When the CPU is looking for the next bit of data to execute and can't find it in the L1 cache, the L2 cache, or system memory, the data must be paged in from the hard drive. This is called a "compulsory miss", because it cannot be avoided; at some point, every piece of data will need to be fed into the cache for the first time.

Prefetching is a nifty trick that helps work around the inefficiency of a compulsory miss.

Log in or registerto write something here or to contact authors.