Matt's responsible for our digital connectivity portfolio.

The operational ideal is a network that flexes easily to meet changing requirements, perhaps supporting the setup of new branches in new countries within days, perhaps rapidly winding down capabilities in a different region. 

The reality is very different, and it’s characterised by inflexibility. Multi-region connectivity is challenging and there are long lead times for service changes. Vendor lock-in is common, limiting choice, and holding back access to innovative cloud services offered by hyperscalers. 

Almost everything has changed in an organisation’s operational world except the networks that underpin their businesses. And this is a significant issue because, the further organisations move into a cloud-centric world, the more the inflexibility of traditional networks will hold them back and increase their vulnerabilities.

Nothing highlights traditional network inflexibility, and the need for change, like network failure. It’s important to recognise that older networks were architected to principally carry traffic between an organisation’s sites - not to connect those sites to hyperscaler providers. Country coverage was key and building a Point of Presence (PoP) was expensive, so most traditional networks still only have one PoP per metro area or region. 

As a result, many organisations don’t have PoPs in the right place to optimise traffic to hyperscalers. So, when an express route goes down and traffic is immediately redirected to the nearest PoP, that PoP can be all the way over in the next availability zone, a longer journey that creates a big increase in latency. This significantly impacts the performance of critical cloud apps. 

What’s more, there’s no realistic workaround to this – organisations can’t rely on SD-WAN overlay networks to reroute application traffic because SD-WAN performance is only as good as the underlay it’s using.

Our data-heavy society is bandwidth hungry, requiring more and more to support an ever-growing reliance on cloud traffic and applications. However, it takes time to make changes to a traditional network, and organisations wanting to avoid long waits for additional bandwidth tend to overprovision to give themselves capacity for expansion. It can also take up to 120 days to get a manual circuit changed, leading to organisations reluctantly incurring extra costs to support ‘spare’ bandwidth to give a degree of flexibility and avoid bottlenecks. This is far from ideal, and severely limits an organisation’s ability to test out new links and locations quickly and easily.​

Building an effective cloud-centric future involves starting from the ground up, and that means to be redesigned networks. As organisations transform, they need to be able to minimise risk and optimise spend. No more hedging bets with overprovisioning. No more accepting big latency increases when connectivity paths go down.

They want to be able to spin bandwidth up and down to fit their evolving connectivity requirements, without waiting weeks, and quickly add apps whenever they need to. They want it to be easy to right size infrastructure build as they go, to keep costs in check. They want the freedom to be proactive, rather than reactive. And, ultimately, they want to keep employees productive, and customers satisfied.

Our revolutionary new Global Fabric network will be designed for the cloud-centric environment of today and tomorrow, with flexibility a guiding principle throughout the build.

Global Fabric is a fully programmable underlay that delivers networking services to the cloud and between clouds, whether public or private, as well as between sites. The key takeaway is that it will be pre-integrated with 630 cloud product and partner services and connected to around 700 data centres worldwide. 

This means it will offer direct connections at the click of a button into 74% of the hyperscalers’ global PoP locations – 48% more than the nearest comparable telco provider today*. It’s worth noting that we’re building at least two PoPs in every metro zone so, should one fail, traffic stays within the same metro and latency is unaffected. Our PoP rollout investment strategy is designed to predict and meet our customers’ evolving coverage and performance requirements so we can always deliver great public and private cloud coverage with low latency. 

However, it’s not just our PoP coverage strategy that will define the success of Global Fabric. Instead, it’s the combination of cloud coverage, latency, resiliency and embedded long-term flexibility to meet evolving customer requirements that will signal true transformational change. Global Fabric customers will benefit from optimised network design to reduce TCO, pre-integration with hyperscalers and data centres, Network-as-a-service (NaaS) to flex connectivity up and down to meet business needs and help with workload migration, plus effective cost modelling that will help to reduce egress charges.

We’re also weaving operational flexibility into how you can use our network – choose from DIY, fully managed, or something in between. Digital management processes have been simplified to ease the burden on IT teams - there’s no need for IT specialists to spend time they can’t really spare on manually engaging with telco processes anymore. The exact division of responsibilities is up to you - you can specify the in-house and outsourcing management balance that works best for your organisation. 

Plus, we’ll be offering offer a choice of services, from services that deliver rapid impacts, through to in-life capabilities for longer-term support.

Visit our dedicated webpage for more information, including our whitepaper and expert blog posts exploring the five transformative benefits of Global Fabric.