Dear Vray user,
We just purchase Vray GPU pre order, and test it today.
First of all, our choice to buy vray have been taken after testing Octane render. We test it on a 3 years old laptop with a quadro FX 1800 and the result was so fast that we have been looking around for other competitors.
After looking around for several week,on CG forum, vray video demo of GPU… We finally decided to go for Vray as it seems to be the standard for Interior architecture due to its fast rendering biais engine.
So we though, Whao!!, if a brand new beta engine like octane can make it so fast on a old laptop with a entry range video card, Vray , as a major player in the rendering engine with it’s new GPU engine will rock.
So we bought a FX5800, installed it in a brand new Dual X5650 computer, with Win7 64bit… everything to be as much as possible on the top line within , of course, a budget in mind.
Now let’s go back to reality : we just setup a basic scene, just one skylight, overide material everywhere with exclusion of the windows and door to let the light from outside to step in and the result is: incredibly slow.
At least a lot slower than a render with the RT cpu and I guess, from the little time we spend on octane, 10 time slower than Octane in beta stage.
So I really wonder, how other Vray user on this forum, enjoy so much this Beta!
We joined the scene, in case somebody got the time to test it on a similar graphic card, just to sort out what’s the problem, if there is one of course. 3dtestVrayGPU.zip (51.4 KB)
Iam really disapointed too about RT GPU. I have GTX 295, two cores and it slow. Its sometimes maybe 3x faster then CPU, but I have Core2Quad at 3GHz, Core i7 what I have at work is 2x faster then my CPU at home, so It will be almost no difference at work.
When I have time, I will try Your scene.
Here are the screen shoot of the Vray RT setup.
Unfortunatly, GPU is not faster than CPU as seen on here, at least not on this basic scene and not with this version of Vray GPU.
It will be a good idea to put the true benchmark on line instead of marketing video spread around the web.
What I don’t like about GPU is the lag I get. But, wireframe views in 3dsmax is incredibly slow compared to XSI or Maya (or anything else), which bothers me a lot more than the lag in vray-rt-gpu. I can’t figure out why 3dsmax wireframe views are so slow. VRay is the best part of 3dsmax, imo.
The beta seems fine. I was expecting something faster. Octane has it’s own problems. Shader creation is really slow and unstable, last beta I tried. They probably are wise to use only NVidia CUDA, though. Only one GPU to worry about, which is probably why it’s so fast. It would be nice if Chaos Group at least tried CUDA-only option. But they are smarter than me so maybe when ATI gets it together with OpenCL support, they will have a great GPU solution, too.
Also, let’s not forget what RT is for. Previewing lighting and shaders w/ G.I. in near RT is extremely helpful. Try going back to mental ray final gather/GI calculation tweaking after using RT for a while and you’ll see what I mean.
Also good to keep in mind that Octane, while capable of doing a few things pretty well, does not have nearly the range that RT does. How much faster is Octane if you want to move two objects relative to each other? Or maybe add a spotlight or three?
Octane is interesting and has some potential, but if it was covering the same ground that RT was it might not be or seem so quick. Just my POV.
b
For RT/GPU, an FX5800 (290 cores) couldn’t be considered even remotely top-of-the-line for GPU rendering. Remember that the Chaos Group’s tests were based on a machine with 4 GTX480s (1920 cores) on a single frame.
I just rendered your test file and it did what I would expect. RT/CPU (i7 3.2GHZ) rendered 726 samples in a little over 2 minutes, with RT/GPU (GTX 480, 336 cores) rendering the same 726 samples in under 40 seconds.
I see, but the point was GPU renderer vs GPU renderer right? Production ready, biased renderer vs beta stage GPU unbiased renderer can’t compete much, it will be unfair to battle them..
I even color corrected and distorted the result + glow to match the photo.
Open my scene and press activeshade (I set timelimit to 10minutes). Then, the guys, who’re willing to compete - just convert scene to octane.
edit
Seems I included old version of my scene Will up later new one.
Hi guys, you cannot compare nvidia CUDA with OpenCL.
Cuda is much faster… octane, arion, and much more … they use CUDA.
Vray, for internal politics, no. It use OpenCL. So it’s normal that is slower.
I see that you cannot compare the number of path because Cuda will be always faster and the images with less noise .
Vrayrt is good, in my opinion, not for production but fast previsualizations. For now…
…And that’s probably why OpenCL may be fundamentally slower than straight CUDA on nVidia hardware. Chaos Group chose OpenCL so that V-Ray/RT users can be free and not tied to a specific hardware company for GPU rendering. The downside of this will be some temporary performance incongruity while developers decide what, how heavily, and when to support standards in the world of GPU rendering.
Unfortunately, nVidia’s chief competitor, ATI, does not have OpenCL implementation that is presently usable by RT. That should change and when it does, you will very likely see a very different picture out there. Just give things a little time.
It may currently seem that in the short term Chaos Group is not making the fastest GPU renderer out there, but knowing these folks like I do I predict that in the long term, V-Ray users will have the best in integration, compatibility, freedom to choose competing hardware, ease of use, and yes, speed in their GPU rendering solution.
Hm. When I compared RT to Arion - I saw a difference in kind of noise they produce. It was the first released working piece of Opencl software that made me want RT so much
And RT was a lil slower, when using meshlight emitters. Vlado promised, it would be faster. One day.
P.S. It would be great, if RT will support AA filters.
Will up later new one.