Why do we need edge computing

With the rise of 5G, the Internet of Things and the popularity of cloud services, the emerging new applications have brought a huge test to the computing capacity and network bandwidth service capacity of the Internet.

Why do you say that?

Consider some numbers: Gartner predicts that 14.2 billion connected objects will be used in 2019, and the total will reach 25 billion by 2021. And each of these devices will generate an enormous amount of data. Take mobile phones as an example, we can think about their monthly use of mobile phones (including WiFi, do not have to pay the flow is also the flow 🙄) will generate how much traffic?

Xiaobian mobile phone traffic in 24h

So how many people in the world are using mobile phones now? How much data is going to be generated all the time without interruption?

But that’s just phones! The rise of 5G and the Internet of Things has turned every object we touch and see into a potential data producer.

According to IDC’s data Age 2025 report, the global annual data generation will grow from 33 zabytes in 2018 to 175 zabytes, or 491 exabytes of data per day. (If you’re not sure about this number, use the following figure to convert it.)

What should we do when we face the disadvantages of high latency and low efficiency of computing power and insufficient bandwidth of backbone network? After all, the vast amount of data is here, and it can’t just be thrown away.

Therefore, edge computing appears as a new computing mode (or computing logic) to meet the application requirements of low delay, large bandwidth and large connection.

How is edge computing implemented?

It’s different from uploading data to the cloud and doing calculations. Edge computing disperses large services originally processed by the central node to edge nodes closer to the user terminal, thus effectively improving data processing speed and transmission speed and reducing delay.

As a result, edge computing covers almost every data-generating scenario in our lives. Such as smart cities, smart transportation, smart retail and video surveillance. Actively promote the traditional “cloud to end” evolution into “cloud – edge – end” emerging computing architecture.

So how does edge computing work?

Let’s look at its main architecture: core infrastructure, edge computing center, edge network, edge device composition.

Core infrastructure

Edge computing center: it mainly provides computing, storage and network forwarding resources, and is one of the core components of the whole “cloud-edge-end collaboration” architecture.

Edge network: it integrates various communication networks to realize the interconnection among edge devices, edge servers and core facilities.

Edge devices: Data consumers and producers in various scenarios.

Enterprise edge computing layout

At present, in order to develop the layout of edge, many enterprises need to purchase edge nodes one by one from operators at all levels and build their own edge infrastructure. However, problems and challenges also come one after another:

Poor elastic expansion: it is difficult to obtain node resources, and the delivery period of new nodes is long. 2. Difficult O&M: A multi-node O&M management platform is developed, which depends on the service capabilities of different operators. High investment cost: a large number of IDC cabinets, bandwidth, IP, servers and other procurement costs

In order to solve the problem of self-built edge infrastructure, JD Zhilian cloud extends its core capabilities to edge computing center by utilizing the accumulated IDC resources, technologies and operation and maintenance service capabilities of core infrastructure, and launches edge computing service in the form of edge cloud: edge physical computing service.

Edge physical computing service provides high-quality edge node resources, sinks computing, network, and storage capabilities to the edge of the network, and provides low-latency, high-bandwidth, high-real-time, localized computing and storage services close to customers, as well as on-site operation and maintenance services. Customized deployment can be made according to customer requirements.

Customers purchase edge physical computing services from JD Zhilian Cloud, which can greatly reduce the cost of self-built computer rooms. The investment in the early stage of computer room construction is “paid” by JD Cloud, and customers only need to choose business nodes to use as needed.

Related solutions

Edge computing has been widely applied in practical business scenarios. Edge physical computing service is committed to creating edge resources of multiple communication operators, and sinking business, content and service to edge nodes for nearby processing, so as to bring higher performance or lower cost business results for customers. In order to meet the needs of customers in different scenarios, JD Zhilian cloud combines all kinds of edge computing products to provide solutions for different application scenarios.

  • Network and iot device edge

The Internet of Things is the Internet, the traditional telecommunications network and other information carriers, so that all ordinary objects can perform independent functions to achieve interconnection network. Iot devices need to be more agile in connection and more efficient in data processing. By deploying edge nodes close to users, they can help master data networks easily process user data, provide timely responses and support intelligent computing of iot devices.

Scenario Features: Supports IoT protocol, privatized deployment, and data desensitization.

Scenario example: Intelligent factory, intelligent healthcare, and industrial machine management

  • Edge of Artificial Intelligence

At present, most artificial intelligence computing is supported by cloud data centers, which will easily cause network congestion. If edge nodes are deployed to build high-performance computing platforms, data can be calculated locally, and real-time environment awareness, human-computer interaction, and decision control can be carried out without networking.

Scenario features: Edge deployment AI model, AI model protection, AI hardware acceleration

Scenario example: Face recognition, image audit, new retail control, intelligent driving, etc

Currently, JD Zhilian cloud [edge physical computing services] is in public beta!

Click on the”reading”Can apply!