[@bill qu] should not matter what size of dedicated video RAM is available as the program is supposed to fall back to CPU only processing if there is insufficient GPU resource type or specification remember they specifically boast about it:
Smart Hardware Acceleration: Automatic CUDA, Vulkan, and CPU fallback cascades ensure lightning-fast processing while preventing system crashes.
Not exactly supportable as you can be assured that with GPU acceleration it will be slow and if it has to fall back to multi-core CPU it WILL BE SLOOOOOOOOOW processing never lightning-fast processing as it’s based upon individual raster frame upscaling so has to be decoded into masses of rasters/bitmap images, processed one frame at a time and then finally lossily re-encoding the video from the upscaled raster images and merged back with the audio streams.