again a priority question please

[miohn]

Hi,

whatever I try with setting priorities and starting a second C4D instance - I still cannot continue
work on the main workstation while rendering something to PV.

So my question:
Is it somehow possible to completely disable GPUs from rendering to PictureViewer
so that only the GPUs from the Slave-Computer are used for the rendering in the PV?
And the (in my example) 2 Titans in the main PC are completely reserved for working
and rendering in LV.

Is this possible?

thanks
Mike

[aggiechase37]

I have mine setup where the monitor is hooked up to the motherboard, which means the computer is going to utilize the hd4600 of the Intel, leaving my titan x to render in the background. This gives me carte blanche to work on other stuff while I have a render going on in the background. Not sure if this is exactly what you're asking but hope it helps.

[miohn]

Hi,

I don't know why its not working here.

In my setup atm I have 2 Titans installed and, to be sure (because I don't know
which one is the one with the monitor attached) I've set both Titans to "Prority Low"
I connected a Slave-Computer (2 GPUs) with GPU-Cluster (4 GPUs) via netrendering. (in total 8 GPUs).
When I start a rendering to PV, I cannot even select a material or scroll in timeline
without waiting seconds for the system to react.

Thanks
Mike

[aybekdes]

Hello,
Sorry if I wasting your topic mionh, but I've also question about priority:
I have 2 GPUs 780Ti -3Gb vram and 980Ti with 6Gb
Screenshot: http://take.ms/QT7lQ
Why octane "chooses or uses" that GPU that has smallest amount of ram?
Is there any workarounds or what should I do?

[miohn]

Hi,

you can do nothing I think.
Octane uses the VRam of the smallest GPU.

Therefore I looked that all my GPUs have 6GB Ram.

regards
Mike

[zoppo]

Hello,
Sorry if I wasting your topic mionh, but I've also question about priority:
I have 2 GPUs 780Ti -3Gb vram and 980Ti with 6Gb
Screenshot: http://take.ms/QT7lQ
Why octane "chooses or uses" that GPU that has smallest amount of ram?
Is there any workarounds or what should I do?

When you have a scene that requires 6GB disable the 780, when you have a scene that only needs 3GB use all cards.

[zoppo]

Hi,
I don't know why its not working here.

Can't you just disable the card with the monitor? Not just use Prio but completely disable it?
Then you should be able to determine very quickly which one is the one with the monitor attached.
BTW what are the GPU statistics showing regarding the load when you use Prio?
You can control if they are correct by using GPU-Z.

And be sure to disable "use all gpu" in the C4D render settings. You can re-enable it if you let your PCs render over night.

[miohn]

Hi zoppo,

thanks for your reply!

GPU stats are showing nearly the same load and temp.
If I change Prio from one GPU to the other I can't feel any difference
on viewport performance.

I came to another solution ( not what I wanted but anyway)
I render the jobs from my slave mashine (as master) and on the WS
I installed slave-daemon with only the 2. GPU active.

This way I have one GPU for working on the WS and the second
GPU is rendering via slave daemon.
This works, but has the hint, that I alwyays have to copy the projects to the slave PC.
I wanted to just start the network render from the master and continue working.
So my initial question was, if it is possible to deactivate a GPU from rendering to PV.
Quasi the other way round.

Hope this makes sense.

thanks
Mike

[zoppo]

Sorry for asking again - are you _sure_ you disabled one of the GPUs in the Octane settings and disabled "Use all GPU" in the C4D render settings?

[miohn]

Hi,

I thought, if I dissable one GPU in settings and then render to PV
only one GPU (the second) is used but for both rendering PV and Vieport.
Not?

regards
Mike