Hi everyone,
I am experiencing some unexpected poor performance when using hybrid rendering with multiple GPU's and trying to determine if this is driver/version related or not.
First Scenario:
-V-ray Next 4.04.03 for Maya 2018
-Production Engine set to CUDA, Full size textures
-Progressive Image Sampler, 50 max subdivs, 0.05 Noise Threshold, default Light Cache settings enabled
-Titan Xp
-GTX 960
-Nvidia Driver 419.67 Game Ready Driver (Or creative driver with the Titan only)
I have a scene with a high poly corvette. When Rendering with CPU+Titan Xp, my render time is about 1 minute. When I add the GTX 960 to the "V-ray render devices selection" my render time jumps up to about 4 minutes, why is this?
Second Scenario:
-Updated to V-ray 4.12.01 for maya 2018
-Same exact render settings as before
-Updated to Nvidia Driver 430.64 Game Ready Driver
-Same GPU's as before
Now when I render CPU+Titan Xp the render time is 3 minutes as opposed to 1 minute on older versions!!! How can this be? Should it not be faster or at least roughly the same? Adding the GTX 960 makes it even slower again and adjusting the rays per pixel/bundle size doesn't help any.
So I would like to know what is the best CUDA Driver to be using with the latest version of Vray Next for Maya?
Is it unusual to see slower render times with different GPU's as such in my case? I would like to use both GPU if there is any benefit to doing so, otherwise I'm rolling back to creative drivers on the Titan Xp and scrapping my 960.
Thank you all for any help and suggestions!
-Harrison
(p.s. I am a customer of v-ray and phoenixFD but I cannot post in the other forums, should I email support about this? thanks again)
I am experiencing some unexpected poor performance when using hybrid rendering with multiple GPU's and trying to determine if this is driver/version related or not.
First Scenario:
-V-ray Next 4.04.03 for Maya 2018
-Production Engine set to CUDA, Full size textures
-Progressive Image Sampler, 50 max subdivs, 0.05 Noise Threshold, default Light Cache settings enabled
-Titan Xp
-GTX 960
-Nvidia Driver 419.67 Game Ready Driver (Or creative driver with the Titan only)
I have a scene with a high poly corvette. When Rendering with CPU+Titan Xp, my render time is about 1 minute. When I add the GTX 960 to the "V-ray render devices selection" my render time jumps up to about 4 minutes, why is this?
Second Scenario:
-Updated to V-ray 4.12.01 for maya 2018
-Same exact render settings as before
-Updated to Nvidia Driver 430.64 Game Ready Driver
-Same GPU's as before
Now when I render CPU+Titan Xp the render time is 3 minutes as opposed to 1 minute on older versions!!! How can this be? Should it not be faster or at least roughly the same? Adding the GTX 960 makes it even slower again and adjusting the rays per pixel/bundle size doesn't help any.
So I would like to know what is the best CUDA Driver to be using with the latest version of Vray Next for Maya?
Is it unusual to see slower render times with different GPU's as such in my case? I would like to use both GPU if there is any benefit to doing so, otherwise I'm rolling back to creative drivers on the Titan Xp and scrapping my 960.
Thank you all for any help and suggestions!
-Harrison
(p.s. I am a customer of v-ray and phoenixFD but I cannot post in the other forums, should I email support about this? thanks again)
Comment