Recent Innovations and Enhancement in Data Computing Platform
Published on : Monday 06-12-2021
Organisations may prefer edge computing for critical applications because smart devices with computation power are connected on the edge of the network, says Jasbir Singh.
Edge and cloud computing are both different and non-interchangeable technologies. They can’t replace each other during computation in operation. The main difference between cloud computing vis-à-vis edge computing is worth looking at as to how data processing could take place. Most of the data processing is performed within the cloud, using a large number of connected IoT devices with a series of centralised servers at multiple locations. Edge computing takes a completely different approach where the processing is away from the centralised servers, and closer to the end users.
Understanding the characteristic and capability
Cloud computing is used to process data, which has no effect on time-latency. Edge computing is preferably used to process time-sensitive data and also widely preferred in remote locations, where there is limited connectivity to a centralised server. Edge computing is computing complete work and delivery of information near the source of the data, whereas for cloud computing it needs to select one from many data centres to do all the work.
The amount of data required to process every second is sometimes not adequately supported by cloud computing. There is a lot of computation/calculations that cloud computing does not provide to cloud-based applications in connected servers. There are two main problems that surfaced during the processing stage – latency in processing and high number of unwanted resources. These issues largely exist in decentralised data centres, cloudlets and mobile edge nodes.
Computing potential of each platform
Edge computing supports companies by turning computed data into competitive information. Edge computing is mainly beneficial to specialise and intelligent machines connected with device servers which are not like regular computing devices like PCs designed to perform multiple specialised functions. This specialised computing is required to respond to particular machines in a specific way. Edge computing is considered for corporate computing, close enough to data centres to cut down latency. It develops the raw power and analytics if allowed to be used as cloud computing selectively in future, although on smaller scales.
Edge computing filter and compute sensitive data at the source and do not send it to the central data centre. Most of the time edge computing is preferred over cloud computing when it requires computation for critical applications in remote locations. Limited or no transfer of sensitive information between devices and the cloud means better overall security. For speed, quantum computing solves problems by carrying out multiple calculations simultaneously and Edge computing comes to the rescue here by enabling data analysis closer to source. By increasing use of artificial intelligence, it adds existing technology with computer networking and data storage.
Limitations
Edge computing is a solution for the challenges faced by organisations but it does not handle all applications, thus, cloud computing will be needed for some functions in many applications to keep cost of operation in a manageable limit.
The critical machines shall no longer depend on cloud loop to provide emergency response due to its operation with edge computing but still works co-functional with cloud computing to manage the IoT devices remotely. Cloud computing remains relevant and works alongside edge computing to provide real-time solutions.
Innovation and way-forward
Edge systems are scaled up if connected with input from devices preferably cheap low-power sensors. Latency, a prime motivator for the development and acceptance of edge computing systems, is already growing in demand by connecting billions of nano-IOT devices, where 5G will not by itself increase the speed of computation.
Nano-materials devices in development, powered with artificial intelligence, connected to a secure network of billions of devices, are advancing the efficient edge computation to make it complete, energy-efficient with ultrahigh performance.
Good way forward for improving edge computing would be, by replacing the inefficient materials used in chips for computing today. The normally available smaller silicon-based diodes/transistors used in chips, experiences conductivity losses and leads to energy loss by generating heat. Using silicon-based elements of carbon nanotubes, by having more efficient electron transport properties demand less energy requirements to function.
Evolution to meet our future need
Search for the elements having more efficient electron transport properties and less energy requirements are the basic need for the development of chips for the future electronic circuits. The natural or man-made materials, which can carry electricity and light without resistance and backscattering are more in demand for the development of nano and efficient devices. Nano systems are in demand for reducing energy need to function and reduce waste heat in the case of electricity requirement to operate.
Materials and technology in support of system advancement
Scientists in Japan have theorised that graphene could function as a superconductor. A team of Japanese researchers managed to make electricity flow through graphene with no resistance in 2016. However, scientists found that it is possible to implement only in an extremely cold environment and by this its application is limiting the experiment. The superconductivity could only be achieved by doping, or by placing the graphene on a superconducting material. Both of them alter the graphene property and compromise the experimental results too.
Researchers at the University of Cambridge found that, by coupling graphene with praseodymium cerium copper oxide (PCCO), graphene itself becomes superconductive. PCCO is an oxide from a wider class of superconducting materials called ‘cuprates’. It implies that PCCO was the material needed to trigger graphene’s inherent superconductivity. Graphene has an incredibly durable material in which electricity can flow easily. Graphene has the potential to be used in a new era in electronics, for more efficient, high-speed devices in nano-sized graphene circuitry.
Superconductivity
Superconductors are those materials that have exactly zero resistance and infinite conductance, because they can have V = 0 and I ≠ 0. This also means there is no Joule heating, or in other words no dissipation of electrical energy as explained well in Wikipedia. These could be used to replace silicon-based transistors, for more efficient and speedier microchips and smart sensors.
Quantum networks and dynamic of edge computing
In quantum computing, instructions can be given in both 0 and 1 simultaneously and calculating information are produced faster for certain tasks. These devices could work more efficiently under extreme traffic loads used to uplift edge-computing networks better performed. Scientists have demonstrated that quantum computing could move information rapidly through the network and vis-à-vis improve edge computing for rapid calculation and produce better results. Quantum effects are employed to transmit data instantly, which can have an enormous impact on the ability to process data on a network from connected nano-devices at remote edge nodes in the same network/facility. These nodes establish a link at peer-to-peer level to operate for quantum applications. Thus, nano-device systems and edge computing may become an inseparable entity in the future, where devices and computing functions interact dynamically.
For edge computing, there are three ways the technology can be engaged:
Fog computing: data is distributed between a centralised computing infrastructure and connected devices.
Mobile edge computing: In the MEC architecture, the cloud’s computational and storage capacities are brought closer to the end-users’ mobile networks.
Cloudlet computing: an infrastructure which uses smaller data centres for offloading data means bringing the cloud closer to the end-users.
Organisations may prefer edge computing for critical applications because smart devices with computation power are connected on the edge of the network. The device itself monitors a setup of machines, if pre-defined metrics given with tolerance levels, and if the results are outside of the prescribed tolerance, a warning signal is generated as soon as the signal reaches to the failure level, resulting in the initiation of shutdown signal to stop the machine within microseconds to avoid losses.
Types of edge computing
Cloud edge: Public cloud is protracted to a series of multiple point-of-presence (PoP) locations.
Device edge: Software runs within the hardware (device and server).
The cloud edge is the extended form similar to the traditional cloud computation, where the cloud provider is responsible for the upgradation, seamless working and maintenance of the entire system. Device edge computes primarily within the hardware itself, to process real-time data in a manner that it becomes very swift and accurate. The main difference and selection between the two lie as per their deployment procedure and the way they are priced to customers. For edge computing the maintenance and time bound upgradation of centralised servers for data computation and storage becomes a challenge for many organisations who have limited system resources, setup and skilled manpower.
Jasbir Singh is an Automation Expert having long experience in Factory Automation, Line Automation, Implementation Strategist, Business Coach, Regular writer on automation, Artificial Intelligence, Robots/Cobots, Digital Technology, Network Communication, Industrial Internet of Things (IIoT), Wireless Communication, Block Chain and use of advance digital technologies. He has established a long association with Business Houses/large production houses to improve factory automation in their production lines as well as productivity improvement in factories in India and overseas; and in advising and designing the units to transform into digital platforms by use of Artificial Intelligence. Email: [email protected]
