We have recently tested out the new V-Ray RT rendering engine from Chaos Group to see if it can speed up our production process and most importantly allow for faster feedback when setting up lighting, cameras and textures in scenes. Chaos also allowed us to trial the full beta version of V-Ray RT GPU which utilises the Graphics Processing Unit or Graphics Card for rendering instead of the usual CPU (Central Processing Unit) or processing power of your computer.
The advantage of V-Ray RT is that you can get almost real time feedback on your scene without waiting to render the scene each time. You can move the camera, change objects, lighting, materials etc and they automatically update within your viewport . The other thing i wanted to test is whether render times would also be improved and if Graphic Processing rather than Central Processing for visualisation was the way forward. Using the GPU isn’t a substitute for the production Vray renderer, but as a tool for quickly setting up your scene ready for render. The CPU and GPU spec of my computer can be seen below:
The CPU processing power used was Dual Intel Core 2 Quad processors @2.83GHz (8 cores)
The GPU processing power used was ATI Fire GL V7700
I got a very similar result in terms of render quality and time taken when using either the CPU or GPU renderer. The computer used for the tests has been set up with high central processing power needed for rendering and animation, which is why it was quite surprising to get a similar result from a fairly standard graphics card. The next step if we are to go with GPU processing will be to upgrade our graphics cards for increased render power and next generation GPU’s like the ‘NVIDIA Quadro Fermi’ will take visualisation workflow far beyond current capabilities.
I have added a couple of links to videos by Peter Mitev below showing the process in more detail: