June 19, 2014

Students Learning HPC

Tiffany Trader

Considering the much bemoaned talent gap that exists within HPC, it’s always rewarding to hear about the uber-talented students that are pushing the boundaries of computer and computational science. The National Institute for Computational Sciences (NICS) at the University of Tennessee is currently facilitating two student-focused HPC endeavors: the SC14 Student Cluster Competition and the Computational Science for Undergraduate Research Experiences (CSURE) 2014 summer internship program.

The Student Cluster Challenges have evolved to introduce students to HPC and to showcase student expertise in a friendly yet spirited competition. Currently there are three such events, at SC, ISC and the most recent addition, the Asia Student Supercomputer Challenge (ASC). Staff members at NICS are sponsoring and mentoring a Student Cluster Competition team for SC14, which will take place in New Orleans, from Nov. 16–21. Mentors Dr. Vincent Betro of the Scientific Computing group and Paul Peltz of Operations at NICS selected the students based on their diverse interests and strengths in core HPC areas, including operating systems, compiling, hardware, and visualization.

“Students are very excited about this competition since it provides them with an opportunity to work with cutting-edge hardware and software as well as round out their HPC education far beyond what is offered at the undergraduate level in coursework,” says Dr. Betro in a feature story authored by NICS science writer Scott Gibson.

The NICS student team has been taking part in weekly lunch-and-learn meetings at ORNL with their mentors, providing them with the opportunity to learn from their mentors and other experts who can start preparing them for the applications that they will need to be familiar with.

Hardware for the competition is being provided by Cray, Intel, Mellanox and NVIDIA. The students will be tasked with assembling, installing, and maintaining the cluster. They will also need to tune the hardware and the software and stay within the 26-amp power limit set by the compeition guidelines.

The other exiting student-focused program at NICS is CSURE – the Computational Science for Undergraduate Research Experiences, a summer internship program, which is also featured on the NICS site. Students in the ten-week program just completed their first week of activities on the campus of the University of Tennessee, Knoxville, and Oak Ridge National Laboratory (ORNL).

Taking part in the program are 10 students from US colleges who were carefully selected from a competing pool of thirty. They are joined by four students from the Chinese University of Hong Kong. While the students have diverse backgrounds and interests, they are connected by a passion for learning. During the first week, they received orientation, had an introduction to scientific computing, and toured the high-performance computing facilities at ORNL.

The Oak Ridge Leadership Computing Facility, which was a main stop on the tour, houses some of the most powerful supercomputers in the world, machines such as Titan, Gaia, and Darter, we well as the now-decommissioned Kraken.

Working with the program mentors and other HPC experts, the students are getting a crash course in the main HPC skill areas, including Linux, C programming, parallel programming, and more. During their stay, they will also be exploring domains such as atmospheric sciences, chemistry and materials sciences, engineering mechanics, numerical libraries on emergent platforms, and systems biology. The mission of the program is to provide the participants the knowledge and skills necessary to begin using computational methods.

CSURE is sponsored by the National Science Foundation Research Experience for Undergraduates, or REU, and conducted by the Joint Institute for Computational Sciences (JICS). JICS was established by the University of Tennessee and Oak Ridge National Laboratory (ORNL) to advance scientific discovery and state-of-the-art engineering, and to further knowledge of computational modeling and simulation.

Share This