Global Capacity: Trends, Drivers, Enables and Data Centers
As published in the May Issue of SubTel Forum Magazine
By Kaushik Sengupta
May 18, 2022
The submarine cable industry continues to see unprecedented growth. Looking at recent reports suggest that during the period 2022 – 2024, we could expect to see some ~$10bn of further investment in new subsea cables, with that growth coming from both established and emerging markets. Let’s take a look at some of the trends, drivers, and enablers behind these staggering numbers.
Firstly, there is a growing trend for data centres to seek to ‘land’ submarine cables, rather than the traditional model of stand-alone cable landing stations feeding cables into the data centre ecosystem. With the subsea industry having been around for much longer than the data centre industry, there are a surprisingly significant number of similarities between the two, which, has led to a logical alignment.
Both sectors are highly capital intensive. Constructing a state-of-the-art, fully equipped, high-quality data centre of a decent size (5–30MW) normally costs between USD30 million and USD300 million. Building a state-of-the-art, multi TB submarine cable network that is several thousands of kilometres in length requires a similar investment.
Both sectors have long lead times. Data centres have lead times of 1–2 years, while cable projects take 1–4 years or longer. Proper capacity planning is required to start building ahead of demand in both cases, be it subsea cables or data centres. Consequently, we are seeing a great deal of alignment in the planning stages of subsea projects with slated data centre builds.
There is always going to be growing demand in such a fast-changing environment. The demand for digital services supported by submarine cables and consequently data centre infrastructure, is growing quickly as evidenced by almost daily announcements of new data centre projects, in all corners of the globe. This also includes the emergence of edge data centres, which in many cases are perfect to ‘double up’ and act as cable landing sites for subsea cables, given that in most instances subsea cables typically land in locations that aren’t necessarily highly populous city centres, but do lend themselves to edge-type facilities.
Disruptive technological innovation continues unabated. The technologies used for submarine cables built in 2021 are considerably improved compared to those used for submarine cables built at the turn of the century. In many ways, you could argue that Shannon’s law limit is looming and will be the next big challenge to overcome. Who remembers the last time something less than a 10G service was activated on a subsea route? Similarly, there have been considerable technological evolutions in the data centre space to push the power usage efficiency (PUE) ratio close to 1, while increasing the density and evolving architecture. Unit economics continue improving in both industries.
Many submarine cables were initially funded and deployed by telecoms operators, as were many data centres (though there were also non-operator providers). Hyperscalers have now entered both sectors in a major way. Hyperscaler’s first built their own data centres and are now building their ownsubsea cable networks.
Subsea cables and data centres are a natural fit and it is certainly a trend we won’t see diminishing in the foreseeable future. One thing to ponder though, is the cyclical nature of all things technology related. In recent history, there have been periods during which there was a relative over-supply of submarine cable capacity. This held back investment into new submarine cable systems for a period in the early 2000’s, even though the demand for submarine cable capacity continued to grow. It is not yet clear if the data centre industry will go through the same cycle in coming years on the back of further efficiency in power usage and the growing evolution from an environment of hardware to one of connectivity.
So, with this highly focussed alignment between subsea cables and data centres, what can we expect to see from the subsea industry in terms of routes and markets in the near term? We could break this down into three key categories. Firstly, hyperscale – where are the new global ‘hotspots’ based on cloud regions, secondly, emerging trend of building edge facilities in tier two cities / markets and finally, resiliency to support the entire ecosystem.
From a hyperscale perspective, two markets to watch over the next couple of years are India and east Asia, where approximately 40% of all data centre developments are happening or are slated to happen. Of this 40%, approximately 60% is geared towards wholesale activity, Vs. retail. By comparison, Europe accounts for approximately 25% new data centre developments, with most of this investment geared towards retail markets and the bolstering & expansion of existing infrastructure in well-established markets. So, let’s follow the activity and look at the implications for the subsea market in this context.
