When using @distributed parallel code, if the amount of data needed to be manipulated is large, will CPU utilization reach 100%, which can be used as a basis for successful parallel？ (of course, direct test runtime is more accurate)
This depends on your task and whether it is CPU-bound, the details of the hardware setup, how the data is accessed, and a lot of other details.
3 Likes