Actually, I am running my graphics applications natively and locally, so performance should be at maximum. Applications are executed via a daemon running on the DC. I am basically building a "graphics cluster" built on ltsp technology. I'll be happy to write up a reference when I'm through.
I think I've started to answer my own question, though. I ran across some notes on the NVidia web site regarding using agpgart and their own built in AGP support. I'll post the solution if I get it working. Thanks, -don On Mon, 7 Jan 2002, David A. Lechner wrote: > Thsi might be hard if I understand the goal and technology right - > AGP 4X has a 1056MB/Sec bandwidth capability - > Fast Ethernet has a 12.5 MB/Sec bandwidth - > In a TC model the server is catching all the graphics card calls and sending > the messages over the network to the clients - whcih then re-create the > graphics locally. But the FE pipe is only about 1% of the AGP4x pipe - So > the performance will, if it works at all, be possibly much less than desired > (or at least less than used to for 3D gaming in a high end machine). > Wr/ Dave Lechner/ > > > > Don Burns wrote: > > > Perhaps this is not the exact forum for this question, but it might be > > relevant to someone who is trying for the same configuration as I. > > > > I am counting on getting maximum 3D graphics performance out of my DC's. > > Currently they use NVidia cards. I've been successful in getting the > > NVidia drivers to work, but not optimally. > > > > glxinfo reports no direct rendering support (BTW, NVidia drivers do not > > utilize DRI): > > > > display: :0.0 screen:0 > > direct rendering: No > > server glx vendor string: NVIDIA Corporation > > server glx version string: 1.2 > > > > etc... > > > > /var/log/XFree86.0.log reports (among many other things) : > > > > (WW) NVIDIA(0): Failed to verify AGP usage > > > > and 'cat /proc/nv/card0' reports : > > > > ----- Driver Info ----- > > NVRM Version: NVIDIA NVdriver Kernel Module 1.0.2314 Fri Nov 30 > > 19:33:20 PST 2001 > > Compiled with: gcc version 2.96 20000731 (Red Hat Linux 7.1 2.96-81) > > ------ Card Info ------ > > Model: GeForce2 MX/MX 400 > > IRQ: 11 > > Video BIOS: 03.11.00.04 > > ------ AGP Info ------- > > AGP status: Disabled > > AGP Driver: > > Bridge: Via Apollo Pro > > SBA: Supported [disabled] > > FW: Unsupported [disabled] > > Rates: 4x 2x 1x [-] > > Registers: 0x1f000207:0x00000000 > > > > I'm loading a kernel that has been built with agpgart compiled in. > > However, if I try an insmod agpgart, I get a "No such device" error. > > > > Anyone else out there already struggled or is struggling with this? > > > > Thanks, > > > > -don > > > > _____________________________________________________________________ > > Ltsp-discuss mailing list. To un-subscribe, or change prefs, goto: > > https://lists.sourceforge.net/lists/listinfo/ltsp-discuss > > For additional LTSP help, try #ltsp channel on irc.openprojects.net > > > _____________________________________________________________________ Ltsp-discuss mailing list. To un-subscribe, or change prefs, goto: https://lists.sourceforge.net/lists/listinfo/ltsp-discuss For additional LTSP help, try #ltsp channel on irc.openprojects.net