India is currently a focus area for both hyperscale data centre development and subsea developers with a multitude of new builds planned to land on both coasts, providing onward connectivity to the east towards Europe and west towards Singapore and beyond. This is driving some ‘new’ thinking about resiliency and single points of failure between Europe and Asia and is reflected in the new routes being contemplated on what has historically been a very traditional path – Singapore – India – Egypt – Marseilles
Currently, there are no less than three new builds under consideration which aim to bypass Egypt, which is considered the ‘traditional’ routing on this path between Europe and Asia. We are now seeing the Kingdom of Saudi Arabia coming into play in a much more meaningful way, with cables looking to traverse the Kingdom north – south, cutting through Israel and Jordan and then onwards to Italy and the FLAP countries, consequently, taking Egypt out of play. This achieves a couple of objectives, namely bypassing what is a massive bottleneck of cables at the northern end of the Red Sea whilst also reducing cost associated with Egypt crossing fees. It’s a very unique route and we are highly confident it will be very popular.
Whilst contemplating the Europe to Asia route, the challenge still remains between India and Singapore. The Malacca Strait is also coming into sharp focus as a narrow stretch of water which represents risk for cables between India and Singapore. We anticipate that developers will look to bifurcate risk by capitalizing on the new trend of edge data centre establishment in tier two cities. Possible routing solutions to minimize the risk of the Malacca Strait exist, such as landing further north in Thailand, in a tier two city such as HatYai which has an evolving edge landscape given its proximity to the Malaysian border, and the coming down to Singapore terrestrially or via the Gulf of Thailand.
Similarly, we are seeing massive development and spending on ‘new’ routes between the US and Asia which is in a not dissimilar situation to the Europe – Asia path. If we look at the traditional routing between Asia and the US, realistically, there have always just been two options. Firstly, Hongkong to the US via the Luzon, or Singapore – Hongkong – Japan – US. As with Asia – Europe cables being planned which bypass Egypt, we are now seeing systems on this Pacific route going directly into Singapore from the US, routed south of the Philippines. There are at least three significant builds to be announced on this path, bringing with it new opportunity for data centre development in markets such as the Philippines, Guam, Indonesia and surprisingly, Darwin in Australia.
On both traditional paths discussed, (Europe – Asia) and (US – Asia) are in a state of evolution from a routing perspective. India’s economic growth as a nation and avoidance of routing risk are both two key drivers, whilst hyperscale and edge data centre development are the enablers for the subsea development on both paths mentioned within. The data centre trend towards edge facilities provides greater flexibility for the subsea industry in terms of routing, offering up new options for landing subsea cables which ultimately, provides for a global fabric which is much more resilient than it ever has been – and it needs to be. Our subsea cables underpin everything digital.
As an industry, we need to cater for the exponential demand from a global population that is hungry for digital and cloud services, be it consumer, enterprise, or governments. Some of the key drivers that will underpin the industry, be it storage or connectivity as we move into the middle of this decade include the Internet of Things which is predicated to see some 150bn devices connected by 2030. Artificial Intelligence is also continuing to see unparalleled growth, leading to great changes in industry, communication, transportation, education, energy, health care, entertainment, government, warfare, communication and research. Industry 4.0 is also a consideration, particularly in emerging markets. Finally, quantum distributed computing or cluster computing also offers great opportunity for the subsea space.
At a more fundamental level, one only has to look at smartphone traffic growth to get a sense of the immense anticipated growth in digital services that is coming. Total global mobile data traffic was around 65Eb per month at the end of 2021 and is projected to grow by a factor of around 4.4 to reach 288Eb per month in 2027. Sure, not much of this traffic will directly traverse subsea cables, but most certainly at an ‘infrastructure’ level, there will be massive M2M traffic flows across subsea cables between servers on a global scale to underpin all of this activity. Again, it’s all about subsea working in concert with data centres to address the demands of a digital hungry global population.
The opportunity for continued growth knows no bounds.
About the Author
Kaushik Sengupta is Senior Vice President of APTelecom. Kaushik is an integral part of the APTelecom team, supporting business growth in India and southeast Asian markets. With over 25 years experience in the wholesale telecoms industry to leverage off, Kaushik supports APTelecom’s clients in the areas of Submarine Cable, Satellite Services and Managed Services, including network consultancy, cloud computing and IoT.