Guidesort: Simpler Optimal Deterministic Sorting for the Parallel Disk Model

TL;DR

Guidesort is a new, simpler deterministic sorting algorithm for the parallel disk model that nearly matches the optimal I/O complexity with a small constant factor, improving on previous methods.

Contribution

It introduces a simpler, deterministic sorting algorithm for the parallel disk model that achieves near-optimal I/O performance with a smaller constant factor.

Findings

Achieves a constant factor C close to 3 of the optimal I/O complexity.

Simpler algorithm and analysis compared to previous work.

Potentially better performance for plausible parameter values.

Abstract

A new algorithm, Guidesort, for sorting in the uniprocessor variant of the parallel disk model (PDM) of Vitter and Shriver is presented. The algorithm is deterministic and executes a number of (parallel) I/O operations that comes within a constant factor $C$ of the optimum. The algorithm and its analysis are simpler than those proposed in previous work, and the achievable constant factor $C$ of essentially 3 appears to be smaller than for all other known deterministic algorithms, at least for plausible parameter values.