3 ms·
Potentially relevant sidenote: Using VirtualGL (http://www.virtualgl.org/About/Introduction http://www.virtualgl.org/About/Introduction) and TurboVNC (http://w
by i336_ 11y ago
Potentially relevant sidenote:
Using VirtualGL (http://www.virtualgl.org/About/Introduction http://www.virtualgl.org/About/Introduction) and TurboVNC (http://www.turbovnc.org/About/Introduction http://www.turbovnc.org/About/Introduction), you can drop a GPU in a server and serve up OpenGL applications rendered using the GPU in the server via VNC over the network.
VGL pools the GPU(s) between multiple VNC displays, allowing for excellent performance.
- tadfisher 11y agoVirgil (https://virgil3d.github.io/ https://virgil3d.github.io/) is a similar effort for virtualizing GPU(s) in QEMU. The virtio-gpu guest DRM driver just shipped yesterday with Linux 4.4, making it possible to accelerate multiple VMs using host hardware. A Direct3D guest driver, if it performs well enough, could finally end the hacky VFIO passthrough setups that are currently in vogue.
- i336_ 11y agoWow, this looks really cool. One curious idea that might be worth exploring: writing a Vulkan guest driver on top of this stack. It would be a novel approach, as the guest would see a Vulkan-centric view of the world (making development really simple), but that "world" would be easily transportable between machines, regardless of said machines' host graphics stacks.