An Empirical Study of Cache-Oblivious Priority Queues and their Application to the Shortest Path Problem

TL;DR

This paper empirically evaluates cache-oblivious priority queues, demonstrating significant speedups over traditional methods in external memory scenarios, but also highlighting limitations when accessing graph adjacency data.

Contribution

It compares two cache-oblivious priority queues and applies them to shortest path problems, providing experimental insights into their practical performance.

Findings

Cache-oblivious queues outperform internal memory techniques in limited RAM scenarios.

Speedups are reduced when graph adjacency list access dominates runtime.

Experimental results validate theoretical advantages of cache-oblivious algorithms.

Abstract

In recent years the Cache-Oblivious model of external memory computation has provided an attractive theoretical basis for the analysis of algorithms on massive datasets. Much progress has been made in discovering algorithms that are asymptotically optimal or near optimal. However, to date there are still relatively few successful experimental studies. In this paper we compare two different Cache-Oblivious priority queues based on the Funnel and Bucket Heap and apply them to the single source shortest path problem on graphs with positive edge weights. Our results show that when RAM is limited and data is swapping to external storage, the Cache-Oblivious priority queues achieve orders of magnitude speedups over standard internal memory techniques. However, for the single source shortest path problem both on simulated and real world graph data, these speedups are markedly lower due to the time required to access the graph adjacency list itself.