While 2015 has been the year of cloud, it has also been the year of High Performance Computing (HPC). As more and more agencies embrace the cloud, they also have to consider how best to leverage and, in some cases, accelerate accompanying compute resources.
The Department of Defense (DOD) has been a true innovator and proactive in taking steps toward modernization by way of HPC with the Global Formation Grid Services Management-Operation program and the High Performance Computing Modernization Program, respectively.
Some of the most impressive DOD examples in leveraging HPC include:
U.S. Sen. Thad Cochran’s office recently shared that DOD will participate in an $82.7 million Army contract meant to increase compute capabilities within the command. As part of the DOD’s High Performance Computing Modernization Program, the U.S. Army Corps of Engineers in Engineering Resource and Development Center (ERDC) and Stennis Space Center in Mississippi will participate in the initiative.
Mississippi Senator Thad Cochran shared that he is “…pleased that the ERDC and Stennis employees and communities have proven to be smart places to invest taxpayer dollars for the most intensive modeling and simulation missions of the Department of Defense.”
As the need for HPC grows, we continue to see the benefits of the Global Information Grid Services Management-Operation (GSM-O) program, established in 2012 by Defense Information Systems Agency (DISA). According to William Welsh of GCN, DISA will “launch a new configuration management tool for the Global Information Grid on Sept. 28 that will give users better visibility into network assets.”
A top DISA official also explained that this is one of many enhancements being implemented in the DOD’s worldwide data network this month. Additionally, Jessie Showers director of DISA’s Infrastructure Directorate shared “On the immediate horizon, what we are planning to do is implement a series of technology insertions centered on global IT service ordering.”
While DOD has embraced HPC and is implementing the proper infrastructure to carry out current initiatives- like the Global Formation Grid Services Management-Operation program and the High Performance Computing Modernization Program, respectively- a supercomputer is not a requirement for high-speed computing.
For example, according to Fabien Sanglier of Software AG Government Solutions, “due to the very high costs of building and maintaining such high performing grids…government agencies…are faced with growing requirements for larger-scale data processing and real-time, data-driven execution – without the budget for supercomputing capabilities.”
Yet, he goes on to explain that ramping up computing power can be done in one of two ways- vertically with a super computer or horizontally with “scaling through in-memory computing.”
According to Sanglier, horizontal scaling can be a viable, realistic option for agencies with strapped budgets as it utilizes existing compute components and capabilities with accompanying scaling capabilities such as Hadoop.
To learn more about how your agency can leverage existing infrastructure for Modernization and High Performance Computing efforts, click here.