Yes, I do have superlu_dist built with petsc. The command I used for launching simulation is mpiexec --mca btl self,vader,tcp -np 4 python3 .../main.py ./input_ls -pc_type lu -pc_factor_mat_solver_type superlu_dist -pc_asm_type basic -cuda_device NONE On Mon, Sep 27, 2021 at 6:43 PM Satish Balay <[email protected]> wrote:
Do you have petsc built with superlu_dist?
Satish
On Mon, 27 Sep 2021, Yiyang Li wrote:
Hello,
I have CUDA aware MPI, and I have upgraded from PETSc 3.12 to PETSc 3.15.4 and petsc4py 3.15.4.
Now, when I call
PETSc.KSP().solve(..., ...)
The information of GPU is always printed to stdout by every MPI rank, like
CUDA version: v 11040 CUDA Devices:
0 : Quadro P4000 6 1 Global memory: 8105 mb Shared memory: 48 kb Constant memory: 64 kb Block registers: 65536
CUDA version: v 11040 CUDA Devices:
0 : Quadro P4000 6 1 Global memory: 8105 mb Shared memory: 48 kb Constant memory: 64 kb Block registers: 6553
...
I wonder if there is an option to turn that off? I have tried including
-cuda_device NONE
in command options, but that did not work.
Best regards, Yiyang