use std::transform_reduce for performance

[scarf005]

Summary

utilize the power of Functional Programming, i.e std::transform_reduce to improve performance

Benchmarks (lightweight structs)

std::parallel_reduce was the fastest across all benchmarks.

• 5 elements: same as ranged for loop and std::accumulate
• 100, 1000 elements: x1.1 faster than ranged for loop, x1.8 faster than std::accumulate
• 10000 elements: x1.2 faster than ranged for loop, x1.4 faster than std::accumulate

5 elements

100 elements

1000 elements

10000 elements

[scarf005]

exciting

we'll have to figure out a way to add tbb to our toolchain first, tho