Construction of a high‐performance computing cluster: A curriculum for engineering and science students

Abstract

In this article, a curriculum on the construction and maintenance of a high-performance computing cluster for students in engineering and science is presented. Laboratory sessions are an important part of this curriculum. Successive completion of the laboratory sessions automatically leads the student group to a homemade full-functioned cluster. Students are encouraged as they see their computing cluster working at the end of the curriculum. This curriculum can be of benefit to the students who are involved with computationally intensive simulations. ß

On a course on computer cluster configuration and administration

“…Other works are focused on building clusters to use them in different courses. In [2], [5] and [15], they implement real hardware clusters while [26] and [30] choose virtualization. In Finally, [26] and [30] draw on virtualization to develop "training" clusters.…”

“…A single cluster, consisting of one master and 10 slave nodes, is built. The cluster is shared between all the students [5]. proposes the construction and management of reduced high performance clusters as part of a course aimed at engineering and science students.…”

Teaching high-performance service in a cluster computing course

“…In [4], [8], [13] and [19], those clusters are implemented using real hardware while [31] and [34] choose virtualization. In general, most courses build the clusters pursuing two goals.…”

“…In general, most courses build the clusters pursuing two goals. First, it allows the students learning how to make and test an HPC cluster [8], [19]. Moreover, in second place, many courses take advantage of the cluster built to run parallel programs [4], [13], [31], [34].…”

Evaluation of parallel programming teaching methodologies: On‐campus versus online methodologies

“…In fact, some works have described experiences teaching advanced undergraduate or graduate parallel programming courses. We refer the reader to for insights. In particular, it is a great challenge to analyze the user's activity in learning management systems (LMS) with the purpose to detect difficulties and deficiencies in online methodologies .…”