Search found 36 matches

by lion
Thu Feb 13, 2014 11:05 am
Forum: General Development
Topic: Hair intersection : approaches ?
Replies: 16
Views: 13220

Re: Hair intersection : approaches ?

What if combine both approaches?
  • Build BVH of tessalated hair.
  • Use triangles at big distance.
  • At small distance compute "curved mesh" ray intersection only if ray hits BVH leaf.
  • Possibly split "curved mesh" to smaller and simpler curves for each BVH leaf\group of leafs.
by lion
Sun Feb 09, 2014 6:49 pm
Forum: General Development
Topic: CUDA and branch prediction
Replies: 2
Views: 3871

Re: CUDA and branch prediction

Yep, I know about that, we need all warp do exactly same instructions to achieve full utilization.
Also I note that there no effect from converting bra\bra.uni instructions in simple kernels (probably hardware branch prediction buffer), but gain some effect in complex kernels.
by lion
Thu Feb 06, 2014 3:18 pm
Forum: Visuals, Tools, Demos & Sources
Topic: Arauna2 announcement video
Replies: 14
Views: 65528

Re: Arauna2 announcement video

ahh, much better. is this scene public available as obj/mtl pair ? Download free version of unity http://unity3d.com/ Download robot lab scene https://www.assetstore.unity3d.com/#/content/7006 Import robot lab scene in new project. Place obj exporter http://wiki.unity3d.com/index.php?title=ObjExpor...
by lion
Sat Feb 01, 2014 11:07 am
Forum: General Development
Topic: CUDA and branch prediction
Replies: 2
Views: 3871

CUDA and branch prediction

Branch is costly operation on GPU, so we want control branch weight if compiler failed to do so. Is there analog for __builtin_expect() or __assume() for CUDA? Same question on nvidia forum https://devtalk.nvidia.com/default/topic/458657/non-divergent-branch-nvcc-and-ptx/ but no one answers. My curr...
by lion
Wed Jan 22, 2014 6:43 pm
Forum: Links & Papers
Topic: Arauna2 new site
Replies: 4
Views: 7723

Re: Arauna2 new site

http://www.staff.science.uu.nl/~tan0010 ... vlasakker/ hehe) images and video looks very good)
Probably you see this, but I leave it here too http://www.luxrender.net/forum/viewtopi ... 34&t=10529
by lion
Wed Jan 22, 2014 1:37 pm
Forum: Links & Papers
Topic: Arauna2 new site
Replies: 4
Views: 7723

Re: Arauna2 new site

Development targets include further performance enhancements, support for complex materials, noise reduction, and post processing of real-time generated images.
noise reduction! xD I look forward to ;)
by lion
Wed Jan 22, 2014 1:00 pm
Forum: General Development
Topic: Csillámtrace - my new graphics engine
Replies: 153
Views: 100319

Re: Csillámtrace - my new graphics engine

There no opencl on link. 7 million triangle/sec - I think you look at wrong place, we look for rays\s, you need to calculate and show it in one of tutorialXX_ispc samples. Also note that arauna have sse optimizations and process packet of rays at once, so you need compare it with _ispc version of em...
by lion
Wed Jan 22, 2014 6:52 am
Forum: General Development
Topic: Csillámtrace - my new graphics engine
Replies: 153
Views: 100319

Re: Csillámtrace - my new graphics engine

I've read about GPU, I mean CPU OpenCL, it's easy way to see what you can get if you use C (not C++) code, so you can use same serial C code for your initial CPU path and OpenCL CPU path. Also take a look at this http://embree.github.io/ it's CPU only, have serial C++ code as well as optimized sse\a...
by lion
Tue Jan 21, 2014 9:04 pm
Forum: General Development
Topic: Csillámtrace - my new graphics engine
Replies: 153
Views: 100319

Re: Csillámtrace - my new graphics engine

I mean lib+headers+sample or viewer which can load some popular format (.obj for example) to try own scene.
PS: have you try use sse\avx to process packet of rays at once? Also you can use Intel\AMD OpenCL (CPU) and achieve sse\avx packet processing using serial C code.