Twelve Ways To Fool The Masses When Giving Parallel-In-Time Results

TL;DR

This paper highlights common pitfalls in evaluating parallel-in-time methods, emphasizing the importance of careful performance assessment to avoid overestimating algorithm efficiency and speedup.

Contribution

It provides a humorous yet insightful guide on avoiding misinterpretations and pitfalls in performance evaluation of parallel-in-time algorithms.

Findings

Identifies twelve common ways to mislead performance results

Provides practical advice to ensure accurate performance assessment

Raises awareness about pitfalls in parallel-in-time benchmarking

Abstract

Getting good speedup -- let alone high parallel efficiency -- for parallel-in-time (PinT) integration examples can be frustratingly difficult. The high complexity and large number of parameters in PinT methods can easily (and unintentionally) lead to numerical experiments that overestimate the algorithm's performance. In the tradition of Bailey's article "Twelve ways to fool the masses when giving performance results on parallel computers", we discuss and demonstrate pitfalls to avoid when evaluating performance of PinT methods. Despite being written in a light-hearted tone, this paper is intended to raise awareness that there are many ways to unintentionally fool yourself and others and that by avoiding these fallacies more meaningful PinT performance results can be obtained.