I’ve been using Blender very lightly over the last 2 years, doing some simple animations here and there. I would like to get more serious with it, but my render times aren’t anywhere near what they should be with my system. My system is a dual Xeon 3.06ghz w/ Hyperthreading on (4 logical processors), Radeon 9800XT 256mb video card, and 2GB of RAM. I remember reading somewhere before about people making batch files to render large animations on a dual system, but with the large amount of upgrades and improvements to Blender recently, I was wondering if there is anything recently introduced for SMP support within Blender, or the best solution someone has come up with to use all available CPU’s to render a scene.
Thanks for your help
i think the os should balance both processors equally. however, what are you rendering/what do you think the render times should be? your system seems pretty fast - each of your processors is clocked five times faster then mine, and you’ve got twice as much, and HT… if you don’t go for 4k radiosity animation, you should be able to produce any kind of renderings blazingly fast.
Blender is not multithreaded. The best you can do with multiple processors to speed up animation rendering is to run two instances of Blender simultaneously, one trained to each proc, each rendering a different set of frames.
Sorry. This is one of a couple areas where Blender is still not ready to play with the big boys.
Thanks for the replies.
Does anyone know the current status of SMP development in Blender? Is this something they’re working on actively?
Converting Blender into a threaded application is not on the developers’ list right now, as far as I know.