Migrating processes to different CPUs under current OS models involves migrating not only the process itself, but any RAM it is using, any files it has open, etc. This is not trivial on an OS level and I think GNU/Hurd is the only operating system even attempting to be something like this.
I'm sure IBM already did this on some type of hardware in the 60's or 70's (IBM has done everything already) on mainframes so the base concepts are likely not new.
If the type of work is something less than "anything" then it's a lot easier to develop a "work server" for that type of work, such as an encoding server, indexing server, etc. The amount of data that needs to be moved back and forth includes a lot less than "everything a process" needs. So this is why things like this tend to be application specific.
Something like an "indexing server" is going to be "multi-node multi-master replicated database" and it's possible to set one up in MySQL but not trivial and not guaranteed to work with whatever software your NAS is working.
You should consider not using your NAS for indexing and use a separate app on a more powerful computer to index files, or consider building your own NAS on more powerful hardware.
You can no more do this than you can hook up an old PC to a new PC ad affect the same sharing of CPU resources as you describe. – JakeGould – 2019-10-08T17:40:39.773