For a long time AMD didn't compete in GPU render software that is basicaly dominated by CUDA codes. Now we finaly see AMD with their own Pro Render spread as a plugin and even integrated in most relevant cad softwares of the industrie, wich is very welcome. I always think that cpu is the way to go for render. you have all the features, no memory limits that GPU has(in comparison). But yes gpu is a lot faster than cpu, because of it make sense to have all your render passes on cpu, but the visualization when modeling or animating on gpu.
But if your scene is so big that you still able to render on cpu, but you can't visualize when modeling? why did you invested in that really expensive card(or a bunch of it), if now you can't utilize it, because of the memory limit of the card?
Here comes my question, as AMD comes with this Radeon Pro SSG card. It's a vega with 16GB memory(or should i say cache memory), and 2TBytes of ssd's on it. AMD made it clear that vega has this technology called HBCC that make a lot more sense to me now. So my question is, my software MAYA (or could be any other) can load a TByte scene if i have a system memory for it, but the thing is: the Pro Render software once identified the card can load my TByte scene on it? Or can only see 16GB and all 2TB is only for specific jobs and that's it?
I obviously didn't talk about performance here, i also can put more cards to help on a render job, but my question is the ability to load the scene. Don't even think i ever have a TeraByte scene, but you got the idea.
So for AMD Pro Render software, the AMD Pro SSG is a 2TB video card or a 16GB video card?
Anyone with the knowledge, please.
sorry for my english.
Hi there. HBCC will allow you to access system memory for out of core rendering. SSD would allow you to have textures out of memory in a file cache for fast access.