AA-sort is an integer sorting algorithm, which exploits SIMD and multi-core.
It is proposed by H. Inoue, T. Moriyama, H. Komatsu, and T. Nakatani at 2007(see "A high-performance sorting algorithm for multicore single-instruction multiple-data processors").
I tried to implement it on x86/x64 with SSE4(for only one processor) and verified that it is 2.8~4 times faster than std::sort(STL) for random data.
The source code is https://github.com/herumi/opti/blob/master/intsort.hpp and implementation detail is AA-sort with SSE4.1.
