Cloud infrastructure environment based on data cube and machine learning technologies.

The usage of traditional high-performance computing (HPC) is a challenge for scientific workflows because of the high cost of acquisition and a lack of expertise to deploy and support such in-house infrastructures. HPC as a Service cloud model has tremendous potential of spreading the outreach of HPC to the broader scientific and industrial community considering the advantages of elasticity, flexibility, and customization. The mismatch between performance-oriented HPC and cost and resource-utilization oriented cloud an infrastructure is an additional barrier to build a cloud-based optimized HPC environment. Therefore, we aim to provision HPC capabilities over the cloud, combining scientific discoveries in cloud data fusion, algorithms, methods, and SaaS through comprehensive performance and energy efficiency analysis. [ABSTRACT FROM AUTHOR]