Evaluating Containerization for High-Performance Computing
Keywords:
Containerization, High-Performance Computing, Docker, Singularity, Resource Utilization, Workload Performance, Author Name, Scopus, Springer, Wissira, Journal Name, Journal Short Form, Wissira Press, Wissira Research Lab, Research Gate, SSRN, ISSN, Academia, UGC Care, PubMed, WOSAbstract
Containerization has become a groundbreaking technology in software deployment and development, providing portability, scalability, and efficiency. Its adoption in High-Performance Computing (HPC) environments is increasing, mainly because of its ability to optimize resource utilization and enhance workload performance.
This paper assesses the viability of containerization for HPC workloads, highlighting its advantages, limitations, and overall performance impact. A detailed statistical comparison between containerized and non-containerized HPC systems determines key performance indicators. The research concludes that although containerization is a feasible solution for specific HPC workloads, its effectiveness is based on workload characteristics, container orchestration techniques, and underlying hardware settings.
