Increasing system and algorithmic complexity combined with a growing number of tunable application parameters pose significant challenges for analytical performance modeling. We propose a series of robust techniques to address these challenges. In particular, we apply statistical techniques such as clustering, association, and correlation analysis, to better understand the application parameter space. We construct and compare two classes of effective predictive models: piecewise polynomial regression and artificial neural networks. We compare these techniques with theoretical analyses and experimental results. Overall, both regression and neural networks are accurate with median error rates ranging from 2.2 to 10.5 percent. The comparable accuracy of these models suggest differentiating features will arise from ease of use, transparency, and computational efficiency.

PDF file

In: Symposium on Principles and Practice of Parallel Programming (PPoPP)

Details