adecemployee wrote:A quick question..
The "Update Geometries" option warns you not to use it in its current state for complex or large scenes. I can certainly understand why.. But what criteria would you say makes a complex scene?
Also down the road, what hardware would you say most drives the ability to keep the live stream between octane and c4D for this live geometry update? I see the plug-in as a potential live design tool while sitting with a customer, but some of our scenes can be extremely large. Currently using one Tesla C2075 6GB GPU with this plugin and loving the results, however down the road we are looking at the new K series Tesla - my thinking is the increase in memory speed may help this live plug-in update faster, and the additional cuda cores help the rendering times.. Or perhaps the updates to the GPU are more of a system memory speed issue? Just wondering your thoughts on where to spend the money.
Thanks again!!
I must give more info about Update Geometries process. This option do same thing with full scene export process, but only for updated objects. So each process has voxelisation phase. This is most time consuming step in update operation. And done by Octane SDK. I'm not sure what is the most slow part. Data transfer or geometry parsing. Before 1.11 version, voxeling was taking long time due single threaded processing.
Best way to test with different cards on same pc configuration. Which will give a idea. Maybe Titan or Tesla K20 can give you boost.
I can compare Titan vs. Gtx 590 for some processes.