YOU ARE AT:AI-Machine-LearningHow the pace of AI is fueled by MOFNs (Reader Forum)

How the pace of AI is fueled by MOFNs (Reader Forum)

The need to build out high-performance networks faster to provide true global connectivity has given rise to Managed Optical Fiber Networks or MOFNs

The demand for bandwidth across the planet is consistently increasing — and there is no sign of it stopping. The International Telecommunications Union in 2022 recorded a 25% growth in international bandwidth usage,

Note the year: 2022. This was just prior to the arrival of consumer-facing artificial intelligence (AI) application ChatGPT. Since then, AI has begun driving even greater levels of bandwidth demand as enterprises and consumers alike avail of humanity’s latest and greatest innovation. According to a recent analysis from research firm Omdia, monthly AI-enriched network traffic is forecasted to grow at approximately 120% (CAGR) from 2023 to 2030. 

Global cloud providers have been on a constant mission to expand rapidly to fuel this growth, even prior to AI, and to do so they need to extend their reach into new and emerging markets. They’ve invested in data centers and submarine cables, but the key is the terrestrial networks to connect them seamlessly – and rolling these out is easier said than done, even for the major cloud providers.

For one, while cloud providers often build their own network infrastructure in many parts of the world, some countries have strict rules for fiber ownership, policies on licensing and standards for workforce qualifications. Additionally, it can be difficult for cloud providers to obtain dark fiber or get licenses to own and/or operate networks in some geographies.

Getting into new markets quickly can be challenging once data privacy regulations and emerging sovereignty laws — which vary wildly from market to market — are taken into consideration. There’s a requirement to store, compute and ship data locally, which means building infrastructure in some areas that may not exist.

The rise of the MOFN

Ultimately, both global cloud providers and service providers need to work together to build out high-performance networks faster to provide true global connectivity.

This gives rise to the Managed Optical Fiber Network (MOFN), a business model which has been in use for a long time but gaining renewed interest and increased traction as cloud providers expand operations and seek efficient ways to interconnect their sprawling infrastructure. MOFN represents a collaborative business model where cloud providers partner with service providers for dedicated, high-performance connectivity in regions where they cannot or do not want to own and operate their own terrestrial networks.

It’s a model also used for subsea cables — the content service provider or cloud provider owns the fiber pair and the operations, but the ownership of the equipment depends on a country’s rules and regulations.

The service provider — the party within the MOFN that is responsible for moving the data quickly for the cloud provider — builds or leases a dedicated fiber optical network based on the cloud provider’s unique specifications and then offers it as a managed service to the cloud provider.

An example of this is Colt, which is taking the reins of connectivity for many of its customers across Europe, Asia Pacific and North America. With a growing need for fiber connectivity to pass through jurisdictions and borders, Colt is partnering with stakeholders at the early stages of new route builds and proactively investing in its network to meet future demand from global cloud providers.

This enables Colt to shorten the time to roll-out MOFNs as the network can be designed in the background, with Colt being relied on as the primary contact for the network build and operations. Ciena and Colt work together to help drive success in the MOFN arena.

Instances like this are why cloud providers are turning to MOFN to speed up their ability to increase network capacity and add new routes in jurisdictions where they may face restrictions.

We’re seeing similar instances of these collaborative approaches across the globe. One North American service provider was able to offer a cloud provider a MOFN of more than 34Tb/s C-band capacity from Seattle to Chicago to support an AI service initiative. And a Canadian service provider was able to offer a MOFN with a C-band line system from Canada to the US to enable fast and cost-effective capacity adds for a cloud provider.

Technology vendors play a critical role here too. In this collaborative model, cloud providers outline their needs for performance and expansion, while service providers showcase their infrastructure capabilities. Vendors then position the right technologies to bridge these objectives, creating win-win scenarios that accelerate deployment and innovation.

Once the MOFN model is established, expansion can be achieved faster and data — where the true value lies — can be monetized by all parties.

Getting into new markets quickly will increasingly be key to cloud providers maintaining growth and scaling — yet increasingly common data sovereignty laws and local requirements mean they can only scale at the speed of regulation. However, increasing the speed of deployment is critical to global cloud providers to keep up with the pace of AI — if not stay ahead of it.

The demand for bandwidth is an appetite that shows no signs of abating. As we move towards and beyond 4K streaming, to workloads increasingly all in the cloud, and particularly as AI continues to evolve, providers — both service and cloud alike — are set to lean on MOFNs as the business model to fuel this demand from users around the world.

ABOUT AUTHOR