Paradigm Shift: From Data Centers to the Rugged Edge
When a federal employee takes a trip out into the field to collect data—whether that is water and soil samples or sensitive data on the movements of troops—they need to get those datasets processed and analyzed as quickly as possible. Most federal agencies, however, are operating with their computing power housed in traditional core data centers, either their own or a cloud service provider’s, requiring the data to move from the field to the core to be processed. This legacy infrastructure set up, doesn’t have high capacity computing power at the edge to do complex processing quickly, so there is latency in how quickly an organization can benefit from data collected in the field.
In a traditional edge to core to cloud computing environment, data is collected throughout the network and sent back to the core servers for processing. This became the standard as most devices connected to the outer edge of the network didn’t have enough compute or storage capacity to process the data where it was collected.
In today’s environments, edge computing technology is not optimized for the emerging Internet of Things (IoT) data-intensive workloads. There is an over-abundance of data being transmitted to, and stored in, core data centers, of which only a small amount can be analyzed in real-time, and much of it is never used at all.
Private industry is now shifting focus from processing data in the data center to processing data at the rugged edge where it is collected, with only the insights from the data transmitted back to the core data center. By processing data at the edge, organizations are able to get near-real-time feedback to optimize everything from supply chain operations to customer experience. In the private sector, the company that moves first and fastest generally wins, so there is a lot of investment to enable more capable intelligent edge computing systems. As technology progresses, the size of devices decreases and their computing power improves exponentially. The private sector has realized how revolutionary having high powered edge computing is for real-time data analysis and developed systems that are rugged, portable and powerful enough to compute real-time calculations anywhere.
Now is the time for the federal government to take advantage of years of private industry investment – computing at the rugged edge is now a commercial domain. Section 256 of the FY20 National Defense Authorization Act recognizes the need to accelerate the Defense Department’s adoption of commercial edge computing, and asks for funding to be injected into the top priority programs. As agencies, like the DoD, look to employ artificial intelligence and machine learning in a deployed environment, investing in high powered edge computing becomes mandatory.
When evaluating rugged edge computing solutions, we generally compared Size, Weight and Power (SWaP), and settled for the amount of compute available in a rugged form factor. Thanks to the ground-breaking developments in the commercial sector, we can now have fully scalable data center quality server nodes in a small, rugged form factor, that are manufactured as commercial-off-the-shelf (COTS) items, at COTS scale and COTS prices. Going forward we need to assess rugged edge computing solutions using SWaP-C3, where the C3 stands for Compute, Cooling and Cost. We no longer need to settle for low-power computing capability for operating in high temperatures, and pay high prices when procuring rugged edge computing systems.
The HPE Edgeline EL8000 system is a great example of this capability. The result of more than two and a half years of research and development to meet the need for rugged high powered intelligent edge computing for both the telco and oil and gas industries, it has four data center quality server nodes in a form factor that will fit in the overhead compartment of a commercial airline and operate at 130 degrees Fahrenheit. Federal and DoD organizations can now move any application/workload from a core data center out to the edge, to include artificial intelligence and machine learning – no more compromising.
For more information on HPE’s edge computing solutions, visit the Iron Bow website.
TechSource in your Inbox
Sign-up here to receive our latest posts by email.