On Wed, Apr 25, 2012 at 4:38 PM, David Fuentes wrote:
> thanks! I seem to be getting the same answer on cpu and gpu in 3d w/
> double.
> However, increasing the dof with -refinement_limit 0.01, the residual is
> differing slightly (see below).
>
> in single precision, do you have an idea of the
thanks! I seem to be getting the same answer on cpu and gpu in 3d w/ double.
However, increasing the dof with -refinement_limit 0.01, the residual is
differing slightly (see below).
in single precision, do you have an idea of the upper bound limit on
degrees of freedom
such that the sieve gpu FEM
On Wed, Apr 25, 2012 at 2:06 PM, Matthew Knepley wrote:
> On Wed, Apr 25, 2012 at 1:50 PM, David Fuentes wrote:
>
>> I'm pretty sure I pulled the latest on this petsc-dev and Generator, but
>> when i change GPU type to PetscReal or double, i'm getting key errors ?
>>
>
> Ah well, I was getting ar
On Wed, Apr 25, 2012 at 1:50 PM, David Fuentes wrote:
> I'm pretty sure I pulled the latest on this petsc-dev and Generator, but
> when i change GPU type to PetscReal or double, i'm getting key errors ?
>
Ah well, I was getting around to fixing this. The problem is a little
deeper. The C versio
I'm pretty sure I pulled the latest on this petsc-dev and Generator, but
when i change GPU type to PetscReal or double, i'm getting key errors ?
hg diff
diff -r 4a642f9b7b70 config/PETSc/FEM.py
--- a/config/PETSc/FEM.py Wed Apr 25 02:13:29 2012 -0500
+++ b/config/PETSc/FEM.py Wed Apr 2
On Mon, Apr 23, 2012 at 10:46 AM, David Fuentes wrote:
> I have some existing libMesh code that I would like to use with this ex52
> GPU code example.
>
> I think it may be reasonable to use the existing code for setup of initial
> conditions and jacobian calc,
> But then use the GPU residual cal
I have some existing libMesh code that I would like to use with this ex52
GPU code example.
I think it may be reasonable to use the existing code for setup of initial
conditions and jacobian calc,
But then use the GPU residual calculation.
Where is the nodal degree of freedom stored in the "DM_Co
thank you Matt,
David
On Mon, Apr 2, 2012 at 9:37 PM, Matthew Knepley wrote:
> On Wed, Mar 28, 2012 at 6:42 PM, Matthew Knepley wrote:
>
>> On Wed, Mar 28, 2012 at 6:25 PM, David Fuentes wrote:
>>
>>> Yes. This would work.
>>> I had trouble compiling in single precision using the some of the
>>>
On Wed, Mar 28, 2012 at 6:42 PM, Matthew Knepley wrote:
> On Wed, Mar 28, 2012 at 6:25 PM, David Fuentes wrote:
>
>> Yes. This would work.
>> I had trouble compiling in single precision using the some of the
>> external package options I was using for double.
>>
>
> Okay, getting on it now.
>
O
On Wed, Mar 28, 2012 at 6:25 PM, David Fuentes wrote:
> Yes. This would work.
> I had trouble compiling in single precision using the some of the external
> package options I was using for double.
>
Okay, getting on it now.
Matt
> On Wed, Mar 28, 2012 at 4:57 PM, Matthew Knepley wrote:
>
Yes. This would work.
I had trouble compiling in single precision using the some of the external
package options I was using for double.
On Wed, Mar 28, 2012 at 4:57 PM, Matthew Knepley wrote:
> On Wed, Mar 28, 2012 at 4:12 PM, David Fuentes wrote:
>
>> works!
>>
>
> Excellent. Now, my thinkin
On Wed, Mar 28, 2012 at 4:12 PM, David Fuentes wrote:
> works!
>
Excellent. Now, my thinking was that GPUs are most useful doing single
work, but
I can see the utility of double accuracy for a residual.
My inclination is to define another type, say GPUReal, and use it for all
kernels.
Would tha
works!
SCRGP2$ make ex52
/usr/bin/mpicxx -o ex52.o -c -O0 -g -fPIC
-I/opt/apps/PETSC/petsc-dev/include
-I/opt/apps/PETSC/petsc-dev/gcc-4.4.3-mpich2-1.2-epd-sm_20-single-dbg/include
-I/opt/apps/cuda/4.0/cuda/include -I/usr/include -I/usr/include/mpich2
-D__INSDIR__=src/snes/examples/tutorials/ ex52
sure. will do.
