From a workforce development perspective, high performance computing is a niche in the computing community, which is itself only one among many STEM fields. Students are not typically exposed to HPC as a discipline during their formal training for STEM careers, unless their university happens to have a computational science major, or has made a concerted effort to produce computationally-capable scientists and engineers. More troubling, students are often not even exposed to the many ways in which technical computing is used as part of everyday practice in their field of study. As a result, many will leave their formal training with significant gaps in the current practice of their field, transferring the responsibility of creating a computationally-capable workforce to the organizations that have chosen to hire them.
There are many forces conspiring to slow the integration of solid computational training into university curricula, from accreditation standards that do not sufficiently motivate STEM departments to include computation to the habits of long-time professors who find it difficult to break with traditional approaches to their topics. These are difficult problems to solve, but if collectively we continue to fail to address them we will ultimately be responsible for slowing down the pace of innovation in our society.
Students must leave universities ready to take up the state of the practice in fields that routinely use computational tools, and ready to advance the state of the practice in fields that have been slower to adopt.
While universities work through their internal processes and adapt to change, the HPC community and government-funded programs around the world have, of necessity, stepped in to help address the computational knowledge gap themselves. The supercomputing program I work for in the U.S. Department of Defense dedicates significant resources to training its users in computational techniques, starting with the most basic concepts. The U.S. National Science Foundation is also a major force in computational education through its many programs, and contributes to the outstanding online resource HPC University, along with the Partnership for Advanced Computing in Europe and many others.
SC13 is also substantially invested in addressing the computational knowledge gap, providing many opportunities for students and young professionals to bootstrap themselves into the application of HPC to their field of specialty, or turn a basic foundation into more advanced skills.
This week I've had the privilege of visiting with one of children of a long-time colleague and friend, a recently graduated engineer that did not get much exposure to computing during her training. As he watched her early career develop, he encouraged her to seek out opportunities to move into a more computational approach to her daily job. As her skills, largely self-taught, developed over the first year of her career, both she and her company realized that even basic computational tools put into routine use would dramatically improve the effectiveness of their designs and make them a more effective company. She made the case for attending SC13 this week, and has spent her time so far in tutorials including Parallel Computing 101 and Effective HPC Visualization and Data Analysis using VisIt. She reports being completely enrapt during these sessions — a good sign that we've got a long-time convert to the field!
At dinner the other night she shared a decision to go back to school and pursue a master's degree in computational science or engineering. We talked about the booths she should visit on the exhibit floor, and the people she needs to meet at the conference as she goes through the process of finding a good school and shaping a degree program.
This is a technical conference at its best — a venue in which a young career is transformed by the personal connections and in-person experiences that only happen when you are in the same space as many other like-minded folks. As economies around the world have struggled in recent years, some organizations, notably the United States government, have radically restricted the ability of their employees to attend technical events. This may be penny-wise, but the pound-foolish consequences of a poorly trained, poorly connected, and disengaged young STEM workforce will be felt for decades to come.
No entries found