Facebook: An evolution of the sea

Facebook: An evolution of the sea

30 November 2020 | Natalie Bannerman

Cover

Facebook’s Ricardo Orcero speaks to Capacity’s Natalie Bannerman about the company’s history in the subsea sector and its future as an innovator in the space.

Motivation is key when assessing any decision. Facebook’s decision to enter into the subsea space – as a social media and social networking provider – was certainly one that made sense but demanded explanation outside of its ecosystem.

“When I started out at Facebook almost seven years ago, we were just running on leases. Our traffic however was increasing and continues to grow. And there was not enough capacity in the market to serve our needs in a scalable way,” explains Ricardo Orcero, network investments - submarine cables EMEA, Facebook.

As a result of these conditions, Orcero says that Facebook’s opex increased at the same rate as its traffic, requiring some drastic steps in order to bend the curve and to have enough supply to serve the company’s needs.

“So those were the main reasons we started to look a bit more holistically at the problem and at potentially joining one of the existing cables, but the commercial ask was quite high.”

Instead, the company decided to partner with two carriers who were leading two new transatlantic cable initiatives, and together with other OTTs became the anchor tenants, which in turn helped them with the financial closure, bringing the supply contract in force to get those systems built.  

“Once we enabled those two systems, we realised where we need more capacity but also more diversity to secure our capacity growth.”

Since that initial taste of subsea participation, the company then partnered with Microsoft and built the MAREA cable system – of course with global infrastructure company Telxius.

Interestingly it was that first project that saw Facebook seek out a new business model that incorporated the best of private and consortia cables.

“Private cables that you can operate independently, the fibres that you are entitled to and from the consortia model that it is a co-build, meaning, that we are not taking the full risk of building a submarine cable alone; but  with other interested parties.“

One of the most recent and probably most talked about systems is the 2Africa Cable – not just due to its 37,000km size but for what it’ll do for the African continent. For Orcero it is the culmination of all the company’s learnings in the subsea space.

“I would say it’s an evolution of the learnings that we have faced in the Atlantic and Pacific. Trans-Pacific or transatlantic cables are serving a high-volume market, connecting Europe to the US or Asia to the US, mainly driven by data centre traffic. In the case of Africa, it was more about serving the connectivity problems that the world’s facing.”

When deciding to develop 2Africa, Orcero shares that there were only a couple of cables in service, built circa 2010, to South Africa. But those systems still operated using outdated consortia business models that that often led to upgrade cycles two years long.

In addition, the technology used in those cables was far behind the innovations developed over the last five to six years, meaning that in three to four years, some of those systems could really reach their capacity limits and hardly be available.

“And that is problematic,” says Orcero. “But we understood that a cable like 2Africa might take three to four years to be built. Then we just initiated discussions with the main players in the African continent who we thought would have interest in securing capacity. “

Leveraging the same model as when Facebook started with MAREA, one thing identified as being very important for the African continent was the open access model.

“2Africa landings across the African continent provide transparent and non-discriminatory access to any local licensed carrier to the capacity on the submarine cable. To ensure this, 2Africa lands or terminates directly in carrier neutral data centres and whenever not feasible in carrier neutral cable landing stations.

Just going out of a cable landing station in Africa, might in certain cases, represent as much as going from the cable landing station back to Europe.”

Ultimately the goal, according to Orcero, is to develop an ecosystem at those landings where data centres and internet exchange points might grow and allow any carrier in those countries to connect to the cable system.

“We believe that landlocked countries will also benefit from the openness of this system. They only need to reach the border crossing to the country where the cable has landed and from there, we hope that they would have multiple choices to reach the cable landing station,” he says.

As for any concerns about the underdeveloped infrastructure that exists in parts of the continent Orcero says that there is a lot of interest in Africa and transforming it into a digital economy.

“With a lot of national plans at country level, we are engaging as well with regulators and policymakers in all those countries as 2Africa, just to see if we can really align interest and make sure that countries will be ready when the cable is ready.”

Also, as 2Africa is a three to four-year project and based on the numerous technology trials and investments currently underway “we believe that probably at the time we land, the demand is going to be there,” adds Orcero.

As conversation turns towards the biggest demanders of capacity, it naturally turns to gaming, which has surge during the pandemic, and Orcero says that it draws comparisons to those in the world of trading.

“High-frequency traders really don’t require a lot of bandwidth but are very sensitive in terms of latency,” explains Orcero.

“In the case of gaming, they have this similar requirement, but you also get the volume that the high-frequency traders don’t have. So, what will likely happen is that when new cables or better new networks, as combinations of terrestrial and subsea, get designed then probably latency will become more relevant than before.”

In terms of technology Facebook is banking on a number of approaches from maximising the capacity per fibre pair on the likes of its MAREA subsea system to integrating newer innovations like Space-Division Multiplexing (SDM). But one key technology is using aluminium instead of copper for the conductor in its cables.

“Aluminium has been used in power cables, subsea power cables, for a longer period of time, but the subsea industry from a telecom perspective has been more conservative. This is really something because the cost of aluminium compared with copper, it’s significantly lower, and we are talking about thousands of kilometres.”

Subsea as we know is a fairly small, albeit close-knit, community that much like the rest of the industry is facing a crisis of sorts to attract a diverse and talented workforce to secure its future – Orcero agrees.

“This is a real concern,” he says. “On the Facebook side, we have a very great diversity policy so that we incorporate people from different genders, different races, et cetera, at the company level. We don’t discriminate and as a result have a very diverse team within our company.”

Looking at the market, particularly at the main suppliers Ocero say that it’s “an area of concern”.

“What we are trying to do is encourage those companies really to look carefully at those diversity policies that they have and bring younger people, as well as diverse people, into their companies.”

The supply chain has also been impacted by Covid-19, complicating an already complex cable laying process with added health and safety harbour restrictions that need to be followed.

With several systems under construction as we speak, the roadmap for Facebook’s subsea business is to “continue building the systems that we have had in the planning phase,” says Orcero.

And as global demand for low-latency connectivity continues to grow, there will be many people hitting the ‘like’ button on that.