Different rendertime on same computer????

Hello
Can you describe me why this happaned?
I have a notebook with gtx660m
And when i run blenchmark addton (or another not matter whitch scene) i get this time
4.15
http://wstaw.org/m/2015/11/08/plasma-desktopJl3983.png
I have another machine - pc with two gtx780
I linked blenders config from pc to laptop via dropbox
and now have a pc config for laptop machine.
Now i try blenchmark agane aaaand…
3.24 !!!
http://wstaw.org/m/2015/11/08/plasma-desktopXq3983.png

Why???

Seems like you failed to actually pick a compute device for the PC, but its CPU is still a bit faster than the notebook’s GPU.

This is only laptop tests

Well, all I know is there’s no compute device selected in the second pic.

No. It is! It mean the 2xgtx780, and give me 8 render buckets (when i select my gtx660m - there is 1 render bucket!)
http://wstaw.org/m/2015/11/08/plasma-desktopag3983.png
http://wstaw.org/m/2015/11/08/plasma-desktopUj3983.png
http://wstaw.org/m/2015/11/08/plasma-desktopqR3983.png

If you’re seeing 8 render threads, you’re NOT rendering with the GPU, but the CPU.
CPU rendering = 1 render thread per CPU thread.
GPU rendering = 1 render thread per device (graphics card).

I guess this is what’s happening here: The config from your desktop tells Blender to render on the two GTX 780s. But your laptop doesn’t have matching cards (obviously), so Blender can’t find the specified render device. Therefore it shows an empty CUDA device field and falls back to CPU rendering.

Blender is most likely rendering faster than before in this case because your ageing notebook-class GPU is no match for your more modern notebook CPU (i7?). What is your tile size? With small tile sizes (e. g. 32 x 32) your CPU should perform even better. You may find that you want to abandon GPU rendering on that particular notebook completely.

BTW, are you sure this really is a topic for “Blender and CG Discussions”? “Technical Support” seems to be a more suitable place, no?

Moved from “General Forms > Blender and CG Discussions” to “Support > Technical Support”