Cloud computing: a way to use the Internet anytime, anywhere, on demand, convenient to use a Shared computing facilities, storage devices, applications, and so on the calculation mode of resource

Cloud computing - a bigger, faster, stronger

To deploy applications to the cloud, can no longer have to focus on those who headache problem of hardware and software, they will be to solve by the cloud service provider of professional team, using the share of hardware, this means that like to use a tool to use cloud services (like plugged in, you can use electricity as simple), just as you need to pay for the cost of the corresponding, and about software updates, resources on-demand extensions can be done automatically,

Edge of computing

Unlike traditional centralized thinking, his main computing nodes distributed deployment and application in data center close to the terminal, which makes the service response performance, or reliability is higher than traditional centralized cloud computing concept,

So, margin calculation and what is the difference between cloud computing?

Actually if cloud computing is a centralized data processing, edge calculation can be understood as edge type large data processing, but different is, only this time, the data need not spread to distant clouds, in the edge of the side can solve,
Edge is more suitable for real-time data analysis and intelligent processing, compared with pure cloud computing is more efficient and safe!
Edge computing and cloud computing are actually both deal with large data calculation run a way,

Edge to calculate more accurately is supposed to be a supplement of cloud computing and optimization.

In short: cloud computing to grasp the overall and edge based on local
focus

Cloud computing problems

Now cloud computing are centralized, namely the server concentrated in one place, in order to use the cloud computing resources, data needs to be transmitted to a long way from the user's data center and then focus on, but many devices can access the cloud, is roughly the following two reasons:

Data: the huge amount of data, the transmission bandwidth cost is hard to accept; Such as general electric have long been aware that industrial sensors on the machine tool to produce a large amount of data need to be on the Edge of the equipment for processing, only will have the most valuable data to the cloud for machine learning and sharing between different equipment (Edge computing could push the cloud to the fringe)
Speed: for low latency, intensive computing intelligent devices, such as a head-mounted VR, robots, unmanned aerial vehicles (uavs), such as limited by network transmission delay and can't enjoy powerful computing resources to cloud computing, these devices also face a common problem, battery life is short;
Edge of the concept of computing is put forward in order to solve this problem, on the edge of computing, sensor, controller and other connection device itself to collect data and analyze the Internet of things, or sending them to nearby computing devices (such as server or laptop) is analyzed, when the data processing and analysis on the edge of the network (as opposed to a data center or cloud), the data can be analyzed immediately and put into operation,

The characteristics of the edge of computing

What are the typical features of edge computing?

Edge in terms of computing has the following traits

1 near real-time data processing: because the data is in the edge nodes were analyzed, and the lower latency, improve the response speed of application

2 reduce data transmission, data do not need to push to the distant clouds and reduce the intelligent devices and data center, the amount of data transferred, save bandwidth costs, at the same time also can reduce the core network congestion, such as facebook and other social software users to upload photos on the edge of adjustment to the resolution of the right and then uploaded to the cloud

3 data security: some sensitive data directly on the edge, don't have to be careful data leakage

4 improve availability: share the offload the center server computing tasks, which eliminated the major bottleneck to a certain extent, and reduces the a single point of failure has the following characteristics may be calculated