I saw that you noted that OpenCL isn't as mature as CUDA for what you're doing and I was wondering what specifically you meant by it. Is it an IDE thing (cross-platform GUI dev) or the actual languages. I know that Chaos Group is working on porting their CUDA V-RayRT permanently to OpenCL and I have this unbiased SmallLuxGPU render benchmark that also uses OpenCL so I'm just wondering what the issue is, especially since you said that you had to rework a lot of code for multi-GPU support, which you get for free with OpenCL. I have been playing around with dual graphics cards in my Mac Pro and all the OpenCL stuff just uses them all (and the CPUs too, if you want):
http://grab.by/grabs/bc6ebe69c796d83714 ... 928219.png
Anyway, I'm not a programmer so I was wondering if this was a real barrier for you or just something you (understandably) don't want to revisit.
curious about OpenCL exclusion
Forum rules
Please add your OS and Hardware Configuration in your signature, it makes it easier for us to help you analyze problems. Example: Win 7 64 | Geforce GTX680 | i7 3770 | 16GB
Please add your OS and Hardware Configuration in your signature, it makes it easier for us to help you analyze problems. Example: Win 7 64 | Geforce GTX680 | i7 3770 | 16GB
It's slow, very very slow.
I'm also not sure about it's future.
CUDA works fast and now, and octane development was started when opencl did'nt exist yet.
Anyways, once i see some numbers (the smalluxgpu's ones are horrible) we can do a port to opencl, it is not that much work.
Radiance
I'm also not sure about it's future.
CUDA works fast and now, and octane development was started when opencl did'nt exist yet.
Anyways, once i see some numbers (the smalluxgpu's ones are horrible) we can do a port to opencl, it is not that much work.
Radiance
Win 7 x64 & ubuntu | 2x GTX480 | Quad 2.66GHz | 8GB
-
SurfingAlien
- Posts: 355
- Joined: Thu Jan 21, 2010 12:05 am
- Location: Italy
Hi!
I'm not programmer too, but I have the feeling the "pure" CUDA-GPU Octane approach is the best as for today. Of course this doesn't mean it will be so for the future, and I think people at Refractive Software have their good reasons for choosing a platform that cut off a half of their potential customers (i.e. ATi powered people).
@cgbeige: from a quick look at your statistics under SLG I think there's something wrong with your settings... your 4870HD is twice as faster than your GTX285 and I don't think this is right (imho).
I'm not programmer too, but I have the feeling the "pure" CUDA-GPU Octane approach is the best as for today. Of course this doesn't mean it will be so for the future, and I think people at Refractive Software have their good reasons for choosing a platform that cut off a half of their potential customers (i.e. ATi powered people).
@cgbeige: from a quick look at your statistics under SLG I think there's something wrong with your settings... your 4870HD is twice as faster than your GTX285 and I don't think this is right (imho).
Last edited by SurfingAlien on Mon May 03, 2010 12:53 pm, edited 1 time in total.
MacBookPro unibody 2x2.4GHz | 4Gb | 9400M+9600GT 256Mb | OSX 10.6.8
i7 2600-K @3.4GHz | 8Gb | GTX560Ti 2Gb | Windows7x64 | Octane 1.0 b2.52
i7 2600-K @3.4GHz | 8Gb | GTX560Ti 2Gb | Windows7x64 | Octane 1.0 b2.52
GTX275(lower model than GTX285) is faster than HD5870(higher model than 4870) .. See 03 and 04 ..
http://www.luxrender.net/forum/viewtopi ... e8d5511ca1
http://www.luxrender.net/forum/viewtopi ... e8d5511ca1
Win7 64 | 2x XEON | 12GB | 8x GTX580
Win7 64 | 2x XEON | 12GB | 8x GTX580
Win7 64 | 2x XEON | 12GB | 8x GTX580
Win7 64 | 2x XEON | 12GB | 8x GTX580
Win7 64 | 2x XEON | 12GB | 8x GTX580
Win7 64 | 2x XEON | 12GB | 8x GTX580
Win7 64 | 2x XEON | 12GB | 8x GTX580
ya, I noticed the slower GTX 285 score. Anyway, my second GTX 285 arrives today so I won't bother trying to figure out what the problem is there. Mixing GPUs is just a bad idea in my opinion, even on OSes that are relatively good at it, like OS X.
Thanks for the response radiance. Good to hear it from someone who knows what they are talking about.
Thanks for the response radiance. Good to hear it from someone who knows what they are talking about.