Crate pmsa 

Parallel Mergeing of two Sorted Arrays using rayon.

This crate provides the highly parallelized algorithm to merge two sorted slices in the following functions.

• par_merge
• par_merge_by
• par_merge_by_key

Benchmark Results

The benchmark runs the sequential and parallel versions on Intel i7-10750H CPU (12 hyperthreads) using release profile. It is tested on two sorted u64 arrays. The parallel version is shown to be ~6 times fater than the sequential counterpart.

per array len	sequential	parallel
10^6	14.56987ms	2.219912ms
10^7	139.675856ms	23.363867ms
10^8	1.427501286s	325.121204ms

Functions

par_merge

Merges two sorted slices in parallel.

par_merge_by

Merges two sorted slices in parallel with a custom comparison function.

par_merge_by_key

Merges two sorted slices in parallel with a custom key function.

seq_merge

Sequentially merges two sorted slices.

seq_merge_by

Sequentially merges two sorted slices with a custom comparison function.

seq_merge_by_key

Sequentially merges two sorted slices with a custom key function.