[Paraview] Paraview with CUDA on a MPI infiband fabric cluster

[Yasha Karant]

We have been using paraview for a number of years on different platforms
but an EL environment (currently, Scientific Linux 6x). I understand
via Nvidia that Paraview will now support CUDA and GPU offloading for
some parts of the execution, but I have not been able to find an actual
implementation. The standard download is for X86-64 linux, not for this
plus GPU CUDA, and over a cluster of such nodes (using a flavor of MPI).
Is there a CUDA enhanced port of Paraview? If so, how does one obtain it?

If CUDA Paraview is built from source, are the CUDA kernels included
along with the various configuration scripts that CUDA uses for the
Nvidia compilation system?

We are running EL 6 production on X86-64 and have CUDA on the GPU
compute engine nodes.

Yasha Karant
ykarant at csusb.edu

[David E DeMarle]

GPGPU support in VTK/ParaView is pretty bleeding edge today.

Thus far vtkPiston (curtesy of LANL) is about the whole of it. vtkDax
(curtesy of Sandia) is in VTK master now and will be coming to ParaView
soon.

To use vtkPiston, you need to have Thrust installed and LANL's Piston
source code available. Then compile ParaView from source with the Piston
plugin enabled. Doing so exposes CPU<->GPU filters that send data objects
to and from the accelerator, and Piston's slice, contour and threshold
filters which operate on GPU resident vtkPistonDataObjects.

cheers

David E DeMarle
Kitware, Inc.
R&D Engineer
21 Corporate Drive
Clifton Park, NY 12065-8662
Phone: [phone]

Post by Yasha Karant
We have been using paraview for a number of years on different platforms
but an EL environment (currently, Scientific Linux 6x). I understand via
Nvidia that Paraview will now support CUDA and GPU offloading for some
parts of the execution, but I have not been able to find an actual
implementation. The standard download is for X86-64 linux, not for this
plus GPU CUDA, and over a cluster of such nodes (using a flavor of MPI). Is
there a CUDA enhanced port of Paraview? If so, how does one obtain it?
If CUDA Paraview is built from source, are the CUDA kernels included along
with the various configuration scripts that CUDA uses for the Nvidia
compilation system?
We are running EL 6 production on X86-64 and have CUDA on the GPU compute
engine nodes.
Yasha Karant
ykarant at csusb.edu
______________________________**_________________
Powered by www.kitware.com
Visit other Kitware open-source projects at http://www.kitware.com/**
opensource/opensource.html<http://www.kitware.com/opensource/opensource.html>
http://paraview.org/Wiki/**ParaView <http://paraview.org/Wiki/ParaView>
http://www.paraview.org/**mailman/listinfo/paraview<http://www.paraview.org/mailman/listinfo/paraview>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.paraview.org/pipermail/paraview/attachments/20131024/8798c1e3/attachment.htm>