Edge Computing in 3 minutes
Gartner said that Edge replacing Cloud is one of the most important trends in 2018. What is Edge Computing ? How can it replace Cloud ? Which population does it target ?
In this article, I will try to simplify technical terms and explain through examples. I recommend reading this if you do not know about Cloud computing.
What is Edge computing ?
The idea of Edge is to do processing near the data source.
That’s all. I think this sentence is sufficient to define it. All that will come after this, is to explain briefly the why, when and how.
Benefits of Edge
Logically, when we push processing close to the data source, we gain some benefits. The most important ones are the following :
- Speed :
It seems that data transmission is faster when its sources are in the same network.
“Yes, genius ! We could have the speed of light :)”
Ok, stay calm. We can consider this only when the time spent to do processing is shorter than the transmission duration. For instance, assume that processing your speech (to be analyzed) takes 2s in local gateway and both its compression and transmission take 3s. In this case, Edge Computing is faster. On the other hand, assume that the complex processing of a photo takes 20s in local gateway and both its compression and transmission take 2s. In this case, it is more reasonable to keep processing in the Cloud.
- Security :
It’s better because data is not traveling over a network.
“Yooopiii ! No more encryption, no more access control vulnerabilities.”
If I were you, I wouldn’t be in such a hurry to say that. Why ? Because some people believe that Edge devices (including IOT) are less secure for the reason that it’s complicated to install patches and updates. Therefore, security in Edge is a double-edged sword.
Edge computing can be considered for use cases where devices have poor connectivity. With a world of IOT, we are most likely in the case of poor networks.
Other use cases have to do with latency-sensitive information processing. This is ideal for situations when real-time is required.
The most telling use case is self-driving cars which run through location not covered by any network while requiring real-time processing.
How to implement it ?
There are some standards to implement Edge computing. The most famous is Fog Computing. The term was created by Cisco in 2014. Fog enables repeatable structure in the Edge computing concept. Therefore, enterprises can push computing out of centralized systems or clouds for better and more scalable performance. Please refer to the Cisco document for further details.
In most cases, Edge cannot do without Cloud. Cloud still plays a critical role in enabling new levels of performance where significant computing power is required to effectively manage vast data volumes from machines.
Edge Computing is becoming a key part of a digital industrial transformation mostly with the new equation of Industrial Internet of Things (IIOT).
Contrary to what some can believe, Edge is far from a threat to Cloud. It can be considered as a complementary brick to meet some needs.