What is edge computing, and what are the use cases?
Nov 2, 2020
At Edgegap, we’re often asked about edge computing by those that are unfamiliar with the term. Once that’s explained, the next question is what can you do with it? Here’s a quick intro to edge computing and a look at what industries it can benefit today, as well as in the future.
How do you define it?
Edge computing is the ability to store and compute data close to its source of origin. This is sometimes evoked as opposed or at least complementary to cloud computing, which is the ability to store and compute data usually far from its source of origin, inside distant datacenters. The benefits of edge computing can be cost-savings, speed (latency improvement), security, and more.
A point to note is that it is theorized that the true value of edge computing will come from use cases that are currently unknown, since this disruptive new technology will bring innovative new uses in our daily lives.
What will be the impact of 5G, AI and the cloud?
5G: Edge computing can be deployed along with 5G antennas to enable computing at the base of the data reception tower. Many Telco companies are very interested in this new technology, especially since they were more or less left behind in the cloud revolution. Data can also be transmitted wirelessly this way, from point of origin to the edge compute, much faster and reliably than before.
AI: AI computes tasks that can now be handled closer to the origin of the data, saving costs for the transfer of data and allowing real-time analysis and interaction.
Cloud: The final impact is still unknown, but the edge is a disruptor for the cloud. It has the potential to woo some customers away from the big cloud providers, but since their value proposition is different, might not have a huge impact on their bottom line. With their expertise, cloud providers are also very interested in developing their own edge and have been in the process of developing their own offering in that sense.
See our post about Edgegap’s integration with AWS Wavelength:
What issues is edge computing facing?
The main issue for the development of edge computing is the chicken-or-the-egg problem:
Without a commercially viable use case, no company is willing to spend the very high amount of capital needed to develop a highly distributed edge computing infrastructure.
Without a highly distributed edge computing infrastructure, many use cases cannot give a good enough value proposition to be commercially viable.
What are some of the use cases of edge computing today?
IoT is gathering interest from many companies, as data from sensors is analyzed locally for a faster and/or cheaper way to collect and analyze data. This is usually very local or at best regional in use, thus not solving the chicken or the egg problem worldwide.
Gaming is a very interesting prospect since it is very latency-dependent. Many sub-categories of gaming can find a benefit in edge computing: Multiplayer game server hosting, Cloud gaming, AR and VR processing, etc.
Other use cases in different industries are currently hypothesized but are longer-term propositions because of the infrastructure changes needed to make them possible: smart city applications, autonomous vehicles, AI virtual assistants, etc.
How can my company use edge computing?
First, there should be a business analysis to find out if handling data closer to the point of origin is either cost-saving or value-adding for your company. For companies that are currently storing and analyzing data in the cloud, there is a good chance that will be the case. Once that correlation has been made, you can either:
Build your own edge, called on-premise edge. This is expensive and complex from a hardware and software perspective, so few companies have the capital and skills needed to build it on their own.
Work with edge vendors that have access to a distributed edge network. You will need to rent servers for a minimum amount of time, 1 month or sometimes 1 year, in all relevant locations, which can also be quite an investment. There are currently few edge vendors that provide worldwide coverage.
Edge IaaS companies. These are companies that have built a solution to provide on-demand infrastructure on a highly distributed network, usually composed of an aggregation of edge and cloud infrastructure. OneEdge, Edgegap’s highly distributed edge network, is such a solution.
Edge computing will be a disruptor in many industries and will create new use cases as it is being developed. The current applications are limited because of the limited distribution of edge infrastructure, and only a few use cases can currently see real benefits such as IoT and Gaming. The true potential of edge computing however will be unlocked once the “chicken-or-the-egg” problem is solved, and it becomes commercially relevant to build highly distributed edge computing infrastructures worldwide. Over time, as more and more technologies become latency-dependent due to the nature of our speed-obsessed society, it might even end up competing with the cloud in terms of usage for the processing and storage of data.