RE: The future of AI – self hosted open sourced models?

Right now, I'm using an AI plugin (OpenVivo) for Audacity (free music software) that separates the stereo audio track into individual instruments. For a 3-minute track, it needs 10 minutes to separate the track into 4 channels: vocal, guitar, bass, other instruments.
I have 12th Gen Intel(R) Core(TM) i7-12700 (2.10 GHz) and 32 GB Ram. Without a graphics card, I'm using integrated Intel. Probably I need to buy some NVidia monster :)