On the morning of July 3rd, at the beginning of 2019 Baidu AI developers conference, the opening speech was given by Baidu CEO Robin Li. During the speech, suddenly one audience rushed to the stage and poured a bottle of water to Robin Li's head. Robin Li was firstly shocked, then calmed down to continue to speak. He pointed out after this event, in the way forward of AI, there will be a variety of things happen, but the determination of people to advance will not change. Eventually, AI will change everyone's life.
Indeed, in the past decade or so, the development of IT technology has encountered various difficulties, and of course, unprecedented innovations. With the development of Cloud Computing, AI and Internet of Things, hundreds of millions of mobile devices are connected to the Internet. These devices generate a lot of data every day. As the pressure of computing increases, computing begins to shift from the core area to the edge device, so the edge computing technology begins to appear.
We take the example of a chef making bread. In the early stage of the traditional computing model, the calculation is done in the CPU core, which is like a chef making bread, all the processes are done by the chef alone in the kitchen. But with the increase of demand, one chef cannot meet the needs of all customers, so we recruit an additional chef, which is what we call multi-core. Next, if it's still not enough, we'll open another branch and recruit more chefs. That's what our servers call Duplex or Multiplex.
But when the big data comes, simply adding more branches and chefs will not meet the demand, so we equip each Delivery man with a portable microwave oven, which only needs simple production in the shop, and the cooking process is completed on the Delivery way, which greatly saves the work of the chef. Such a workflow is like Edge Computing.
Requirements for Data processing using Edge Computing are becoming more and more. For example, Auto-pilot cars need to process data quickly so that AI can make quick judgments and avoid accidents. Many new data computing requirements have also brought up new technology/business models, which is the background of the explosion of edge computing in recent years.
At the same time, the development of 5G technology will further boost the unprecedented development of Edge Computing market. From smartphones to wearable devices, from medical treatment to automobile and industrial manufacturing, Edge Computing will help to promote the development of AI and other modern technologies.
To support Edge Computing in 5G era, HSN innovatively introduces the Edge Node concept. As a vital supplement to Miner Nodes, Edge Node mechanism takes over the intensive computing services. It helps reduce the response delay and bandwidth cost, and helps meet the needs of various smart scenarios under the decentralized architecture model.
As Edge Nodes are the source of big computing and mass storage resources in the HSN network, in the future all terminal devices with certain computing or storage capacity can be Edge Nodes. By securing Edge Nodes with Miner Nodes, HSN ensures that Edge
Nodes can provide efficient, reliable, and credible blockchain network services for big data storage and ultra-high-speed smart contract edge computing.
Time-consuming computing is performed by combining the idle computing power and storage capacity of numerous Edge Nodes into a distributed computing and storage platform. Edge node usage scenarios include AI applications, image processing, gene sequencing etc. After loading the intensive computing tasks off the cloud to the Edge, power consumption of the entire system is reduced by over 40%, and time of data integration, migration and other operations can be cut by over 90%.
At the same time, HSN network design conforms to the edge computing architecture. It can make full use of the node’s computing power, is about to meet the docking requirements for IoT computing and storage devices and improves timeliness of the perception-computation-response process in the IoT. Among the IoT applications, many scenarios require low-latency response, which makes cloud computing unprofitable. But, HSN’s Edge Computing model provides a new, efficient solution.