Tag: Hank Childs

Achieving Scalable Performance on the Intel Xeon Phi for Data-Intensive Workloads

Presenter: Kirsten Dawes

Co-Presenter: Elliott Ewing

Mentor: Hank Childs, Computer Science

Oral Presentation

Majors: Mathematics and Computer Science

As new computer architectures are developed, various concerns are created. Understanding scalability of performance is an important concern for many legacy programs on Intel’s Xeon Phi, a many integrated core architecture. These legacy programs use a significant amount of computational resources, which poses the issues of time, power, and memory consumption as the problem sizes increase. We will be looking at the scalability for the execution time of a program as the number of processors increase. Given this constraint, we should expect to see a decrease in the time of execution of a program using more processors. Exploring data-intensive workloads will allow us to be able to track how certain legacy programs mitigate the issue of scalability on this new architecture. We explored this new architecture by creating two competing mini applications, where each mini application replicated a variation on a legacy program called “VisIt”, which works on data intensive workloads. The two mini applications explored thread affinity and work affinity by implementing image processing filters with a data flow pipeline. We conducted experiments creating different size output and input images; number of image processing filters; type of image processing filters; and number of threads running. The experiments produced timing data based on CPU time as well as time per thread. With the quantitative data, we produced, we were able to conclude that data-intensive workloads like “VisIt” should produce excellent scalability on this new architecture.

Future Flood Risk in the Columbia River Basin Under Climate Change

Presenter(s): Laura Queen

Faculty Mentor(s): Hank Childs & Phil Mote

Oral Session 2 M

The Columbia River has long provided resources as a cultural, economic and ecological agent in the Pacific Northwest. People have congregated along the Columbia’s banks throughout history, from the earliest settlements to contemporary metropoles, but this close proximity poses a serious threat when extreme flooding occurs. Understanding how climate change will affect the future flood risk throughout the Columbia River Basin is imperative for risk mitigation and infrastructural planning. To address this question, we are using an ensemble data set which provides daily streamflow values (1950-2100) for 172 different future projections for 396 locations in the Columbia Basin. To run just one future projection, a modeler must make four choice decisions: the representative concentration pathway (RCP), global climate model (GCM), meteorological downscaling method (MDM), and the hydrological model setup. This ensemble dataset contains 172 projections created by a modeling decision chain containing 2 RCPS, 10 GCMs, 2 MDMs, and 4 setups. With an ensemble dataset produced by multiple hydrologic model parameterizations, we are able to diminish the influence of human-made modeling decisions and find a trend in flood risk change amongst the 172 projections. From the daily time-step streamflow data, we fit probability distributions to extreme events from each water year and estimate flood statistics for floods with 10, 20 and 30 year return periods. From this analysis, we find a substantive increase in flood risk for all outlets sites in the Columbia River Basin and are beginning to study the correlation between sub-basin snow-dominance and increased flood risk.