Edge Data Center
October/November/December 2020 I 37
Latency, which is essentially the time it takes for data
to traverse a network to its destination, is measured in milliseconds
(ms). Due to the way data travels from switch
to switch to server and back again, latency is unavoidable
and varies based on the transmission medium, distance,
number of switch hops and overall efficiency of the infrastructure
and equipment. While 50 to 200 ms to ping
Google or 10 seconds to load a web page is something
most people are used to, there is no place for latency
when it comes to supporting next-generation 5G, internet
of things (IoT) and industrial internet of things (IIoT)
technologies like self-driving vehicles, smart traffic systems,
virtual and augmented reality, artificial intelligence
(AI), machine-to-machine (M2M) communication,
enhanced video streaming and online gaming.
A self-driving car needs to be able to respond virtually
instantaneously to smart traffic signals and road sensors,
and M2M devices rely on guaranteed response times and
real-time processing of data. National Highway Traffic
Safety Administration (NHTSA) studies indicate the average
reaction time of a human in traffic is around 1 to 2
seconds. For self-driving vehicles to be safe and reliable
and to gain consumer confidence, they require 5G’s 5 ms
latency, which is practically real time.
One part of the solution to reducing latency lies in
edge computing that brings IT and cloud-based services
closer to end users and devices, which is why there is rapid
growth in the deployment of edge data centers (EDCs)
closer to the source of the data being processed—anywhere
from the central office, 5G cell tower or local street
corner, to the enterprise office, manufacturing facility or
remote mining operation (Figure 1). According to Gartner,
the global market for EDCs is projected to triple over the
next five years with 75 percent of enterprise-generated
data expected to be created and processed outside
of traditional centralized data centers.
Cloud
Data Center
Internet of Things
Data Processing
(Storage, Capture, Reporting)
Data Cashing
(5G, E-commerce, Streaming)
Data Analytics
(Machine Learning, AI, M2M)
FIGURE 1: Edge data centers are located closer to users and devices to support emerging low-latency applications.