1. Does NeuroServer support running two Starlight Mini instances simultaneously on Apple Silicon?
2. If so, how does the total rendering time scale on an Apple Ultra with 256 GB of unified memory compared to running a single instance? Or on Apple M5 MAX with 128 GB RAM?
Yes, you can export 2 Starlight Mini exports simultaneously, although it is not recommended. Speeds will be about half as fast as processing 1 at a time.
I was able to run two Starlight Mini instances on my 96 GB M3 Ultra. However, that wasn’t the case with Starlight Precise. The 2nd instance of SLP was queued.
My understanding was that Starlight Mini was called via FFmpeg rather than Neuroserver. Perhaps I’m mistaken there, but I’m certain that Starlight Precise IS via Neuroserver.
Thanks.
Andy
