A Descriptive Study of Amazon EC2: Performance Analysis

Abstract

This research examines the performance of Amazon EC2, an online computer rental cloud service, for various tasks. We gathered data from a student survey, IT experts and cloud engineers, and also examined public data sets and performance benchmarking. We concentrated on factors such as computer speed (CPU), memory, storage, network response time, latency, and expense. The findings indicate that EC2 is most effective if the correct type of instance is selected for the job. For instance, compute-optimized instances are quickest for intensive computations, memory-optimized instances are suitable for large data, storage-optimized instances are most suitable for disk I/O, and accelerated computing instances are extremely quick for AI or machine learning but are expensive. Some issues such as lag, intermittent downtime, and wasted resources were also observed. It proposes how to improve EC2, for example, reducing costs, improving resource management, and increasing data centers in various regions.