those wiki pages. What I'm doing is a cholesky decomposition,

8/22/23

Robert Knop, Mark Gates3

8/1/23

Basic usage : tile_node->existsOn(device) failed ; what am I doing wrong?

On Tue, Aug 1, 2023 at 12:25 PM Robert Knop <rak...@lbl.gov> wrote: OK, I figured out my

unread,

Basic usage : tile_node->existsOn(device) failed ; what am I doing wrong?

On Tue, Aug 1, 2023 at 12:25 PM Robert Knop <rak...@lbl.gov> wrote: OK, I figured out my

8/1/23

bcsj, Mark Gates6

7/31/23

Merging two distributed matrices?

Great! I wrote the following function which reads a text file (comma delimited by default) into a

unread,

Merging two distributed matrices?

Great! I wrote the following function which reads a text file (comma delimited by default) into a

7/31/23

Baptiste Legouix, Mark Gates4

7/13/23

fromLAPACK band matrix

from(Sca)LAPACK create a SLATE matrix with tiles that point to existing (Sca)LAPACK data. The data

unread,

fromLAPACK band matrix

from(Sca)LAPACK create a SLATE matrix with tiles that point to existing (Sca)LAPACK data. The data

7/13/23

Arturo Ges, Asim YarKhan3

7/12/23

How to use the Lapack compatibility interface in C language

Hi Asim, The information you provided has been very useful. We started compiling the example with the

unread,

How to use the Lapack compatibility interface in C language

Hi Asim, The information you provided has been very useful. We started compiling the example with the

7/12/23

bcsj, … Mark Gates8

6/15/23

Distance matrix computation

On Thu, Jun 15, 2023 at 4:38 AM bcsj <bjornje...@gmail.com> wrote: Is there a method for

unread,

Distance matrix computation

On Thu, Jun 15, 2023 at 4:38 AM bcsj <bjornje...@gmail.com> wrote: Is there a method for

6/15/23

Ilkhom Abdurakhmanov, Mark Gates7

6/13/23

Testing slate performance on AMD MI 250 GPUs

On Tue, Jun 13, 2023 at 12:54 AM Ilkhom Abdurakhmanov <i.abdur...@gmail.com> wrote: Thank

unread,

Testing slate performance on AMD MI 250 GPUs

On Tue, Jun 13, 2023 at 12:54 AM Ilkhom Abdurakhmanov <i.abdur...@gmail.com> wrote: Thank

6/13/23

bcsj, Mark Gates5

6/8/23

Out of memory error in call to tester

On Thursday, June 8, 2023 at 10:06:54 PM UTC+3 mga...@icl.utk.edu wrote: On Thu, Jun 8, 2023 at 12:45

unread,

Out of memory error in call to tester

On Thursday, June 8, 2023 at 10:06:54 PM UTC+3 mga...@icl.utk.edu wrote: On Thu, Jun 8, 2023 at 12:45

6/8/23

bcsj, Mark Gates3

6/1/23

FATAL ERROR: CCE OpenMP does not yet support GNU entry point "GOMP_taskgroup_start"

Thanks for replying :) I've thought some on the issue since I posted the first message and I'

unread,

FATAL ERROR: CCE OpenMP does not yet support GNU entry point "GOMP_taskgroup_start"

Thanks for replying :) I've thought some on the issue since I posted the first message and I'

6/1/23

Search

Clear search

Close search

Google apps

Main menu