ompf2.com

Really? They patented it or something? I didn't see anything mentioned about that in the paper or the website where I downloaded it from.

See http://www.ece.ucsb.edu/~psen/Papers/EG13_RemovingMCNoiseWithGeneralDenoising.pdf for an example of a recent one. This looks pretty amazing, even the denoising as a simple post-process(MLD in the paper) looks very good. This could be a very important contribution to real-time gi if the method w...

Which noise reduction techniques? E.g., adaptive wavelet rendering, various multi-scale filtering methods, or maybe even some offshoot of that wacky gradient-domain MLT stuff. That adaptive wavelet rendering paper looks really interesting (if you can live with the artifacts, which games most probab...

Which noise reduction techniques?

Actually, 13 years is not really that bad. It is overseeable and a few years could be shaved off by using multiple gpu's. I like the realistic attitude (35x speedup over today's hardware required) NVIDIA has adopted, much better than the marketing from the past.

In one presentation at SIGGRAPH (http://on-demand.gputechconf.com/siggraph/2013/presentation/SG3106-Building-Ray-Tracing-Applications-OptiX.pdf), NVIDIA implies that there is an upcoming version that might address this. Slide 5 from that presentation called 'Real Time Path Tracing' is interesting. ...

For those who hadn't seen it yet, some benchmarks of the Xeon Phi.

About 2x bvh build performance of an 8-core Xeon(slide 37) and about 3.5x path tracing incoherent rays performance of an 8-core Xeon(slide 45).

The 'hybrid traversal' method that is mentioned in the slides looks interesting.

Shared virtual memory looks interesting. I didn't read anything on C++ support in kernels however. That is a feature I would have expeced from OpenCL 2.0.

It all sounds rather complex and time consuming what you have planned there. Moreover, what works well on a certain manufacturers gpu's, does not necessarily work well on other manufacturers gpu's. And optimizations for one generation of gpu's may turn out not to work well on newer generations of th...

Interesting Haswell review from the German website computerbase: "Das ist gut zu sehen in Cinebench x-CPU, in dem ein 4770K acht Prozent schneller ist als ein 3770K, das Gesamtsystem aber auch 13 Prozent mehr Leistung benötigt" . Haswell 8% faster on Cinebench but draws 13% more power. Really discou...

Yes, OpenCL 1.1 has been supported for a while, but the performance is usually not up to par. NVIDIA has fixed OpenCL Titan drivers few weeks after the release. There was some known problem with multiple OpenCL applications (not only LuxMark) at the time of the release. I assume they are now on par...