super fast mbvh4 traversal and triangle intersection

[sngan]

we have developed 2 new traversal methods for mbvh4. One for coherent transport and one for incoherent ray traversal (e.g. diffuse bounces >= 2nd order). on top of that new triangle intersectors have been implemented to further speedup the process of rendering triangle meshes. the traversal methods outperform current embree kernels  by up to 4x for camera rays, primary occlusion, ao and 2x for incoherent ray transport. we also did some measurements against gpu hardware (titan and others). for some setups our kernels perform better than any state of the art gpu implementation (e.g. "understanding the efficiency of ray traversal on gpus") when only simple shading was activated. a highly optimized cpu path tracer (developed with the techniques above) outperforms easily any gpu pt when it comes to complex light transport. intels upcoming avx-512 enabled cpus will push this gap even further. more infos are available here: http://rapt.technology/

[Ondra]

Looks interesting. Do you have a library implementation we could try available? What is your licensing scheme?

[arqrenderz]

So you are offering your new tech to corona ? We will always welcome the extra speed in any form :)
Hope you can make business with them

[Juraj]

yay, new tech :- ) merge, merge, merge...

I have this friend coding for state-of-art photogrametry/lasercan soft who remarked raytracing has been over-researched and nothing further can be done I don't know what he's been smoking...

[sngan]

Looks interesting. Do you have a library implementation we could try available? What is your licensing scheme?

hallo, we can put something togehter for initial testing. for coherent transport we have to find a simple-to-use interface
for you. the other one just needs a list of rays in any order.

[Ondra]

The absolutely best option for us would be if you forked embree and added it as another algorithm