By Mark Sokol, Chief Technology Officer, AtlasEdge
For all the hype around the emergence of a 3D, immersive internet, (think metaverse, augmented and virtual reality), there are signs that it could well deliver on its vast potential.
Citi forecasts that the total addressable market for the metaverse could be between $8 trillion and $13 trillion by 2030, with total users numbering around five billion by the end of the decade.  If delivered effectively, it has the potential to reshape the Internet and upend entire industries.
This sounds suitably impressive, but its success ultimately hinges on two critical factors: seamless user experience and affordability. Without these, the metaverse could fold, wiping away billions in capital expenditure in the process. If it is to deliver on its immense promise, then Edge computing has a vital role to play.
Making it work
The processing and rendering requirements of an “embodied internet” are staggering. Whether it be an operating theatre or a Foo Fighters concert, generating a dynamic virtual environment, which users can interact with in real time, could require anywhere between 2 to 5 Gbps (Gigabit/s) as a typical bandwidth requirement.  The technology demands near-zero latency, with any form of lag or jitter undermining the entire user experience.
Put simply, the metaverse will not work if network connectivity is poor. The transfer of information from “device” to router via Wi-Fi, from router to exchange, and from exchange to metaverse server, needs to take place in 25 milliseconds or less. Any deviation from this, even if minimal, will cause issues.
Dan Rabinovitsj emphasised this in his white paper, outlining how “making the metaverse a reality will require significant advancements in network latency, symmetrical bandwidth and overall speed of networks”. 
Edge computing will ultimately bear much of the workload for delivering this seamless user experience. By bringing critical processing capabilities closer to end-users and eliminating the need to transfer data between networks, latency is minimised and real-time interaction enabled.
We view composable edge, which is configured and deployed to optimise the minimal compute and storage resources available and supports highly distributed infrastructure deployments, as the principal way metaverse players will achieve the QOS (Quality of Service) that is integral to their success.
Processing and storing the data generated by metaverse applications at the edge of networks reduces overall traffic coming in and out of servers. This is in turn frees up vital bandwidth for other potentially more critical tasks and ensures networks are not overwhelmed.
Remote processing via a composable Edge architecture also allows for vast amounts of user data to be analysed in real time, a key benefit for a decentralised, sprawling virtual world that will ultimately rely on real-time, performance insights.
Making it accessible
The success of large platform companies has been predicated on their services and products being low-cost. If the companies developing the metaverse are to match this success and reach a mass global audience, then their offerings need to be similarly affordable.
Again, the principal challenge here centres around the need to smoothly render a synchronous virtual or augmented world featuring trillions of objects. The easiest way to achieve this is by housing a dedicated GPU (Graphics Processing Unit) in the devices used to interact with the metaverse.
This approach, however, will limit usage to wealthy consumers who can afford a high-end device. The other issue with installing a GPU in devices is overheating, with a high-end GPU in a gaming console, VR headset, or television drawing approximately 400W in energy and incurring significant cooling costs.
The solution here are proximity hub edge locations. These hubs, which are close to end users, are capable of hosting shared GPUs which can be accessed on demand. Several businesses are experimenting with this approach already.
This improves both cost per unit, while also resolving some of the issues around cooling, with proximity locations typically saving up to 20–30% in costs depending on the Power Usage Effectiveness (the measure which determines energy efficiency) of the site.
How exactly the metaverse will take shape over the coming months and years remains to be seen. While it certainly has its detractors, its potential to radically overhaul the way we interact with the digital world is unquestionable. To deliver on this immense potential Edge computing will need to play a fundamental role.
Composable edge will be at the heart of meeting its processing and rendering requirements, guaranteeing the real time interaction demanded by audiences. Proximity hub locations, meanwhile, will be key to ensuring the technology is affordable and accessible to a diverse, global user base.
Many of the tech giants have bet big on the metaverse, and Edge computing may well be their trump card.
 The next big connectivity challenge: Building metaverse-ready networks, Dan Rabinovitsj, 27 February 2022
Duncan White | [email protected] | +44 7970 974832