Search

The complexity and size of scientific and engineering challenges are continually increasing to the point where they are approaching exascale computing. One of the main challenges is to be able to run scientific software applications on these extremely large systems in an energy efficient manner, as power consumption will become one of the dominant cost factors that will govern the next generation of large high performance computing data centers. In this paper, we present the results of node level power measurements on a petaflop HPC system.

Energy consumption will become one of the dominant cost factors that will govern the next generation of large HPC centers. In this paper we present the Dynamic Voltage Frequency Scaling (DVFS) Plugin to automatically tune several energy related tuning objectives at a region-level of HPC applications. This plugin works with the Periscope Tuning Framework which provides an automatic tuning framework including analysis, experiment creation, and evaluation. The tuning actions are based on changes in the frequency via the DVFS. The tuning objectives include the tuning of energy consumption, total cost of ownership, energy delay product and power capping. The tuning is based on a model that relies on performance data and predicts energy consumption, time, and power consumption at different CPU frequencies.

We introduce a model for the illumination of multilayered surfaces. The model enables the simulation of numerous optical characteristics based upon the physics of light, such as thin film interference that occurs between layers. The technique can also account for spatial distribution, scattering and diffraction characteristics associated with subsurface particles