Learn the structure and how the Edge Computing model works

Home  /  Blog

thumb Edge Computing


As demand for the Internet increases, more bandwidth means running a data center costs a lot. Edge Computing has become a trend for popular IoT solutions in recent years. So how is it structured and how does it work? Let’s find out with Adtech through the following article.

Learn the Edge Computing model

With the growth of the Internet of Things (IoT), billions of devices are generating huge amounts of data. But to store or analyze all that data in real time is nearly impossible. This is where edge computing becomes relevant.

Edge Computing Model
Edge Computing model

With edge computing, we process data closer to the source, such as an IoT device, an IoT gateway, an edge server, a user’s smartphone or computer. The idea is to move the intelligence to the edge and let the edge nodes/devices do the real-time analysis.

This reduces application latency and saves network bandwidth. The architecture is distributed, as opposed to centralizing all processing in the cloud. It minimizes long distance client-server communication.

In many cases the edge can be a local workstation that is connected to the Internet but still located at the edge of a LAN, but sometimes the edge is a local data center that handles processing for an area. geography.

In general, the edge does not have to be on the LAN or connected over the Internet, as long as it is closest to the data source and is available to provide real-time low-latency processing.

The structure of the edge computing model

To better understand and study this model in depth, you need to understand the components and their roles. Here are 5 basic components that make up the edge computing model:

  • Cloud Server: can be a public or private cloud, or a data center. These clouds host and run applications used to coordinate and manage the various edge nodes. Edge workloads, endpoints, and clouds interact with each other during processing.
  • Edge device: is a device with integrated computing capabilities similar to ATMs, cars, digital cameras. Edge devices often have limited computing power, handling only transient requests that require low latency.
  • Edge node: An edge node is a generic term for any edge device, edge server, or gateway on which edge computing can be performed.
  • Edge server: An edge server is a general-purpose computer located in a remote facility such as a factory, retail store, hotel, distribution center, or bank branch. . Edge servers are typically built using industrial PCs with CPUs with 8 – 16 cores or more, 16GB of memory, and hundreds of GB of local storage. An edge server is typically used to run enterprise application workloads and shared services.
  • Edge gateway: an edge gateway is typically an edge server that, in addition to handling enterprise application workloads, performs network functions such as protocol compilation, firewall protection, or connectivity wireless.
Components That Make Up The Edge Computing Model
Components that make up the edge computing model

How edge computing works

To imagine how Edge Computing works in practice, smart devices are the best example.

Imagine that an IoT sensor can generate millions and tons of data a second, the data is instantly transferred to the cloud database, the center where it is processed and stored.

The Fast And Accurate Operation Process Makes Edge Computing Widely Used
The fast and accurate operation process makes edge computing widely used

When there are requests, the server will give a response back to the device upon receipt, perform analysis of the received data, etc. The whole process takes less than 1s to complete, but there are still issues. affecting the stage such as: slow, weak network connection, data center located far from equipment, etc.

Meanwhile, for Edge Computing, you don’t need to send data collected from IoT sensors. The nearest device or network node will be responsible for automatically processing the data and then responding.

Sensors, devices deployed remotely require real-time. Electromagnetic cloud systems tend to be slow when faced with this reality, especially when decision-making needs to be done in microseconds.

What hardware is available to implement edge computing?

What hardware is available to implement edge computing?

For the edge, RISC processors such as ARM, ARC, Tensilica, and MIPS are preferred over CISC. While ARM Cortex is suitable, ARM also offers Neoverse specifically for edge use cases. ARM Cortex-M55 and Ethos-U55 are AI edge computing chips.

Commonly Used Edge Computing Implementation Hardware
Commonly used edge computing implementation hardware

NVIDIA Jetson GPUs are designed for the edge. For example, the Jetson Nano is a 128 core GPU. Jetson TX2 and Jetson Xavier are intended for industrial and robotic use cases. There is also the NVIDIA EGX Platform that provides GPU edge servers.

Intel has Movidius and Myriad 2. Myriad 2 is also part of Intel’s Neural Compute Stick (NCS) that draws power from the host device via USB.

Mainflux Labs provides the MFX-1 IoT Edge Gateway. Huawei’s AI Atlas computing platform includes AI accelerators, edge stations, and servers based on Huawei’s Ascend AI processor.

Scale computing powers the HC3 Edge and HE500 systems. APC’s EcoStruxure micro data center promises physical security, standardized deployment, and remote cloud-based monitoring. There’s more of a focus on the micro data center.

Edge computing is still receiving a lot of attention, because it is suitable for computations that need large processing power. With IoT technologies still in their infancy, the development of IoT devices will also have an impact on the future development of edge computing.

Avatar Of Admin



Leave a Reply

Your email address will not be published. Required fields are marked *