OTOY Forums

Hello all,
I recently finished testing octane on one of our 8 GPU servers and thought I'd share the results with the group. Saw some very fast render times and also noted how well the performance scaled up to 8 GPUs. I've been testing many GPU renderers lately and octane seems to be the only one that takes full advantage of all available GPUs (99-100%)

Here's the demo:
http://www.youtube.com/watch?v=CZ1IRQTq ... r_embedded

-joe

Cool, how fast is the rig with the benchmark scene in pathtracing?

You talk about unbiased raytrace engine but use direct light kernel? Whats up with that.

technically, direct lighting in Octane is unbiased

Its ambient occlusion.
It gives the wrong impression when flipping aroundin a model in AO mode and speak of GI
All im saying is that the speaker says something by do something else.

I see no secondary bounces, but if there is then im cool with that. As in the new mode

yeah, i see no GI either... but, ambient occlusion or not, it's still unbiased (just semantic BS) since it gives no biased errors, and Variance is reduced by 'square root of n' for n samples as per wiki terms... as in I have too much free time...
the rig is ridiculously fast btw, good job

Its misleading for customers. Nothing more. Nothing less. I also posted it in another thread.

All this great markting scene browsing isn't that impressive if they would use PMC and have low samples with 8 GPUs. Plus people get the wrong impression of Octane Render if its explained as a unbiased render scene but only brwosed with direct lighting. What do people think if they see this video and compare it to Maxwell Renders Fire mode which is CPU based?
Mostly I guess "Hey this is unbiased? Looks like crap compared to Maxwell!".

So far I understand, Direct light is an ambiant occlusion mode, so It cannot be unbiased!
Can an Refractive software member team clarify this?

i wish to see too what this rig will do with PMC and DL . . .