[BScout]

Audio is inherently a serial process (if you have a signal chain, try to process the compressor before the eq that sits before it in the signal path; if you have 4 notes in a bar, try to process the fourth note at the same time as the first note.)

That is why you end up with a distribution like that. It is also why, for audio work, higher clock speed is more important than more cores.

(to be clear, if you have multiple independent tracks, then you can load balance those better since the tracks have no dependency on each other until the are gathered for summing at the end for the output)