English

Optimal Round and Sample-Size Complexity for Partitioning in Parallel Sorting

Distributed, Parallel, and Cluster Computing 2023-05-30 v6

Abstract

State-of-the-art parallel sorting algorithms for distributed-memory architectures are based on computing a balanced partitioning via sampling and histogramming. By finding samples that partition the sorted keys into evenly-sized chunks, these algorithms minimize the number of communication rounds required. Histogramming (computing positions of samples) guides sampling, enabling a decrease in the overall number of samples collected. We derive lower and upper bounds on the number of sampling/histogramming rounds required to compute a balanced partitioning. We improve on prior results to demonstrate that when using pp processors, O(logp)O(\log^* p) rounds with O(p/logp)O(p/\log^* p) samples per round suffice. We match that with a lower bound that shows that any algorithm with O(p)O(p) samples per round requires at least Ω(logp)\Omega(\log^* p) rounds. Additionally, we prove the Ω(plogp)\Omega(p \log p) samples lower bound for one round, thus proving that existing one round algorithms: sample sort, AMS sort and HSS have optimal sample size complexity. To derive the lower bound, we propose a hard randomized input distribution and apply classical results from the distribution theory of runs.

Keywords

Cite

@article{arxiv.2204.04599,
  title  = {Optimal Round and Sample-Size Complexity for Partitioning in Parallel Sorting},
  author = {Wentao Yang and Vipul Harsh and Edgar Solomonik},
  journal= {arXiv preprint arXiv:2204.04599},
  year   = {2023}
}

Comments

12 pages

R2 v1 2026-06-24T10:43:28.968Z