On 04/21/2013 04:46 PM, dsimcha wrote:
2. Different hardware than I tested on, maybe with better memory bandwidth.
Your implementation performs a lot of copying. Maybe an in-place parallel sort algorithm would perform better, e.g. parallel quicksort.
3. Expensive comparison functions. I didn't test this in D either because I couldn't think of a good use case. I tested the D parallel sort using small primitive types (ints and floats and stuff).
String sorting is a good use case with slightly higher comparison cost.
