Today, IoT data is driving an explosion in computing located at the edge of the network. Gartner believes that by 2020, an astonishing ~50% of all enterprise-generated data will be created and processed outside of a traditional centralized data center or cloud environment – up from just 10% today*. With a multitude of IoT-enabled applications and their requirements for low latency, autonomy, and security, it is becoming critical to get the compute function physically closer to where the data is being generated.
The demands at the edge are driving three key requirements IT architects must consider for any effective edge computing solution.
Requirement 1: Rapid Decision Making
Many modern IoT applications need to quickly make critical decisions as new data is created and/or arrives. Transporting data back to a central location with a distant compute device isn’t realistic. A round trip of data to the cloud and back does not meet the latency requirements of applications that are emerging at the edge. Decisions such as how to avoid traffic collisions to keeping factory equipment from failing need to be made rapidly in order to provide the type of outcomes these modern applications demand.
Requirement 2: Fast Data Compression and Cleansing
It’s believed that over 500 zettabytes (1 zettabyte = 1 trillion gigabytes) will be generated by 2019**. Managing, sifting through, and analyzing the massive datasets being created at sensors requires an efficient means for cleansing and aggregation. Not only will large volumes of data need to be analyzed close to the source but they will need to be aggregated and cleansed for transport back to the data center or the cloud. Providing effective data compression and cleansing near the data source reduces the cost of transport and compute in the cloud, ensuring the best use of compute resources across all points within the IT infrastructure.
Requirement 3: Security for Local Operations
One estimate states that the number of access points through IoT will grow from over 8 billion devices in 2017 to over 25 billion devices by 2021*. These devices span highly sensitive areas where tight security is essential such as in retail, banking, military, as well as other verticals. Each additional access point increases the surface area of attack and presents an additional opportunity for security intrusion. Edge computing demands compute nodes be designed from the ground up with integrated security in mind.
Announcing Two New Server Solutions for the Edge
Today we’re announcing two new exciting PowerEdge servers uniquely suited for this new era of edge computing. The PowerEdge R6415 and R7415 exceed expectations in meeting the key requirements for edge computing. Both servers drive fast decision making with up to 2TB of in-line memory and up to 32 cores. For fast data compression and cleansing, both systems have up to 128 PCIe lanes enabling abundant amounts of low-latency storage. The 1U R6415 supports up to 10 NVMe drives and the 2U R7425 can provide up to 24 NVMe drives! Efficiency is an important criteria for edge computing and both of these servers are ideal – delivering outstanding TCO. Both are single socket servers yet are still able to support large amounts of storage, memory, and cores. With software such as vSAN licensing priced per socket, having single socket servers with rich storage, memory, and core counts provides the balance of performance and optimal cost for the life of the investment the edge demands.
Finally when it comes to security – the R6415 and the R7415 offer the outstanding security capabilities found in all of our 14th generation PowerEdge servers. With the PowerEdge, security is not an afterthought but is built in and integrated into the entire solution. This includes a number of industry-leading features such as System Lockdown, drift detection, Boot Guard, BIOS recovery, Rapid OS Recovery, Silicon Root of Trust, and System Erase. PowerEdge servers are the only servers with an end-to-end security ecosystem and lifecycle – absolutely essential for any edge deployment.
At Dell EMC, we’ve seen these three critical requirements consistently with our customers. We are really excited about these two new server solution offerings as an excellent combination of technologies to meet the challenge of the edge. For more information on compute at the edge listen to this conversation I had with Scott Aylor from AMD for our Luminaries podcast, where we discuss this emerging trend in depth. For more information on today’s launch of these new PowerEdge servers, check out this blog by Ashley Gorakhpurwalla, President and General Manager, Server Solutions Group.
Join the conversation with me on Twitter at @Ravi_Pendekanti
* Forecast Analysis: Internet of Things — Endpoints, Worldwide, 2017 Update, Dec 2017, Gartner
**Business Insider, Edge Computing in the IoT: Forecasts, key benefits, and top industries adopting an analytics model that improves processing and cuts costs, Oct 2016