Although instruction yield (or 27, 27, 27) , introduced with ISA v2.06, is listed as a way to change PPR's (Program Priority Register) field PRI to provide a performance hint that performance will probably be improved if shared resources dedicated to the executing processor (or thread) are released for the other threads, it was never in fact implemented in hardware of POWER7 and above, have the same effect of a 'nop' instruction. mdoio, mdoom, and miso, just like yield, are listed as other ways to manipulate the hint performance from a thread but in fact was never implemented for POWER as well.

That said, it's possible to achieve a similar effect by using other 'or' instruction to set thread priority to low or very low, which helps, for instance, on spinlocks, as ISA says:

By setting the PRI field, a programmer may be able to improve system throughput by causing system resources to be used more efficiently. E.g., if a program is waiting on a lock, it could set low priority, with the result that more processor resources would be diverted to the program that holds the lock. This diversion of resources may enable the lock holding program to complete the operation under the lock more quickly, and then relinquish the lock to the waiting program.