High Performance Cluster Scale-Out
Pharma/biotech, financial services, federal agencies, and other research-focused industries depend on High Performance Computing (HPC) for modeling, quantitative analysis, and other complex informatics. Companies invest millions each year building out their clusters, but this is often driven by peak usage for resource-intensive jobs that only run for days or weeks. As a result, internal resources can be oversubscribed during peak periods but underutilized the rest of the time.
Instead of constantly building out your internal clusters or grids at enormous cost, you can scale out to the cloud for greater flexibility and reduced capex:
- Build a cluster of hundreds or thousands of servers in minutes, paying only for what you use
- Run HPC jobs in the cloud as if they were running internally, using the same queuing processes and management/monitoring tools
- Use the cloud as an extension of your security perimeter, providing the same protection and control available in your data center
Using CloudSwitch, you can build a secure HPC cluster in the cloud in minutes that matches your internal environment. With a few clicks, you first deploy or migrate the distributed computing management (DCM) server that coordinates jobs across compute nodes. With a few more clicks, you can then stamp out dozens, hundreds, or thousands of virtual nodes to form the cluster. Using your existing tools, deploy the operating system on each node, then the DCM software that enables jobs to run in parallel. The result is a fully-functioning HPC cluster in the cloud, with the same look and feel as if it was running locally.
A well-known pharmaceutical company is turning to the cloud to provision massive computing clusters on demand for complex informatics, rather than waiting days or weeks. Researchers submit jobs to the queue as they always have, but now those jobs are distributed across a cluster of compute nodes in the cloud that automatically scales up or down depending on the length of the queue. The company has integrated the cloud into its infrastructure strategy, with flexible, secure access to massive computing resources and endless scalability.
