The emergence of technologies such as artificial intelligence, machine learning and blockchain has exponentially multiplied the volume of data used, putting current computer systems to the test. Despite the fact that the arrival of 5G promises an improvement in these limitations, it is not enough to respond to the performance demands of this latest generation technology. An alternative to this is edge computing.
We live surrounded by smart devices. However, they need connectivity to transmit the data that allow us to access our email, enjoy content streaming or automate a factory through IoT (Internet of Things) techniques. These are all connected to a network that has to respond immediately to that data demand, but the reality is that they are centralized systems that have limitations such as latency, bandwidth, and even data privacy and autonomy.
The deployment of 5G technology promises greater speed in the circulation of all this information, as well as a drastic reduction in latency. But the new network is not enough by itself; it needs to be complemented by edge computing, a distributed computing paradigm that addresses the limitations of centralized systems by bringing data processing closer to their generation source, in other words, to devices and users.
"Processing the data closer to where it is generated means processing information faster and having the ability to make decisions more quickly," says Matías Díaz, BBVA´s Global Head of Edge Security in Security Architecture.
Edge computing: What does it consist of?
In centralized systems, data has to travel from the place where it is generated (the devices), to a central node for processing, and then return to its place of origin. It is a process that involves a huge amount of information and which consumes a lot of bandwidth and sometimes, when the data travels round-trip, it can cause latency that affects the proper functioning of the devices themselves.
Furthermore, with the emergence of technologies such as 5G, artificial intelligence, machine learning and blockchain, the volume of data used has multiplied exponentially. According to estimates by International Data Corporation (IDC), in 2025 there will be 41.6 billion devices connected to IoT that will generate 79.4 Zettabytes of data. This means that processing them will require a huge amount of joint computing power. But if you move to a distributed system, with intermediate nodes, large data centers will see their workload reduced, benefitting the processing of all those requests. For this reason edge computing is key.
"Alternatives such as edge computing that bring data processing closer to the place where it is generated, seek to find solutions to these problems, offering alternatives for making the operating model remain not only efficient, but also scalable and sustainable,” says Juan Carlos López, Professor of Computer Technology at the University of Castilla-La Mancha and member of the COIT Governing Board.
Thanks to edge computing, "the possibilities of the current centralized cloud model have increased and expanded, supporting the evolution and deployment of IoT devices and admitting innovative applications, thus offering a great evolution for digital businesses as a result," adds Ignacio Velilla, Spain´s Managing Director of Equinix, a multinational specialized in internet connection and data centers.
Greater investment of resources
A basic deployment of edge computing consists of a device that generates information and that requires information from other sensors or devices to modify its behavior or make decisions. “For this reason, a nearby infrastructure where all this data is stored and processed is necessary. This way each device can immediately access not only its data, but all the rest, to take advantage of the information generated, " Velilla explains.
For Díaz, it's necessary that this process be linked to an investment of resources. "Companies need to have the best communications networks along with optimal points of presence to implement our architecture in order to process data in real time using the best algorithms and techniques of machine learning and artificial intelligence," he explains .
This is something that the professor at the University of Castilla-La Mancha agrees with: “If we look at massive data processing, with its large number of transactions and use of more complex algorithms which are necessary for providing a better service with fast solutions and decision making in real time, it will require greater computing capacities and shorter response times, as well as greater flexibility. ”
The advantages of edge computing for banking
One of the basic capabilities that edge computing will offer banks is the ability to decentralize their computing model and reach their customer more directly. BBVA´s head of edge security believes that "by eliminating latencies and improving computing speed, we will be able to offer customized products, improve fraud detection systems and operate in financial markets in a decentralized manner, along with cost savings."
BBVA, which in 2019 became the first financial institution in Spain to deploy 5G technology at its headquarters, has been working on building an edge computing platform for some time. "With the Ether platform we have the basis to successfully face the new challenges that edge computing brings us. These pillars are globality, reusability, automation, resilience, and embedded security,” says Díaz.
According to Díaz, these last few months have shown us the need to interact with customers wherever they are. For this reason, he believes that "we must take advantage of edge computing to reach our clients in a more agile and secure way, and be able to interact with them as if we were face to face in order to offer them customized solutions and products.
Agility and security are two key factors in this new technology, which are also requirements to successfully bringing that personalization to users. “Today we have a new service architecture that combined with our global network, brings us very close to our customers. Our applications are connected to globally distributed nodes that deliver content and process part of our services in an agile and secure manner, " explains Díaz. Looking towards the future, he adds that "in the coming months we will evolve to a computing model based on edge computing which will facilitate the creation of innovative financial products. ”