On 8/22/13 10:04 AM, John Clyne wrote: > The Quadro 2000, and I think the 550, are based on the older Fermi > architecture. I wanted to get confirmation that ViritualGL was supporting the > newer Kepler based products. I would think Kepler-Quadro would be supported, > but the Tesla cards target computation, not graphics. And the nVidia Grid is > a brand new product that apparently has some form of integrated remote viz > capabilities.
Yeah, you're referring to VGX. I am keenly interested in knowing more about it, but specific information is hard to find. It seems to require Citrix XenServer, and it is basically providing a mechanism for either assigning a full GPU or a "vGPU" (portion of a shared GPU) to a particular virtual machine. I heard a rumor through one of my customers that Linux guests aren't currently supported-- only Windows, but if anyone knows more about it, please share. It unfortunately isn't anything that we could take advantage of with VirtualGL at the moment, but I really wish nVidia, ATI, etc. would play nice and produce an EGL interface for accessing their cards' rendering capabilities without an X server. It would make my job a hell of a lot easier. Wayland may necessitate them doing that. Anyhow, you guys know more about supporting VGL on headless devices than I do. I can only afford a Quadro 600 to play with. ------------------------------------------------------------------------------ Introducing Performance Central, a new site from SourceForge and AppDynamics. Performance Central is your source for news, insights, analysis and resources for efficient Application Performance Management. Visit us today! http://pubads.g.doubleclick.net/gampad/clk?id=48897511&iu=/4140/ostg.clktrk _______________________________________________ VirtualGL-Users mailing list VirtualGL-Users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/virtualgl-users
