skip to main content

Women, Minorities Gain Parallel Computing Knowledge

7/13/2010 3:56:00 AM

Yeh Center
Yeh Center

By Megan Kelly, CSL

What factors impact the start-up and success of small businesses? Can someone control a machine using only his or her thoughts? These are just a couple of the questions students in Research Experiences for Undergraduates (REU) aim to answer using parallel computing.

The National Science Foundation awarded Illinois a three-year REU grant totaling $358,160. The project, “Passionate on Parallel – A Summer Research Program for Women in STEM,” funds undergraduate parallel computing research with a special focus on women, minorities and students attending universities without advanced research opportunities.

Civil and Environmental Engineering Associate Professor Susan Larson and Computer Science Associate Professor Craig Zilles lead the endeavor as principal investigators. The project’s inspiration also stems from a vision shared by Bill Sanders, Director of the University of Illinois’ Coordinated Science Laboratory (CSL), CSL Professor Wen-Mei Hwu, and CSL Senior Research Scientist Umesh Thakkar. The proposal was sponsored by the CUDA Center of Excellence, housed in CSL.

Parallel computing is a form of computation that makes simultaneous calculations using the idea that large problems can be divided into smaller ones and solved concurrently. Hwu, co-director of Illinois’ Universal Parallel Computing Research Center, said parallel computing is indispensible to the future progress of science and engineering, yet is not often offered at the undergraduate level.

“Parallel computing will be the main venue for computational tools in all science and engineering disciplines to gain more speed and more capability,” Hwu said. “We need to drastically increase the number of scientists and engineers who are well trained in parallel computing.”

To address this challenge and prepare the undergraduates for their research projects, Zilles developed a one-week boot camp to give the students a firm foundation in the principles of parallel computing.  This lab-centric, pair programming-based course uses a "bottom-up" approach inspired by the text "Introduction to Computing Systems: From Bits and Gates to C and Beyond," co-authored by CSL's Sanjay Patel.  The REU project is making the course materials freely available to facilitate similar programs at other institutions.

Thakkar said studies indicate that very few women and minority students, particularly African-American and Latino students, major in computer science, and they become even more underrepresented in graduate school. Thakkar and colleagues decided to incorporate a diversity focus to facilitate the development of more women and minorities in computer science and give them an edge in tomorrow's workforce.

The project began June 8 and includes 10 students representing six universities and multiple ethnicities. The students are working on six parallel computing projects, spanning a variety of topics.

“There are many disciplines outside of engineering and computer science that use parallel computing, for example, economics and atmospheric sciences,” said Larson, who is also the director of Women in Engineering and an assistant dean. “We deliberately looked for diverse projects to show how vital this area of research is applied to other disciplines.”

The Illinois REU project also aims to initiate student interest in graduate school.

“Many students in the REU program might not have considered attending graduate school without this opportunity,” Larson said. “Because Illinois is such a fabulous research school, it’s a wonderful place to get this experience.”

 In Fall 2007, Intel and Microsoft recognized Illinois as one of two Universal Parallel Computing Research Centers in the nation.

“Illinois is by far the most vibrant community in parallel computing,” Hwu said. “There is a very large community of researchers and educators on all aspects of parallel computing. We have the responsibility to lead the charge in educating the coming generation of scientists and engineers on parallel computing.”