How Hyperscalers Are Rewiring Global Connectivity

Hello, today is Friday, November 21^st, 2025. I’m Tony Mangino from TC2, and this is Staying Connected.

TONY –Today, we’ve put on our wet suits to dive into something murky in IT and telecoms deals – and that is, subsea cable projects, and why hyperscalers’ ownership of them could benefit enterprise customers. Joining me today is Deb Boehling, a senior partner at LB3 that knows a thing or two about subsea cables.

Deb, in my experience, it is rare, but not unheard of for a client to delve into the underlying subsea cable provider for its ICT requirements, particularly those customers that spend a lot of time on diversity and business continuity.  That’s always been between the telecoms carrier and the subsea cable provider, an entity most had never heard of and had no interest in.  What’s different now?  Why should enterprises purchasing connectivity think about this? 

DEB – Tony, you are right.  This is new.  In some places already and in many more to come, telecoms carriers will not dominate IT connectivity across the globe.  They will lose to or at least be challenged by the dominant hyperscalers andcloud computing providers – Microsoft, AWS, Meta, and Alphabet, referred to by some as MAMA (left out the last A as Apple is not currently a “player” in subsea investments). The hyperscalers investment in and ownership of subsea cable projects suggest a motivation to expand their reach.  Folks need to be thinking ahead of the benefits and risks of purchasing connectivity from their cloud computing providers.

Hyperscalers like AWS, Microsoft, Google and Meta are driving a major surge in subsea cable infrastructure, with hyperscaler-owned projects creating some of the largest cables globally.  Following a recent, lackluster earnings call, AWS announced AWS Fastnet, a subsea cable that will connect Maryland, USA to County Cork, Ireland. 

Meta’s Waterworth Project, one of more than 20 other similar projects, is poised to be the world’s longest subsea cable at 50,000 km.

For the past 15 years, Google has invested in subsea cables, with private ownership becoming the focus in recent years.

These projects clearly demonstrate hyperscalers shifting from relying on traditional carriers to building private cable systems tailored to their massive data demands.  

TONY – That’s a lot of investment, but again, why should enterprises care?

DEB – Thanks for keeping me on track. Let me lay out the pros and cons of their ownership.  

Hyperscaler-owned cables offer enterprise customers numerous benefits compared to existing, sometimes older, subsea cables.

Newer systems often include significantly more fiber strands, with cutting-edge technology and routes that provide superior, dedicated connectivity with ultra-high capacity, and ultra-low latency.

The cables are typically designed to have different routes with different landing points outside of the traditional cable corridors of existing systems providing additional options for route diversity. And the cables will be buried deeper, with better protection such as steel armoring and enhanced monitoring to improve incident detection.

Another potential benefit is that hyperscalers can integrate directly with enterprise-grade cloud computing and edge and connectivity services. Data processing becomes faster and closer to end-users, accelerating digital transformation projects and streamlining vendor management.

TONY – So, Deb, it seems hyperscaler dominance of new subsea cables is good, but that’s a bit naïve, right?

DEB – Nothing is all good or all bad. Every decision an enterprise makes has its pros and cons to weigh, just like they do when you and I assist with their IT deals.

TONY – So true.  You’ve let us know the pros, so what about the cons?

DEB – Tony, the challenges with hyperscaler connectivity and bundles may sound familiar, but the risks are heightened when relying on hyperscalers with already concentrated market power.

As we know, timing is critical for IT deployments. Customers worry when anything needs “special construction.” Yes, actual construction costs are a concern, but delays can be untenable and ultimately more costly. Further, the timing of government approval for landing within a country can be complicated by territorial waters and geopolitical factors. Finally, the designation of subsea cables as critical infrastructure can raise uncertainty and infrastructure bottlenecks.

Tony – Deb, what about undersea cable cuts – the oceanic equivalent of the age-old backhoe in the parking lot?

Deb  – Ha, yes! The critical role subsea cables play in global data traffic—with over 95% of internet and financial transactions depending on them  is certainly cause for concern.  Newer cable systems seek to address this, but physical cables remain vulnerable to maritime accidents, sabotage, cyber threats and accidental damage that can cause significant outages.

And its not just cable damage that can impact your business.  Deep integration with a single hyperscaler can create business continuity risks if that provider faces technical outages, regulatory interference, or disputes impacting infrastructure access. While hyperscalers should be good at diversifying their network to keep it up and running, diversifying may not be enough.  A single problem can have spill on effects as was evident with the October 29^thMicrosoft Azure and Microsoft 365 outage. Add to this, that ownership allows hyperscalers to enforce proprietary protocols and limit visibility for customers regarding routing, capacity, and security controls. This can hinder multi-cloud or disaster recovery planning outside the provider’s ecosystem.

TONY –What should you do now while waiting for new routes to be completed?

DEB –  Every enterprise will have its own calculation of risks and rewards. The calculation often depends on where the vast majority of its operations occur, and the sensitivity of operations to latency and subsea cable routes for connectivity. Regulated enterprises, particularly financial institutions and providers of health care or related services will be focused on latency, business continuity and security and likely seek diversity.

Internal work is needed first to develop a list of their hyperscalers, each hyperscalers’ provided cloud services, which divisions or groups use each, the use cases each supports, what data is processed, etc.

With that information in hand, we suggest conducting due diligence on the specific subsea cable projects supporting your potential hyperscaler connectivity services, noting their completion status and geopolitical risks.

When ready to issue an RFP for the IT connectivity, an enterprise should require the hyperscaler to provide explicit details on cable routes, landing jurisdictions, operational transparency, and incident response times. You will also want to request detailed security documentation and response playbooks and align your internal cyber resilience and business continuity planning accordingly.

Understand the tradeoffs between the superior performance of hyperscaler private cables and the potential risks from concentrated control of players with significant market power. Consider diversified connectivity strategies with multiple providers or cables crossing different routes to avoid over-reliance on a single cable or hyperscaler network, especially for mission-critical applications.

TONY – Any closing thoughts?

DEB –Keep the hyperscalers in mind when issuing an RFP, but make sure to maintain supplier, route diversity, and that business critical applications can be accessed in short order when problems arise.

Tony – Deb, thanks for joining me today for this deep sea, deep dive! 

To our listeners, if you would like to discuss the how hyperscalers might play into your network technology and sourcing roadmaps, or if you’d like to discuss other technology strategy, sourcing and cost reduction needs with Deb, me, or any of our LB3 and TC2 colleagues, please give us a call or shoot us an email. 

You can also stay current by subscribing to Staying Connected, checking out our websites, and following us on LinkedIn.