Analysis of users' first contact with High-Performance Computing

Abstract

Machine Learning and Deep Learning algorithms have become a great revolution in many research fields such as Robotics and Artificial Intelligence. They have applications in such different areas as Meteorology, Cybersecurity, Biology, etc.; though their use in these areas is not so extended. High-performance Computing (HPC) is the most powerful solution to get the best results by using these algorithms. HPC requires various skills to use, such as parallel programming and shell scripting on linux system which may require a long time to acquire and might be intimidating for research with small or no background in Information and Communications Technology (ICT) such as meteorologists or biologists. This work intends to encourage the use of HPC techniques among non-ICT researchers. In order to do so, we plan to analyze the response of such researchers when they are presented some new techniques and possibilities. A set of experiments are being carried out with a group of ethology researchers at Eötvös Loránd University. We will use a three-step methodology. First, researchers will fill out a questionnaire about their knowledge about and attitude towards HPC techniques. Then, they will attend an introductory talk about HPC in general and some specific use cases which may aid them in their research, after which they will receive a follow up questionnaire. After this a subset of the attendees will receive hands-on training in the use of specific HPC applications. Finally, a last round of questionnaires will be carried out with the participants. We expect to identify some key indicators which allow us to identify the main advantages and drawbacks that non-ICT researchers face when discovering High-performance Computing.