@mic they both use the same capture and training ideas, but the engine powering the trained data is different

NAM uses WaveNet, which is known to use quite a lot of CPU.
AIDA-X uses RNN style (how they different exactly I cant say, but they are from different scientific papers and research)

Sadly the NAM plugin doesn't support Linux. :(
The LV2 port is still WIP and not too stable in my opinion.