Going parallel

The previous sections are applicable to either serial or parallel computing. Even in the most parallelizable number crunching program, a great deal of serial code is written, so these observations are very applicable. After a certain point, however, parallel concerns come to dominate. We will start this section by introducing some terminology, before looking at a simple example.


Wall-clock time is the amount of time that passes from the beginning of execution of a program to the end of its execution, as measured by looking at a clock on the wall. Wall-clock time is the measure people usually care about.

Cycle time is the time obtained by summing up the number of cycles taken by the program during its execution. For example, ...

Get Mastering IPython 4.0 now with the O’Reilly learning platform.

O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.