Page 1 of 1

Load scene into GPU | Speed

Posted: Mon Dec 21, 2020 11:49 am
by LichtUndSchatten
I´ve some general question about the possible bottleneck of the time which octane naeeded to load a complex scene/objects into the GPU RAM. Which are here the main things which could be optimized to get it faster? In production there is a lot of time needed until a complex scene is loaded into. And if the scene is loaded into GPU sometimes we are with active ipr able to add new models to the scene without reloading and sometimes not. In which constellation should this work and in wich case not.

Thanks a lot :)

Re: Load scene into GPU | Speed

Posted: Mon Dec 21, 2020 9:16 pm
by juanjgon
Improving the scene extraction time using multiple threads is a planned feature, but I'm afraid that currently releasing the Octane 2020.2 and Octane 2021.1 plugin builds has more priority, and there is still a lot of work to do.

Thanks,
-Juanjo

Re: Load scene into GPU | Speed

Posted: Mon Dec 21, 2020 10:39 pm
by Lewis
Yes we need multi threading of that part, currently that's my biggest need in octane, I'd change that for any 10 new features :).