The requirements for HPC in the cloud are not much different than the requirements for HPC on-prem. When you have a project that requires accelerated compute or if you have a computing problem that scales up and/or out, it’s a good time to consider HPC in the cloud, hybrid, or on-prem. In many cases, the decision to go to the cloud is based on the availability of on-site resources and the time frame for the project being executed.
When HPC in the Cloud makes sense:
When you’re working on an on-premises HPC cluster running at 80% capacity and your organization gets a large project that would out-strip your current compute capacity, it’s a good time to consider bursting to the cloud to cope with your occasional HPC projects. If your organization only runs sporadic HPC jobs or has determined that owning hardware is not the best investment, it makes more sense to burst to the Cloud. The primary benefit of bursting HPC to the cloud is that the cloud provider, such as Nimbix, will be running the newest and most up to date systems, allowing your organization’s computing dollar to go farther.
Proof of Concept
When looking to install an HPC platform that will reside on-premises hardware must be purchased. A cloud provider such as Nimbix can serve as a proof of concept (POC) center and allow you to run benchmarks on a variety of hardware configurations, without the cost of on-site hardware. This ensures that your organization is making the best possible purchase for your current and anticipated needs.
When an organization is burdened by a lack of real estate to devote to an on-premises system and the ever increasing operational costs of maintaining older systems, cloud platforms make sense. A cloud solution solves two issues; alleviating space constraints and staffing issues by outsourcing the labor and real-estate issues to the cloud provider.
-The original author was Tom McNeil