Latency-sensitive 5G use cases require edge computing investment
If you surveyed the greater telecoms industry, you’d likely get a long list of definitions of what (or where) exactly the edge of the network is; similarly, you’d probably return with a litany of potential deployment strategies–operator-led, enterprise-led, a neutral host model, etc…But despite the lack of clarity, there is consensus around one thing–many monetizable 5G use cases, particularly those that require low latency, require an investment in edge computing infrastructure.
Dell Technologies has a long history of innovation in providing data center equipment and services to enterprises of all sorts, including a range of global communications service providers. In a series of product announcements, it’s clear that Dell sees its data center expertise and existing customer relationships extending out of centralized data centers and out to central offices, enterprise campuses and even cell sites, all potential locations for edge computing infrastructure.
Dell’s Jeff Boudreau, president of the Infrastructure Solutions Group, characterized the 2020s as the “data decade” and said in a statement, “The challenge moves from keeping pace with volumes of data to gaining valuable insights from the many types of data and touchpoints across various edge locations to core data centers and public clouds.”
Dell’s edge products, likely set to be made public around the now-cancelled Mobile World Congress so as to garner telco attention, include:
- Dell EMC PowerEdge XE2420 is a ruggedized “short-depth” server designed to fit more easily into the challenging physical environments that are part of the edge.
- Dell EMC Modular Data Center Micro 415 is tailored to fit customer requirements but includes IT, power, cooling and remote management. Dell said it’s smaller than a parking space.
- Dell EMC iDRAC9 Datacenter software is a remotely deployable tool for analyzing streaming data.
- Dell EMC Streaming Data Platform is a scalable tool for data ingestion and storage.
OK, so what does all that mean? To give a hypothetical example, consider a manufacturing line. As products are finalized, prior to shipping, there’s a quality assurance step. Instead of a costly, time-consuming manual inspection, cameras can view the product, compare it against examples of defective units, and use artificial intelligence to either approve or kick out whatever is being produced.
In order to do that, you’d need servers to run the compute workloads and those servers would need to be close enough to where the data is produced, the camera, to provide real-time data analysis and feedback. Once you’ve got the servers running the AI tools, you need a platform to handle and analyze the data. And that’s edge computing.