Performance Anomalies in Concurrent Data Structure Microbenchmarks

TL;DR

This paper investigates how microbenchmark design influences performance evaluation of concurrent data structures, revealing significant variability and potential misinterpretations in reported results.

Contribution

It identifies key microbenchmark design factors affecting performance results and provides best practice recommendations for accurate benchmarking of concurrent data structures.

Findings

Performance results can vary 10-100x based on microbenchmark implementation.

Flawed microbenchmark design can invert performance rankings.

Advanced features in Setbench improve benchmarking accuracy.

Abstract

Recent decades have witnessed a surge in the development of concurrent data structures with an increasing interest in data structures implementing concurrent sets (CSets). Microbenchmarking tools are frequently utilized to evaluate and compare the performance differences across concurrent data structures. The underlying structure and design of the microbenchmarks themselves can play a hidden but influential role in performance results. However, the impact of microbenchmark design has not been well investigated. In this work, we illustrate instances where concurrent data structure performance results reported by a microbenchmark can vary 10-100x depending on the microbenchmark implementation details. We investigate factors leading to performance variance across three popular microbenchmarks and outline cases in which flawed microbenchmark design can lead to an inversion of performance results between two concurrent data structure implementations. We further derive a set of recommendations for best practices in the design and usage of concurrent data structure microbenchmarks and explore advanced features in the Setbench microbenchmark.