You are not logged in.
For starters, sorry if theoretical questions belong elsewhere, feel free to move it where it belongs if so
Anyhow, this came to my mind today while Newegg spec'ing my new system. Is it possible to have, in one Linux-only system, an ATi HD6870 (could be any ATI, really) for graphics output (as it's easily one of the, if not THE best GPU in its pricerange) and an nVidia 550Ti (or any other nVidia) for CUDA and/or Folding purposes (As in, no displays connected to it)? I don't see why not, but the main issue I'd see would be drivers. Which, speaking of, don't tell me about how bad fgrlx sucks, I'm quite aware. The reason I'd want the cards in that setup is because the HD6870's an order of magnitude a better performer graphically than the 550Ti, and because ATi GPU Folding is just about a no-go on Linux.
Thanks to anyone who knows, I'm sure it's a pretty trivial question, but eh. That happens
Currently running Arch on a Samsung Chromebook Pro (dual booted with ChromeOS), and various VPSes and Docker containers.
Offline
Although, I suppose it'd be better off to just get a 570 at that point. The price would be about the same and would perform up to par with the 6870, while having CUDA/Folding support. Still curious if it's possible.
Currently running Arch on a Samsung Chromebook Pro (dual booted with ChromeOS), and various VPSes and Docker containers.
Offline
xf86-video-ati asks for libgl, while nvidia seems to require nvidia-utils, which oviously in conflict with libgl.
Offline
in theory it could work if you patch the nvidia driver to become a Cuda-only driver.
Disliking systemd intensely, but not satisfied with alternatives so focusing on taming systemd.
clean chroot building not flexible enough ?
Try clean chroot manager by graysky
Offline