The Rise in Edge Computing is Bringing Network Accessibility and Data Analytics Closer to Home
Living on the Edge.
Everything we do these days generates more data — and therefore, requires more computational power. Shopping via cell phone, posting on Facebook, streaming Netflix, smart appliances, and wearable technologies have all driven data computation to the edge. The edge data center, that is.
According to a report by MarketsandMarkets, the edge computing market is expected to grow from last year’s $1.47 billion to a $6.72 billion industry by 2022, an annual growth of more than 35% over the next five years. Tasked with offloading latent-dependent services from large, enterprise data centers across the country, edge data centers bring the computing power closer to the user, often in major metropolitan areas. This benefits both the enduser and those collecting the data in many ways. It’s no wonder edge data centers were named a Top 10 infrastructure trend by Gartner this year.
Here’s five reasons why edge data centers may be for you.
Power to the people. Bringing more computational power to the “edge” of the network, aka, closer to the enduser, means increased speed and better performance when downloading, sharing, or streaming content at their current location. Businesses also benefit. Whether a distribution warehouse or a growing inventory of online shipping data is being gathered, it’s processed in the edge data center quickly and then the information that is needed is sent to the enterprise facility for storage and further analysis.
Lower cost. Edge data centers rely on the business’ production facility to carry most of a company’s work load, and as a result, the edge data center can be quite a bit smaller and not as operationally intensive. Because they are smaller, often just a single server or several megawatt’s for larger companies, they’re often housed in a colocation data center (see #4) alongside other business’ edge data centers in a shared environment where operation costs are spread out amongst the tenants. Additionally, edge data centers can help reduce connectivity costs by filtering the data for relevance prior to sending it over the network. Only what needs to be sent is sent and analyzed. Alternatively, if all data was sent from a device over the network, the sheer cost would outweigh any benefit of tracking and controlling a device.
Internet of Things (IoT). As everyday items become more connected, more compute needs to be added within a close enough range to avoid latency issues. Many of these connected devices, including TVs, thermostats, security monitoring sensors, etc., are not meant to store or process data to maintain lower power profiles and reduce the overall cost of deployment.
Every user is different. Because every business will have different infrastructure requirements, the edge data center tries to be as flexible and scalable as possible. Typically, providing Tier III, N+1 infrastructure reliability at minimum, edge data center tenants can pick and choose their requirements based on need. Some tenants will require office space, others will have more demanding network needs, and some may require higher electrical and cooling densities. For example, some edge data centers will require direct connectivity to Amazon Web Services (AWS) or Microsoft Azure to access data in the cloud.
Speed to market. Edge data centers are seeing steady demand because they can deploy quickly within key areas like downtown urban environments, thanks to colocation (co-lo) data centers. Across the country in metropolitan markets, co-lo data centers are springing up in old warehouses or printing press facilities, and former telecom buildings are being converted to data centers.
THE NEXT FRONTIER
The edge data center is an extension of the larger, remote enterprise data center. It can reduce costs, deliver data faster to the enduser, and collect more data in real-time. It will provide endusers with a better user experience when gaming and online shopping, let businesses collect and track more data with the IoT, and help driverless cars navigate their path because of their proximity to the enduser.
Today, the world is connecting to the IoT through billions of devices. With the advent of 5G connectivity, more and more data will be collected due to faster networks. This will give way to even more applications for tracking data through our daily process of driving, manufacturing, healthcare, life-safety, cyber security, and more.
Tomorrow, it’ll be trillions. Living on the edge — or bringing data computation physically closer to the population — is a must for businesses that want to successfully transition tomorrow’s IoT users. n
While the term “edge” is relatively new (3+ years), the concept is not. Google, for example, has used this model to reach the populous for over eight years, developing smaller data centers within urban environments. Originally called “caching” data centers, the concept was to develop small (500kW) data centers to act as “RAM” for faster compute. One of the key elements of the original cache data centers was to develop data centers near central offices, or COs. The term “edge” is primarily the same concept: latency and bandwidth are most important.