-
Environment:
[runDocker.txt](https://github.com/nicknytko/numml/files/12908070/runDocker.txt)
Docker Image nvcr.io/nvidia/pytorch:23.09-py3
Driver Version: 525.125.06 CUDA Version: 12.0…
-
Using the latest `v0.6.2.0` tag, clSPARSE fails to build on Debian testing / unstable with the following error message:
```
clsparse-0.6.2.0/src/library/include/clSPARSE-private.hpp:31:21: fatal erro…
-
Hello,
I would like to ask if it is possible to get the lower triangular matrix from the factorization, or its parts are distributed among the processes.
Best Regards,
Esmail
-
CHECK_CUSPARSE( cusparseLtMatmulDescriptorInit(
handle.get(), &matmul, opA, opB,
&matA, &matB, &matC, &matC,
…
-
@trilinos/tpetra
@cwpearson
## Bug Report
We observe that the `TpetraCore_FECrs_MatrixMatrix_UnitTests` test hangs in our Cuda build in Debug mode.
We use `Cuda` 12.4.1 compiled with `nvcc`…
-
I'm currently working with matrices where >99% of the values of 0, and it's very unlikely most of the blocks will be larger than 1x1 in size. However, the libsmm_acc code makes me think it could still…
-
## Bug Report
@trilinos/muelu
### Description
The following tests fail in Hip builds of Trilinos with rocm/5.6.1
**MueLu_UnitTestsTpetra_kokkos_MPI_1**
```
...
23:30:32 Smoother (level 1)…
-
@trilinos/muelu @jhux2
## Summary
Tentative factory produces a matrix of special form. This special form can be used to speed up matrix-matrix multiply procedure. For example, when nullspace is …
-
In order to support parallelism on more exciting architectures, we must redesign the unfurl function to avoid carrying closed variables from one dimension to another, so that we can re-label those sym…
-
### What is the expected behavior
- Test pass
### What actually happens
```
/clients/tests/../include/utility.hpp:484: rocsparse_local_spmat::rocsparse_local_spmat(int64_t, int64_t, int64_t, voi…