Today’s scientific applications usually take considerable time to run, and hence parallel computing environments, such as Grids and data centers/Clouds, have emerged. Indeed, traditionally, much research in high-performance computing has been conducted with the goal of executing such applications as fast as possible. However, energy has recently been recognized as another crucial goal to consider, because of its negative economic and ecological implications. Energy-driven solutions in these environments are mostly focused on the hardware and middleware layers, but little efforts target the application level. We revisit a catalog of primitives commonly used in object oriented-based scientific programming, or micro-benchmarks, to identify energy-friendly variants of the same primitive. Based on this, we refactor three existing scientific applications, resulting in energy improvements ranging from 2.58% to 96.74%.
CITATION STYLE
Mateos, C., Rodriguez, A., Longo, M., & Zunino, A. (2016). Energy implications of common operations in resource-intensive java-based scientific applications. In Advances in Intelligent Systems and Computing (Vol. 444, pp. 739–748). Springer Verlag. https://doi.org/10.1007/978-3-319-31232-3_69
Mendeley helps you to discover research relevant for your work.