Telcos must help realise the AI vision through the edge

If 2022 was the year of the metaverse, 2023 was the year of Gen AI. And in 2024, that AI frenzy has only continued. The optimism around new AI-driven services and efficiencies, across all sectors, has started to manifest in real investment across the infrastructure layer, too. Levnovo’s Kirk Skaugen recently suggested that “75 percent of data compute is moving to the edge” as a result of AI, and a forecast from IDC seems to back this up, projecting that global spending on edge compute this year is expected to come in at $232 billion, an increase of 15.4 percent from last year.

Delivering applications closer to the end user is a critical factor for a lot of AI applications. Rather than vast volumes of data moving to big data centres as we’ve historically seen, the algorithms are now moving to the edge instead — to where the data processing is happening. This enables the real-time decision-making required by some AI applications. It’s also much more efficient, and dispersing edge compute across geographies has the potential to improve the time taken to process data and help manage the massive demands data centres place on energy grids.

But as core data centre buildouts continue to unfold across vast expanses of the world — notably across Scandinavia, Asia and the Australia-New Zealand (ANZ) regions — the shift toward edge data centres presents both challenges and opportunities for telecoms operators.

To the edge!

In North America and Europe, where the pace of large-scale data centre deployments has tempered slightly, the focus is now shifting to the edge. Edge data centres are poised to become the cornerstone of modern telecoms infrastructure, strategically positioned closer to end users to minimise latency and enhance user experience. This decentralised approach to peering represents a departure from the conventional reliance on central data hubs, such as those in Manchester and London.

Similarly, in Germany — where traffic traditionally funnels toward Frankfurt – telecoms operators are reevaluating centralised models. This shift might be a response to rising land and energy costs across the Western hemisphere. By distributing computing resources across multiple locations, operators are optimising network performance and addressing two key challenges in deploying AI: network capacity and power supply.

Driving sustainability through edge compute

The environmental impact of data centres continues to be a critical concern. According to an International Energy Agency (IEA) forecast, Ireland’s data centres are set to account for 32 percent of national electricity demand in 2026. Globally, the IEA predicts that electricity consumption from data centres, cryptocurrencies, and artificial intelligence could double between 2022 and 2026. Analysts at Morgan Stanley expect Gen AI to drive more than three-quarters of global data centre power demands in 2027, based on 2022 figures.

A key factor behind the increase of edge data centres has been the need to enhance sustainability in storage and processing, thereby reducing the burden on existing power grids.

Telecoms operators and data centre providers are exploring various techniques to achieve this, including adopting the latest optical technology, implementing more efficient cooling methods, and utilising alternative power sources. Decentralising data processing and storage also helps distribute the massive capacity demands across different parts of the network. In regions like the UK, where large data centre sites have historically been concentrated around hubs in London or Manchester, there are some indications that the growth in data centres is putting pressure on the availability of electricity for homes. Decentralisation will continue to be a priority to alleviate some of this strain on power grids.

In countries like India where coal remains the primary energy source, the environmental footprint of data centres looms large. With data centre capacity per capita lagging behind North America and Europe, India is a great example of the global imperative to address energy consumption in tandem with digital expansion.

As data centres are increasingly built with sustainability in mind, Norway stands poised to capitalise on the shift. With its abundance of renewable energy, cool climate, and historic and ongoing investment in international terrestrial and domestic connectivity, it will be interesting to see how many new large deployments land here in the next few years.

The low latency opportunity/challenge for telcos

By capitalising on the proximity to end users and the ability to process data closer to the source, telcos can develop and support a plethora of localised services. For themselves and their customers, real-time data analytics – coming from IoT devices and the like – will result in ever more responsive business decision-making and an explosion in service innovation.

However, to realise these gains, the networks within, and between, data centres must be upgraded. Cutting-edge AI services can't run inside everyday data centre servers. Instead, they need computers with high-performance graphics processing units (GPUs). And those high-performance clusters of GPUs running AI services need high-speed networks to move AI-related data inside a data centre and then out to the wider world. Outside the site, high-speed and high-capacity data centre interconnect (DCI) networks must remain front of mind for investment.

Fortunately, the performance and economics of modern coherent optical technology means telecoms operators have the means to build the necessary infrastructure for an AI-driven future. The success of AI doesn’t just hang on the availability of GPU technology; telecoms operators will also be shaping future of data centre infrastructure and must ensure we’re moving to more efficient and sustainable technology solutions, as we scale to meet demand. 

Jürgen Hatheier is Ciena’s International Chief Technology Officer. In this role, Jürgen is responsible for aligning Ciena’s product portfolio to the business challenges and opportunities of customers across the EMEA and Asia-Pacific regions. With a wide range of international experience spanning North America, Europe, and Asia, Jürgen brings more than 20 years of telecommunications experience in software, hardware, and network operations to Ciena.

Prior to joining Ciena, Jürgen held leadership positions in product management and R&D for ARRIS and Technetix where he served tier-1 network operators across the globe. Jürgen holds a master’s degree in business administration and economics from Johannes Kepler University Linz and a bachelor’s degree from HTBLA Leonding in Austria. You can find him on LinkedIn.