Finalized:Monday, May 29, 2017
Author(s):Wyngaard, J., H. Lynch, J. Nabrzyski, A. Pope and S. Jha
Given the current scientific questions of societal significance, such as those related to climate change, there is an urgent need to equip the scientific community with the means to effectively use high-performance and distributed computing (HPDC), Big Data, and tools necessary for reproducible science. The Polar Computing RCN project (http://polar-computing.org) is a National Science Foundation funded Research Coordination Network, which has been tasked with bridging the current gap between the polar science and HPDC communities. In this paper we discuss the effectiveness of “hackathons” as a model for implementing both the pedagogical training and the handson experience required for HPDC fluency. We find hackathons effective in: (i) Conveying to a science user how and why HPDC resources might be of value to their work, (ii) Providing a venue for cross discipline vocabulary exchange between domain science and HPDC experts, (iii) Equipping science users with customized training that focuses on the practical use of HPDC for their applications, (iv) Providing hands-on training with a realistic domain-specific application in a community of one's peers, but are (v) an incomplete training model that requires supplementation via domain science specific HPDC training materials. In addition to their pedagogical benefits, hackathons provide additional benefits in terms of team building, networking, and the creation of immediately usable products that can speed workflows both for those involved in the hackathon as well as others not involved in the hackathon itself.
J. Wyngaard, H. Lynch, J. Nabrzyski, A. Pope and S. Jha, "Hacking at the Divide Between Polar Science and HPC: Using Hackathons as Training Tools," 2017 IEEE International Parallel and Distributed Processing Symposium Workshops (IPDPSW), Lake Buena Vista, FL, 2017, pp. 352-359. doi: 10.1109/IPDPSW.2017.177This material is based upon work supported by the National Science Foundation under Grant No. 1542058. Opinions, findings, conclusions or recommendations expressed are those of the authors and do not reflect the views of the NSF.