Nvidia CUDA on MRG Realtime 2.1

Latest response

Not sure if this is the place to post this question but I have been searching to see if Nvidia CUDA is supported on MRG Realtime 2.1  I found a post online in 2008 that said there was no planned support for CUDA in MRG realtime.  I have not been able to find and conclusive information on the RHN site.   I was hoping someone could point me in the correct directions

 

Thanks

Jeff Searle

Responses

The support question is something that nVidia would be better able to answer.  CUDA is an offical product from nVidia

It wouldn't be a bad idea for Red Hat and Nvidia to work together on building RPM packages for CUDA and even having them available via some extra channel or something.

 

Last time I was trying to deploy CUDA on a bunch of machines, there was no RPM available, making the install process tortuous.  The developers didn't help by messing things up and demanding rebuilds of servers.

 

The ability to deploy large grids of CUDA capable servers, however, would be a killer feature for Red Hat - especially in the MRG area.