A year ago I bought the GTX 580 from Gainward. It was 100% compatible back then with CS6, and it had good performance for its price.
But in the last couple of months I've been noticing some trouble with my GPU. Things aren't going as fast anymore. One factor is probably because I'm getting so used to the current speed that I have.
Especially rendering 3D objects (ray-traced 3D) inside a Premeriere sequence doesn't go as fast as I hoped for.
So, I was thinking to just buy a new GPU since Adobe CC support more card now and there are many new cards out there.
Now there are 2 cards that have my attention. And that are the GTX 690 and the GTX 780. But you also have Ti series, and different kinds of brand (Gainward, EVGA, Asus, ...). And there is where it's confusing me.
I have a budget of max €/$ 1000 (really the max.). And I'm just just looking for the best performance in After Effects, Premiere Pro and Speedgrade for its price.
Any suggestions on this? I can also look at the specs, but I've been told that it aren't always the specs that does it. And I don't really know where to look (much CUDE cores, much GRAM, fast speed in Mhz, .. ???)
The second question that I have, is if I can use 2 GPU's. My current GTX 580 and my future one. I would like to hook up one monitor on my GTX 580 and my second monitor on my future card. But is this really going to work?
By doing that I'm hoping to use my GTX 580 for random stuff and my future card for pure rendering. But I'm not sure if I can tell my PC or software to do that? So, should I just better sell my GTX580, or can I still gain advantages from it as a second GPU? (GTX 580 doesn't have SLI)
Thanks in advanced!
> Especially rendering 3D objects (ray-traced 3D) inside a Premeriere sequence doesn't go as fast as I hoped for.
That will never use the GPU. Dynamically linked compositions do not use the GPU.
> The second question that I have, is if I can use 2 GPU's.
Yes. That is directly addressed here:
"If you have multiple GPUs installed, the GPU-accelerated ray-traced 3D renderer will use the CUDA cores on all of them, as long as they are of the same CUDA compute level. (See the technical specifications of your GPU for its CUDA compute level.)"
Thanks. This explains a lot why rendering was so slow. Is this by chance anything the developers are working on? It kinda takes away the advantage of having dynamic link..