Page MenuHome

Cuda device do not display when using NVIDIA 334 drivers
Closed, InvalidPublic


System Information
Operating system and graphics card
Gentoo amd64 on asus g75VW with gtx660m
Blender Version
Broken: all

Short description of error
My blender can't find cuda device
but at the same time i'm already have
i have nvidia-cuda-toolkit-5.5.22, nvidia-drivers-334.21 and blender from git
blender cmake log
All worked fine in winter.
Please help to fix my problem
Exact steps for others to reproduce the error
Based on a (as simple as possible) attached .blend file with minimum amount of steps

Event Timeline

Can you double check your nvidia installation specificity nvidia-uvm and its device files in /dev/.

Martijn Berger (juicyfruit) changed the task status from Unknown Status to Invalid.Apr 6 2014, 1:35 PM
Martijn Berger (juicyfruit) claimed this task.

I just tested master to be 100% sure but I will close the issue now as this is not a blender but an issue with your Cuda install.

On linux it should not matter if you have a kernel compiled or not as blender will first try and load (dlopen) and try to use the api then when requested it will even try and use nvcc to compile the kernel on the fly. The first but however is not working.

It could be that you do not have a in your library path or that newer cuda uvm features are causing breakage for you.
It could also be that you need to update something to make your gentoo system handle this new uvm related stuff or that permissions on the device are wrong.
I would try and download blender 2.69 or 2.70 official and try that. If that does work I will reinvestigate.

Hello, juicyfruit. Thank you for answer
nvidia-uvm is ok
I found a problem - it is nvidia-driver-334x
I downgrade to nvidia-driver-331x and my cuda-5.5 work again!
Sorry for noise, but it was a big problem!)))

Martijn Berger (juicyfruit) renamed this task from Cuda is lost. Please help me to fix. to Cuda device do not display when using NVIDIA 334 drivers.Apr 6 2014, 4:11 PM
Martijn Berger (juicyfruit) lowered the priority of this task from 90 to 50.