Text

theses

The computational demand of high-performance computing (HPC) applications has brought major changes to the HPC system architecture. As a result, it is now possible to run simulations faster and get more accurate results. But behind this, power and energy are becoming critical concerns for HPC systems, e.g. Titan’s electric cost is about $9 million per year.  Energy efficiency has become a critical challenge for the exascale research challenges, and U.S. Department of Energy’s (DOE) gives the goal to achieve exascale performance with a power budget of 20MW. Current research efforts have studied power and performance tradeoffs, and how to balance these, e.g., using DVFS to meet power constraints, which significantly impacts performance. However, scientific applications may not tolerate degradation in performance and other tradeoffs need to be explored to meet power budgets, e.g., involving the application in making energy-performance tradeoff decisions.  In this research, we focus on studying the properties and exploring the performance and power/energy tradeoffs of Low-Mach-Number Combustion (LMC) application which is an Adaptive Mesh Refinement (AMR) algorithm. Our experimental evaluation provides an empirical evaluation of different application configurations that gives insights into the power-performance tradeoffs space for this LMC or AMR-based application workflows. The key contribution of this work is a better understanding of the running behavior of this AMR-based application and proposed a power-performance tradeoff for this application, which can be used to better schedule power budgets across HPC systems.

ETD_7275

M.S.

Includes bibliographical references

by Yubo Qin

doi:10.7282/T3S184P7

ETD graduate

The author owns the copyright to this work.