# Parallel Sorting on GPU: Part 1

This post series is dedicated to the description and implementation of sorting array on a graphics processing unit. It’ll be split into two parts.
In this part, we’ll try to derive one of the first efficient parallel sorting algorithm from the sorting networks class. What’s specific is that before performing some sorting algorithm from this class, we already know which fixed positions elements we’ll compare-and-swap at.
Then we’ll calculate the algorithm’s serial and parallel time complexities.