On Wed, Mar 28, 2012 at 1:23 PM, Matthew Knepley wrote:
> On Wed, Mar 28, 2012 at 1:14 PM, David Fuentes wrote:
>
>> thanks! its running, but I seem to be getting different answer for
>> cpu/gpu ?
>> i had some floating point problems on this Tesla M2070 gpu before, but
>> adding
On Wed, Mar 28, 2012 at 1:14 PM, David Fuentes wrote:
> thanks! its running, but I seem to be getting different answer for cpu/gpu
> ?
> i had some floating point problems on this Tesla M2070 gpu before, but
> adding the '-arch=sm_20' option seemed to fix it last time.
>
>
> is the assembly in si
thanks! its running, but I seem to be getting different answer for cpu/gpu ?
i had some floating point problems on this Tesla M2070 gpu before, but
adding the '-arch=sm_20' option seemed to fix it last time.
is the assembly in single precision ? my 'const PetscReal jacobianInverse'
being passed i
On Wed, Mar 28, 2012 at 11:45 AM, David Fuentes wrote:
> The example seems to be running on cpu with '-batch' but i'm getting
> errors in line 323 with the '-gpu' option
>
> [0]PETSC ERROR: IntegrateElementBatchGPU() line 323 in
> src/snes/examples/tutorials/ex52_integrateElement.cu
>
> should th
The example seems to be running on cpu with '-batch' but i'm getting errors
in line 323 with the '-gpu' option
[0]PETSC ERROR: IntegrateElementBatchGPU() line 323 in
src/snes/examples/tutorials/ex52_integrateElement.cu
should this possibly be PetscScalar ?
- ierr = cudaMalloc((void**) &d_coeffi
On Tue, Mar 27, 2012 at 2:10 PM, Blaise Bourdin wrote:
>
> On Mar 27, 2012, at 1:23 PM, Matthew Knepley wrote:
>
> On Tue, Mar 27, 2012 at 12:58 PM, David Fuentes wrote:
>
>> Hi,
>>
>> I had a question about the status of example 52.
>>
>>
>> http://petsc.cs.iit.edu/petsc/petsc-dev/file/a8e2f2c19
On Tue, Mar 27, 2012 at 7:18 PM, David Fuentes wrote:
> thanks! when generating the header file ex52.h
>
> what should the input arguments be to the PetscGenerateFEMQuadrature.py
> file ?
>
Andy Terrel already busted me about this today. I pushed a new tarball to
the ftp site, so this will work
thanks! when generating the header file ex52.h
what should the input arguments be to the PetscGenerateFEMQuadrature.py
file ?
SCRGP2$ $PETSC_DIR/bin/pythonscripts/PetscGenerateFEMQuadrature.py 3 1 8 1
laplacian ex52.h
[{(-1.0, -1.0, -1.0): [(1.0, ())]}, {(1.0, -1.0, -1.0): [(1.0, ())]},
{(-1.0,
On Mar 27, 2012, at 1:23 PM, Matthew Knepley wrote:
> On Tue, Mar 27, 2012 at 12:58 PM, David Fuentes
> wrote:
> Hi,
>
> I had a question about the status of example 52.
>
> http://petsc.cs.iit.edu/petsc/petsc-dev/file/a8e2f2c19319/src/snes/examples/tutorials/ex52.c
> http://petsc.cs.iit.edu
On Tue, Mar 27, 2012 at 12:58 PM, David Fuentes wrote:
> Hi,
>
> I had a question about the status of example 52.
>
>
> http://petsc.cs.iit.edu/petsc/petsc-dev/file/a8e2f2c19319/src/snes/examples/tutorials/ex52.c
>
> http://petsc.cs.iit.edu/petsc/petsc-dev/file/a8e2f2c19319/src/snes/examples/tuto
Hi,
I had a question about the status of example 52.
http://petsc.cs.iit.edu/petsc/petsc-dev/file/a8e2f2c19319/src/snes/examples/tutorials/ex52.c
http://petsc.cs.iit.edu/petsc/petsc-dev/file/a8e2f2c19319/src/snes/examples/tutorials/ex52_integrateElement.cu
Can this example be used with a DM obj
24 matches
Mail list logo