Re: Status of MesaGL on headless GPU machines

I think your best shot at Cuda-OpenGL integration will be with 
NVIDIA's own Cuda/OpenGL drivers.  I don't follow Mesa's support for 
NVIDIA GPUs too closely, but I doubt anyone's really addressed Cuda 
interoperability.

-Brian


On 02/22/2013 12:13 AM, Biddiscombe, John A. wrote:
> I have not been following progress of developments closely and am
> still using mesa 7.6.1 on most of my CPU only machines (because it
> works reliably for the plain OpenGL stuff).
>
> I’d like to run some Cuda-OpenGL interop tests on one of our GPU
> enabled machines with Tesla K20X GPUs installed (272 of them).
>
> Is there any version of mesa that I can use on these type of systems
> that allows me to
>
> a)Make use of the hardware
>
> b)Pass arrays from cuda to mesa GL for VBOs etc to avoid CPU/GPU
> exchange of arrays at each render.
>
> In the past the lack of any X installation was a problem with hardware
> access, but with mesa 9.x perhaps this has changed and it’s now
> possible for me to use an accelerated GL on these machines.
>
> I’d welcome any info – I looked through the nouveau wiki but didn’t
> see anything that was encouraging, so I’m guessing I can’t use the
> GPUs for graphics.
>
> Thanks.
>
> JB
>
> --
>
> John Biddiscombe, email:biddisco @.at.@ cscs.ch
>
> http://www.cscs.ch/
>
> CSCS, Swiss National Supercomputing Centre | Tel: +41 (91) 610.82.07
>
> Via Trevano 131, 6900 Lugano, Switzerland | Fax: +41 (91) 610.82.82
>
>
>
> _______________________________________________
> mesa-users mailing list
> [email protected]
> http://lists.freedesktop.org/mailman/listinfo/mesa-users