Exploiting Cloud Computing for Algorithm Development

Abstract

We consider the application of cloud computing to the process of algorithm development. We introduce a case study focusing on the development of a novel algorithm in computational electromagnetics, illustrating several challenging areas for algorithm developers where cloud-based architectures can deliver enhanced productivity and potentially save costs. The development, verification and tuning of our algorithm have all been assisted by cloud-based technologies. Our preliminary results both demonstrate the potential of the algorithm to solve the problems accurately, and of cloud-based architectures to accelerate the development and verification process. We propose that cloud-based architectures will in the future play a greater role in the development of algorithms; saving costs by improving hardware utilisation, and reducing turnaround time.