Several of my colleagues at Oak Ridge National Labs (including myself) got involved in high performance computing through the Student Cluster Competition. More about the competition can be found at the competition website or in the post I wrote earlier when I competed, but essentially it challenges undergraduate students to design, build, and manage a small HPC cluster to run a set of scientific applications quickly and efficiently. This competition does a fantastic job of getting students involved in scientific computing and system engineering, and as such one of my coworkers wanted to host a team for 2019.
At the start of the summer we selected a team of students from local universities to compete. We then worked with them for the next several months to teach them how to submit jobs, compile and profile scientific code, manage a diskless cluster, and several other HPC fundamentals. Additionally we were able to teach them a lot about large scale system management, and by the start of the Fall they were able to stand up a diskless cluster with xCAT. In November the team traveled to Denver for SC19 and competed against teams from around the world. I was extremely lucky to have been able to work with such a talented and motivated group of students, and I’m very thankful I had the opportunity to do a small amount to help further their education.