Exploring Parallelization with a Raspberry Pi Cluster
Description/Abstract/Artist Statement
In this work, I introduce an affordable and scalable project to explore parallel computing concepts for undergraduate STEM students. A cluster computer composed of Raspberry Pis is presented along with a specific use case to explore the performance of the Pi cluster and examine the consequences of unbalanced task distribution across the cluster. The performance of the cluster is tested using both simple numerical integration and adaptive integration methods. In the case of simple integration, results show that the cluster provides speedup in accordance with expectations due to the equal time-complexity of individual computations. Adaptive integration serves as a use case to examine the importance of equitable task management across the cluster since the tasks assigned to individual threads may be of different time-complexities.
Faculty Advisor/Mentor
Ayman Elmesalami, Soad Ibrahim
College Affiliation
College of Sciences
Presentation Type
Oral Presentation
Disciplines
Numerical Analysis and Scientific Computing | Programming Languages and Compilers | Theory and Algorithms
Session Title
Colleges of Sciences UG Research #3
Location
Zoom
Start Date
3-19-2022 3:30 PM
End Date
3-19-2022 4:30 PM
Exploring Parallelization with a Raspberry Pi Cluster
Zoom
In this work, I introduce an affordable and scalable project to explore parallel computing concepts for undergraduate STEM students. A cluster computer composed of Raspberry Pis is presented along with a specific use case to explore the performance of the Pi cluster and examine the consequences of unbalanced task distribution across the cluster. The performance of the cluster is tested using both simple numerical integration and adaptive integration methods. In the case of simple integration, results show that the cluster provides speedup in accordance with expectations due to the equal time-complexity of individual computations. Adaptive integration serves as a use case to examine the importance of equitable task management across the cluster since the tasks assigned to individual threads may be of different time-complexities.