I've developed a new binary search that is up to 2 times faster than the binary search used by virtually every software application. It solves the problem of how to perform a boundless binary search by combing two novel adaptations.
I've licensed it under GPL 3.0 and it is available here: https://sites.google.com/site/binarysearchcube/binary-search It examines a few binary search variants, with benchmark graphs at the bottom, but the one of interest is: int boundless_binary_search(int *array, int array_size, int key) The algorithm is used in a sorting algorithm here: https://sites.google.com/site/binarysearchcube/sorting-variants Which allows to stress-test and verify that the algorithm works properly. As the graph shows, my binary merge sort outperforms quicksort and it should be a stable sort as well. Depending on the software implementation the performance gain can be quite significant. I am hoping you find this an interesting development and are willing to double check my work and/or forward it to interested parties. If my algorithm gains traction it should have a positive impact on mankind's carbon footprint. :) Kind regards, Gregorius van den Hoven
