Global data generation and consumption has risen at an exponential rate over the course of the last decade, with Statista figures predicting that it will have reached a figure of around 181Zetabytes by 2025. Increasing use of cloud-based services, the advent of 5G communication and the ongoing roll-out of the Internet of Things (IoT) have been identified as three of the main drivers for continued growth.
Though there will still be a clear need for centralised data centre sites, increasing installation activity at the edge will also be necessary in the years ahead. There are several reasons for this, with markedly reduced latency levels being the main one, as the turnaround time for data to go to and from the source will be otherwise too great to make certain applications viable. Augmented reality is an obvious example. Another reason will be to reduce congestion, or else the data coming from countless connected IoT nodes could easily overwhelm the network. Ensuring that transmitted data is secure and that network operational costs are reduced are also key factors.
An edge-based approach will mean that data storage and processing resources are situated much closer to IoT nodes, mobile subscribers, and business users. This will provide them with more convenient access and significantly faster responsiveness, as well as enabling network operators to make the best use of their available assets and avoid any potential logjams.
Although all this sounds straightforward, there are certain issues that are making everything a little more complicated. The infrastructure’s component elements will come from numerous vendors and be optimised for different sets of use cases. This hardware diversity means that there is a heightening risk of interoperability, and that is something that simply cannot be allowed to happen. Also, there are question marks in relation to how, given the complexity of the resulting architecture, the data involved can be kept secure. Finally, the volume of IoT nodes and the scale of the networks needing to be attended to should not be overlooked.
Arm has come up with a way of addressing, at scale, the interoperability and security problems that hardware diversity in edge infrastructure currently poses. Its Project Cassini initiative was first announced back in late 2019. Working with various technology partners, the objective of this venture is to enable cloud-native experiences to be derived at the edge. It will facilitate the software development process so that work can be completed quickly and without excessive engineering resources needing to be assigned.
There are three core aspects to the initiative (as illustrated in Figure 2). These are as follows:
- The establishing of robust standards that software development work will need to adhere to – in order that any installed software works seamlessly across all the different forms of deployed hardware.
- The creation and subsequent utilisation of architecture-agnostic APIs in relation to security.
- Access to detailed use cases and open source reference solutions, developed in conjunction with ecosystem partners. These will cover a broad spectrum of different industry sectors and allow engineers to get a head start on their software development work.
EBV is set to play an integral role in the Project Cassini initiative, contributing to the expansive ecosystem that Arm is currently building around this. One of the major investments that EBV will be making is the founding of a centre of excellence to support those needing to gain compliance. There will be further announcements relating to this initiative during 2022.