I am interested in improving efficiency of applications and statistical algorithms that have direct impact in industry. My current focus is analyzing application behavior on diverse high-performance computing (HPC) hardware.
As part of my Ph.D. research, I used statistical models to capture computational behavior and predict performance of several HPC applications, and scientific and sparse-matrix kernels, prior to their run time. I was funded through the Lawrence Graduate Scholar Program while I worked at Lawrence Livermore National Laboratory for a part of my Ph.D.
I have been at Intel for the last year-and-a-half while wrapping up my dissertation to graduate in December 2016. My dissertation is titled “Statistical Techniques to Model and Optimize Scientific, Numerically-Intensive Workloads.”
I can be reached at steena.HPC at gmail.