Reduced Cost as no need for dedicated hardware and pay-as-you-go model
Better utilization of Resources
Cloud provider takes care of failure management
Just spin up the HPC Cluster in Cloud, complete it ,scale down and finish)
Existing Solution
Virtualization Overhead
Resource Contention
Separate resource pool for Cloud and HPC
Problem Identification
"To design a Cloud platform that is capable of hosting both commodity and HPC applications with Near Native performance and Better Performance Isolation."