liveBook · Manning

Parallel section

As explained in Grokking Concurrency

A parallel part is a portion of an algorithm whose independent operations can execute concurrently.

A parallel section is the part of an algorithm where independent operations can run at the same time. An algorithm may also contain sequential sections that must happen in order, so parallel execution cannot make every task infinitely fast. section 2.11

Searching index cards

Divide a mixed pile of cards among several people, let everyone search independently for cards about concurrency, then combine the findings. Dividing and combining remain sequential; searching is the parallel section. With two helpers, the five-second process falls to three seconds rather than one, because the sequential work still takes time. section 2.11

How Grokking Concurrency covers it

How deep it goes
Defines independent parallel work and contrasts it with sequential steps through a timed example, without implementation details.
Examples
Uses a mother-and-baby analogy and a cooperative index-card search to illustrate concurrency and speedup limits.
What you can do after
Evaluate which algorithm steps benefit from parallel execution and why sequential work limits total speedup.
How it connects
Connects parallel sections with sequential execution, parallel algorithms, Amdahl’s law, and speedup.

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Grokking Concurrency. The section links point to the parts of the book it draws on.