How to connect to Octane Server?
Posted: Fri May 16, 2014 9:50 am
Hello,
at the moment my set up is fine but I have an unused GTX480 which I want to use for Octane.
My mainboard allows another GPU but at the moment I must use one PCIe 3.0 slot for a display card (FX 580).
I tried to get the internal intel hd to work for several days. Without an nvidia driver the system works fine and stable with ubuntu 14.04 LTS but I can't use Octane.
After installing the nvidia driver sometimes the system hangs and is not usable for production use. I must deactive the internal intel gpu. To avoid problems I decided to use an other nvidia card for display.
Now my idea was to use the system only as renderer by changing the display Quadro FX580 against the unused GTX480. I want to start OctaneServer and connect to this machine from my old system which is good enough for modeling. So my setup should look like this:
OctaneServer
OS Ubuntu 14.04 LTS CPU Intel Core i5-4670 RAM 2x4GB
Display none
Render-GPUs NVIDIA GTX 780/3GB (2x), NVIDIA GTX 480/1.5GB
Driver 331.62 CUDA-Toolkit 6.0
OctaneBlender
OS Ubuntu 14.04 LTS CPU Intel Xeon W3530 8 x 2.8 GHz RAM 2x3GB
Display NVIDIA Quadro FX 580 - 512 MB
Render-GPUs none
Driver 331.62 CUDA-Toolkit 6.0
Before I spend (maybe) hours to get this stack to work I want ask whether this should generally should work?! And what I have to know for this scenario?
I have an standalone and a blender plugin (OctaneBlender) license.
Nice side effect would be that the renderer doesn't heat my feet under my desk and annoying my ears all the time. I could put that thing in the climatised server room.
at the moment my set up is fine but I have an unused GTX480 which I want to use for Octane.
My mainboard allows another GPU but at the moment I must use one PCIe 3.0 slot for a display card (FX 580).
I tried to get the internal intel hd to work for several days. Without an nvidia driver the system works fine and stable with ubuntu 14.04 LTS but I can't use Octane.
After installing the nvidia driver sometimes the system hangs and is not usable for production use. I must deactive the internal intel gpu. To avoid problems I decided to use an other nvidia card for display.
Now my idea was to use the system only as renderer by changing the display Quadro FX580 against the unused GTX480. I want to start OctaneServer and connect to this machine from my old system which is good enough for modeling. So my setup should look like this:
OctaneServer
OS Ubuntu 14.04 LTS CPU Intel Core i5-4670 RAM 2x4GB
Display none
Render-GPUs NVIDIA GTX 780/3GB (2x), NVIDIA GTX 480/1.5GB
Driver 331.62 CUDA-Toolkit 6.0
OctaneBlender
OS Ubuntu 14.04 LTS CPU Intel Xeon W3530 8 x 2.8 GHz RAM 2x3GB
Display NVIDIA Quadro FX 580 - 512 MB
Render-GPUs none
Driver 331.62 CUDA-Toolkit 6.0
Before I spend (maybe) hours to get this stack to work I want ask whether this should generally should work?! And what I have to know for this scenario?
I have an standalone and a blender plugin (OctaneBlender) license.
Nice side effect would be that the renderer doesn't heat my feet under my desk and annoying my ears all the time. I could put that thing in the climatised server room.