>We do not currently have any code for using 64 bit integer sizes on
the GPUs.
Thank you, just wanted confirmation.
>Given the current memory available on GPUs is 64 bit integer support
needed? I think even a single vector of length 2^31 will use up most of the
GPU's memory? Are the pract
piaijcusparse.cu:452
[0]PETSC ERROR: #5 MatMult() at
/local/home/rohany/petsc/src/mat/interface/matrix.c:2599
```
Thanks,
Rohan Yadav
will be more expensive than some
> unpreconditioned CG iterations (that don't come close to solving the
> problem -- use multigrid if you want to actually solve this problem).
>
> Rohan Yadav writes:
>
> > Hi,
> >
> > I'm developing a microbenchmark that run
he
common petsc optimizations of preallocating my matrix row counts, so I'm
not sure what's going on with this input generation.
Thanks,
Rohan Yadav
--with-cuda=1 --with-petsc4py && make && make check
>
> Satish
>
> On Thu, 18 Aug 2022, Rohan Yadav wrote:
>
> > Hi,
> >
> > I'm trying to use `petsc4py` on a machine with CUDA GPUs. Is there a way
> to
> > download a version of `pe
Hi,
I'm trying to use `petsc4py` on a machine with CUDA GPUs. Is there a way to
download a version of `petsc4py` from pip or conda with GPU support? If
not, how do I go about installing `petsc4py` with GPU support from source?
I can't find much documentation around doing this.
Thanks,
Rohan
Alright, thanks for the help everyone.
Rohan
On Thu, Feb 3, 2022 at 2:01 PM Barry Smith wrote:
>
>
> On Feb 3, 2022, at 4:28 PM, Rohan Yadav wrote:
>
> To be concrete, the first matrix was
> https://sparse.tamu.edu/LAW/arabic-2005 and the second was
> https://sparse.tamu
bad
> performance (CPU or GPU); it is not a "PDE-ish" type of nonzero structure.
>
>
>
> On Feb 3, 2022, at 2:59 PM, Rohan Yadav wrote:
>
> I'm sorry, I did a little switch here. The original log view I sent for 2
> runs was on a different input matrix. Based
; about 98% on GPU.
>
> Barry
>
>
> On Feb 3, 2022, at 12:29 PM, Rohan Yadav wrote:
>
> >Please send the code that builds the sparse B matrix and the
> > setMatToConstant()
> routine.
>
> Setting to a constant:
> ```
> void setMatToConstant(Mat mat, Pe
---
>>
>> Memory usage is given in bytes:
>>
>> Object Type Creations Destructions Memory Descendants' Mem.
>> Reports information only for process 0.
>>
>> --- Event Stage 0: Main Stage
>>
>> Matrix37 30 2867511840 0.
>> Viewer 2 00 0.
>> Vector 4 1 1792 0.
>>Index Set 2 2 1495248 0.
>>Star Forest Graph 3 00 0.
>>
>> Average time to get PetscTime(): 3.83e-08
>> Average time for MPI_Barrier(): 7.874e-07
>> Average time for zero size MPI_Send(): 3.4035e-06
>> #PETSc Option Table entries:
>> -bench spmm
>> -enable_gpu
>> -log_view
>> -mat_type aijcusparse
>> -matload_block_size 1
>> -matrix /p/gpfs1/yadav2/tensors/petsc/arabic-2005.petsc
>> -n 20
>> -vec_type cuda
>> -warmup 10
>> ```
>>
>>
>> Thanks,
>>
>>
>> Rohan Yadav
>>
>>
>>
>>
>
> --
> Stefano
>
ption Table entries:
-bench spmm
-enable_gpu
-log_view
-mat_type aijcusparse
-matload_block_size 1
-matrix /p/gpfs1/yadav2/tensors/petsc/arabic-2005.petsc
-n 20
-vec_type cuda
-warmup 10
```
Thanks,
Rohan Yadav
I see, thanks all!
Rohan Yadav
> On Dec 14, 2021, at 3:55 PM, Junchao Zhang wrote:
>
>
> MatDenseGetArrayWrite() only returns the pointer to the local array, so it
> does not need any extra memory.
> 2130: PetscErrorCode MatDenseGetArray_SeqDense(Mat A,PetscScal
tDenseRestoreArrayWrite.html#MatDenseRestoreArrayWrite>(B,&x);
>
> --Junchao Zhang
>
>
> On Tue, Dec 14, 2021 at 1:05 PM Rohan Yadav wrote:
>
>> Hi,
>>
>> I'm having trouble setting all entries of a matrix to a constant value,
>> similar to the
Hi,
I'm having trouble setting all entries of a matrix to a constant value,
similar to the `VecSet` method on vectors. I have a dense matrix that I
want to initialize all entries to 1. The only related method I see on the
`Mat` interface is `MatZeroEntries`, which sets all entries to 0. The
obviou
14 matches
Mail list logo