OpenRL can be seen as a ray tracing oriented version of OpenCL with the capability of taking advantage of all available processing units in a system.
OpenRL is the world’s first cross platform, heterogeneous API for ray tracing. It runs on Linux, Mac OS X, and Windows, and will identify and take advantage of CPUs, GPUs, and custom hardware devices all simultaneously. Ray tracing application developers can download and evaluate OpenRL free-of-charge from our website.
This technical example demonstrates OpenRL’s heterogeneous computing ability by leveraging a Macbook Pro’s Intel Core 2 Duo processor, together with an NVIDIA GeForce 9400M graphics processor to solve the ray tracing problem. Essentially whichever device is ready to do more work then OpenRL will assign more rays to it. The result is the fastest performance of ray tracing on a laptop today<.
another opencl toys_!
Since they’re using OpenCL on a G80 as an example, I wonder if they’re going to ditch there niche Caustic hardware entirely and just focus on the OpenRL API.
“It runs on Linux, Mac OS X, and Windows, and will identify and take advantage of CPUs, GPUs, and custom hardware devices…”
Maybe the custom hardware devices is their Caustic hardware. More speedups with it because it’s hardware, dedicated asics!