[Paraview] [Non-DoD Source] Re: ParaView GPU support (UNCLASSIFIED)

[Robert Maynard]

Hi Simon,


The ability to enable CUDA support in the ParaViewSuperbuild has been
added. You will need to have the latest superbuild, and enable
`ENABLE_cuda` for support to be added.

On Thu, Sep 21, 2017 at 1:08 PM, Su, Simon M CIV USARMY RDECOM ARL
(US) <simon.m.su.civ at mail.mil> wrote:
> CLASSIFICATION: UNCLASSIFIED
>
> Hello Robert,
>
> thank you for your reply and we are looking forward to your email. Joe has
> made some progress on that front on one of our newly installed system. We are
> seeing some 12x performance gain on our compute nodes with GPU hardware. We
> are looking into system wide deployment and having to be able to do it from
> superbuild will significantly help simplify our deployment process.
>
> thanks
> -simon
>
> -----Original Message-----
> From: Robert Maynard [mailto:robert.maynard at kitware.com]
> Sent: Thursday, September 21, 2017 12:07 PM
> To: Su, Simon M CIV USARMY RDECOM ARL (US) <simon.m.su.civ at mail.mil>
> Cc: Utkarsh Ayachit <utkarsh.ayachit at kitware.com>; paraview at paraview.org
> Subject: Re: [Paraview] [Non-DoD Source] Re: ParaView GPU support
> (UNCLASSIFIED)
>
> All active links contained in this email were disabled.  Please verify the
> identity of the sender, and confirm the authenticity of all links contained
> within the message prior to copying and pasting the address to a Web browser.
>
>
>
>
> ----
>
> Hi Simon,
>
> Currently the VTK-m support that is offered by the ParaView superbuild doesn't
> support enabling CUDA. I am in the process of adding this option to the
> superbuild and can notify you when this feature has been merged in. In the
> short-term you would want to manual modify the vtk-m CMAKE_ARGS entry in the
> superbuild to have VTKm_ENABLE_CUDA:BOOL=ON.
>
> On Wed, Sep 20, 2017 at 9:45 AM, Su, Simon M CIV USARMY RDECOM ARL
> (US) <simon.m.su.civ at mail.mil> wrote:
>> CLASSIFICATION: UNCLASSIFIED
>>
>> Hi Utkarsh,
>>
>> is there a sure way to tell if our ParaView VTK-m compilation has the
>> correct CUDA support?  For the superbuild process, is there a way to
>> force the detection of CUDA library and fail the superbuild process if
>> CUDA library is not correctly detected? thank you
>>
>> thanks
>> -simon
>>
>> -----Original Message-----
>> From: Utkarsh Ayachit [Caution-mailto:utkarsh.ayachit at kitware.com]
>> Sent: Friday, September 15, 2017 12:36 PM
>> To: Su, Simon M CIV USARMY RDECOM ARL (US) <simon.m.su.civ at mail.mil>
>> Cc: paraview at paraview.org
>> Subject: [Non-DoD Source] Re: [Paraview] ParaView GPU support
>> (UNCLASSIFIED)
>>
>> All active links contained in this email were disabled.  Please verify
>> the identity of the sender, and confirm the authenticity of all links
>> contained within the message prior to copying and pasting the address to a
>> Web browser.
>>
>>
>>
>>
>> ----
>>
>> Simon,
>>
>> The answer depends on whether you want to ParaView to use GPU for
>> rendering alone or also for computation e.g. contours. For former,
>> it's automatic via OpenGL. So long as your environment and/or your
>> build is set up such that ParaView picks up the OpenGL implementation
>> provided by your GPU drivers, you'll get GPU acceleration. For
>> GPU-accelerated algorithms, you have to enable VTK-m support at build time
>> and then enable CUDA support for VTM-m.
>> Currently, you need create special filters via the VTM-m plugin [1] to
>> use accelerated versions of common algorithms like contour. In future,
>> this will be automatic, and you will be able to simply use the generic
>> filters.
>>
>> Hope that helps.
>>
>> Utkarsh
>>
>>
>> [1]
>> Caution-Caution-https://blog.kitware.com/paraview-5-3-0-release-notes/
>>
>> On Thu, Sep 14, 2017 at 3:42 PM, Su, Simon M CIV USARMY RDECOM ARL
>> (US) <simon.m.su.civ at mail.mil> wrote:
>>> CLASSIFICATION: UNCLASSIFIED
>>>
>>>  Hello,
>>>
>>> How do we configure ParaView to take advantage of available GPU
>>> hardware? Do we have to compile with different compile flags? Or
>>> ParaView automatically detects availability of GPU hardware and take
>>> advantage of it?  Any insight is greatly appreciated
>>>
>>> Thanks
>>> -simon
>>>
>>>
>>> CLASSIFICATION: UNCLASSIFIED
>>> _______________________________________________
>>> Powered by Caution-Caution-www.kitware.com
>>>
>>> Visit other Kitware open-source projects at
>>> Caution-Caution-http://www.kitware.com/opensource/opensource.html
>>>
>>> Please keep messages on-topic and check the ParaView Wiki at:
>>> Caution-Caution-http://paraview.org/Wiki/ParaView
>>>
>>> Search the list archives at:
>>> Caution-Caution-http://markmail.org/search/?q=ParaView
>>>
>>> Follow this link to subscribe/unsubscribe:
>>> Caution-Caution-http://public.kitware.com/mailman/listinfo/paraview
>>
>>
>> CLASSIFICATION: UNCLASSIFIED
>>
>> _______________________________________________
>> Powered by Caution-www.kitware.com
>>
>> Visit other Kitware open-source projects at
>> Caution-http://www.kitware.com/opensource/opensource.html
>>
>> Please keep messages on-topic and check the ParaView Wiki at:
>> Caution-http://paraview.org/Wiki/ParaView
>>
>> Search the list archives at:
>> Caution-http://markmail.org/search/?q=ParaView
>>
>> Follow this link to subscribe/unsubscribe:
>> Caution-http://public.kitware.com/mailman/listinfo/paraview
>>
>
>
> CLASSIFICATION: UNCLASSIFIED