A fast vectorized sorting implementation based on the ARM scalable vector extension (SVE)

TL;DR

This paper presents a fast, vectorized sorting implementation optimized for ARM's SVE, demonstrating significant speedups over standard algorithms by adapting to SVE's unique features.

Contribution

The paper introduces a novel vectorized sorting method tailored for ARM SVE, addressing its unique predicate and variable vector size, and achieves substantial performance improvements.

Findings

Achieves 4x speedup over GNU C++ sort

Efficiently handles different data types including integers and doubles

Adapts well to ARM SVE's predicate and variable vector size

Abstract

The way developers implement their algorithms and how these implementations behave on modern CPUs are governed by the design and organization of these. The vectorization units (SIMD) are among the few CPUs' parts that can and must be explicitly controlled. In the HPC community, the x86 CPUs and their vectorization instruction sets were de-facto the standard for decades. Each new release of an instruction set was usually a doubling of the vector length coupled with new operations. Each generation was pushing for adapting and improving previous implementations. The release of the ARM scalable vector extension (SVE) changed things radically for several reasons. First, we expect ARM processors to equip many supercomputers in the next years. Second, SVE's interface is different in several aspects from the x86 extensions as it provides different instructions, uses a predicate to control most operations, and has a vector size that is only known at execution time. Therefore, using SVE opens new challenges on how to adapt algorithms including the ones that are already well-optimized on x86. In this paper, we port a hybrid sort based on the well-known Quicksort and Bitonic-sort algorithms. We use a Bitonic sort to process small partitions/arrays and a vectorized partitioning implementation to divide the partitions. We explain how we use the predicates and how we manage the non-static vector size. We explain how we efficiently implement the sorting kernels. Our approach only needs an array of O(log N) for the recursive calls in the partitioning phase, both in the sequential and in the parallel case. We test the performance of our approach on a modern ARMv8.2 and assess the different layers of our implementation by sorting/partitioning integers, double floating-point numbers, and key/value pairs of integers. Our approach is faster than the GNU C++ sort algorithm by a speedup factor of 4 on average.