Benchmarking the Performance of Scientific Applications with Irregular I/O at the Extreme Scale

In this paper we hypothesize that irregularities of I/O patterns (i.e., irregular amount of data written per process at each I/O step) in scientific simulations can cause increasing I/O times and substantial loss in scalability. To study whether our hypothesis is true, we quantify the impact of irregular I/O patterns on the I/O performance of scientific applications at the extreme scale. Specifically, we statistically model the irregular I/O behavior of two scientific applications such as the Monte Carlo application QMCPack and the adaptive mesh refinement application ENZO. For our testing, we feed our model into an I/O skeleton tool to measure the performance of the two applications' I/O under different I/O settings. Empirically, we show how the growing data sizes and the irregular I/O patterns in these applications are both relevant factors impacting performance.