It supports CUDA, or ROCm (AMD) or CPU but as I read it certain dependencies like Pytorch only supports ROCm on Linux (not Windows or MAC) and on MAC it only supports CPU. You can see this here if you select the operating system: https://pytorch.org/get-started/locally/
So, when something in the dependencies gets updated, it might break compatibility with AMD or MAC releases. And Steinberg has to work around that. This might be harder than you think.
my hesitations are multiple for upgrading from SL11 to 12
1_I’m nearing the end of a long project; I’m cautious to not overly disturb my workflow
2_Will SL12 run as well on my current computer as SL11 does? I’m not in position to upgrade computer at this time. Running a trial seems like a better solution for me here…so I await the trial to be offered
Getting SL12 has been a life-saver, it’s been one of those things that I couldn’t imagine life without now that I’ve tried it.
I didn’t pay much attention to it and then I seen a youtube video of a hip-hop artist using it to remove background noise from vocals and they had a sale in May so I grabbed it. It’s resolved a few irritating issues that would have been impossible otherwise, and I find it’s amazing for precise editing like where you want to smooth over a little imperfection between 1-3khz without destroying the audio.
I only spent a month with SL11, it was great - but the new modules are awesome.
My only gripe is that the ARA integration has a long ways to go (it can corrupt projects), but I just discovered that standalone mode can still show your track in beats/measures so I prefer to just use standalone and save my own versions.