Dear Paul,<br><br>Thank you very much for your consideration.<br>Actually, if the codes have been finished for complex number-based computation in PETSc, I can do some testings.<br>It will be ok for me. If lots of codes need to be written regarding what I said, I am not sure whether I have enough time. <br>
After all, I am not familiar with PETSc and CUDA.<br><br>Best,<br>Yujie<br><br><div class="gmail_quote">On Mon, Jan 23, 2012 at 5:09 PM, Paul Mullowney <span dir="ltr"><<a href="mailto:paulm@txcorp.com">paulm@txcorp.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><u></u>
<div bgcolor="#ffffff" text="#000000">
Yujie,<br>
<br>
I'm developing GPU code for PETSc that will, at some point, support
complex numbers. However, I cannot give you a time frame as to when
that will be complete as I have several other tasks that are higher
priority right now.<span class="HOEnZb"><font color="#888888"><br>
<br>
-Paul</font></span><div><div class="h5"><br>
<blockquote type="cite">Dear Barry,<br>
<br>
For complex number-based implementation, will the functions in
Vec, Mat, KSP, and PC having been realized for real number work at
least at the running level?<br>
Do I need to do more coding for them?<br>
<br>
To my knowledge, Victor Minden finished most of the coding work
for GPU implementation in PETSc with your and Matt's help.<br>
However, it seems there is not people to further realize and
optimize other GPU functions in PETSc.<br>
Do you think GPU-based computation is not important or has low
efficiency compared to current CPU-based implementation in PETSc?<br>
<br>
<br>
Thank you very much,<br>
Yujie<br>
<br>
<br>
<div class="gmail_quote">On Mon, Jan 23, 2012 at 3:46 PM, Barry
Smith <span dir="ltr"><<a href="mailto:bsmith@mcs.anl.gov" target="_blank">bsmith@mcs.anl.gov</a>></span>
wrote:<br>
<blockquote class="gmail_quote" style="margin:0pt 0pt 0pt 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><br>
Yujie,<br>
<br>
As I said we have never tried it. You are welcome to try it
following the directions I gave. You must realize that we do
not have an army of people to support GPUs so you are largely
on your on in terms of getting things done and need to use
your initiative in trying things yourself and working through
the problems that may arise.<br>
<span><font color="#888888"><br>
Barry<br>
</font></span>
<div>
<div><br>
On Jan 23, 2012, at 3:36 PM, recrusader wrote:<br>
<br>
> Dear Paul,<br>
><br>
> Please find the changelog in the following link for
CUSP 0.2.0<br>
> <a href="http://code.google.com/p/cusp-library/source/browse/CHANGELOG" target="_blank">http://code.google.com/p/cusp-library/source/browse/CHANGELOG</a><br>
><br>
> They has added cusp:complex class to support complex
number-based operation.<br>
><br>
> does it not work for PETSc?<br>
><br>
> Thanks a lot,<br>
> Yujie<br>
><br>
><br>
> On Mon, Jan 23, 2012 at 3:28 PM, Paul Mullowney <<a href="mailto:paulm@txcorp.com" target="_blank">paulm@txcorp.com</a>>
wrote:<br>
> I would claim that Petsc does not support GPU
capabilities for complex numbers right now.<br>
><br>
> The CUSP library is templated over the scalar type
(i.e. float or double), however I don't think it supports
complex numbers. Although, I could be wrong on this.<br>
><br>
> It could be somewhat straightforward to move from
CUSP to CUSPARSE (yes, these are different Nvidia
libraries). I believe CUSPARSE supports SpMV for complex
types. It also supports triangular solve for complex
types.<br>
><br>
> -Paul<br>
><br>
><br>
> This is an installation issue, please send all
installation issues to <a href="mailto:petsc-maint@mcs.anl.gov" target="_blank">petsc-maint@mcs.anl.gov</a>
with the entire configure.log<br>
><br>
><br>
><br>
><br>
> You can try editing config/PETSc/packages/cuda.py
and removing the two lines<br>
><br>
> if self.scalartypes.scalartype == 'complex':<br>
> raise RuntimeError('Must use real numbers with
CUDA')<br>
><br>
> As it says we have never tested for complex so I
do not know how far it is from working.<br>
><br>
> Barry<br>
><br>
> On Jan 23, 2012, at 1:24 PM, recrusader wrote:<br>
><br>
> Dear PETSc Developers,<br>
><br>
> I am compiling PETSc-dev using GPU in complex number
mode.<br>
> However, when I configure PETSc-dev, I get the
following errors<br>
> "<br>
>
*******************************************************************************<br>
> UNABLE to CONFIGURE with GIVEN OPTIONS
(see configure.log<br>
> for details):<br>
>
-------------------------------------------------------------------------------<br>
> Cannot use cuda with complex numbers it is not coded
for this capability<br>
>
*******************************************************************************<br>
> "<br>
> I have checked CUSP. The latest version can support
complex number.<br>
> What should I do for CUDA?<br>
> Thank you very much.<br>
><br>
> Best,<br>
> Yujie<br>
><br>
><br>
<br>
</div>
</div>
</blockquote>
</div>
<br>
</blockquote>
<br>
</div></div></div>
</blockquote></div><br>