The world is changing fast and to keep up you need local knowledge with global context.
JOHANNESBURG — The advent of undersea broadband cables such as Seacom on Africa’s eastern coastline has had a major impact on the quality and pricing of faster internet broadband in South Africa. More cables, such as WACS, have also boosted the country’s ability to connect to the world. But while there are cables running up the eastern and western coastlines of the continent, there still isn’t a cable running east to west. A US company called Seaborn Networks plans to fix that by building the first-ever subsea cable between Cape Town and Brazil. The cable will then ultimately connect to the US, shaving off milliseconds from South Africa’s internet latency speed. Here’s Seaborn COO Andy Bax who told me more about this project. – Gareth van Zyl
On the line from CT I have Andy Bax, who is the COO of a US-based company called Seaborn Networks. Now, this company is going to build the first and only direct subsea broadband cable between CT and Brazil. Andy, why build this cable?
Hi, Gareth, thanks for having me on, first of all. So, why build the direct cable? So, if you look at the ecosystem that exists today for not just CT but Africa as a Continent, in general. Connectivity to the rest of the world involves going either up the West-Coast or up the East-Coast, and then some rather tricky traversents through the Middle-East, and Southern Europe and into Northern Europe. Then across maybe the Atlantic to the US.
What that basically means is that to get to where, as a consumer you don’t necessarily see that, but as a provider, whether you’re an ISP or mobile network, or one of the content providers. What you basically see there is multiple hubs to basically get to one place. When I say ‘multiple’ you could be talking about passing through dozens of different locations to get to the endpoint you want to get to, and it’s a common endpoint for most companies, whether it be on the East-Coast or West-Coast of the US or into one of the big data centres. It’s a common endpoint that the idea behind this cable is that its twofold.
One is to rectify that situation and we often talk about it at Seaborn in terms of thinking of it like an airline. If you were booking a flight and you have a choice that was one stop between CT and NY, and a flight that had 12 stops. You’re never going to take the flight that has 12 stops and the reason for that is twofold. The one is it’s disruptive to your life and secondly, the chances of getting through all those airports without a delay or some kind of issue in making all those flights is slim, at best.
This is kinda similar. It’s a similar scenario so, if you’re having to jump through, keep it simple. If you’re jumping through a dozen different data centres to get from CT, book the cost, through the Middle East, through Europe, and across the Atlantic to NY. The likelihood of there being a fault in one of those data centres throughout the week and in the year that you use it, is fairly high. If your other option is to go on a route where you pass through one other cable station, and the important difference here is that it’s a cable station, it’s not a data centre where there’s people in and out every day. These are basically locked down facilities. Hardly no manual interaction goes on in a cable station on a day-to-day basis.
So, you’re passing through one other building to get to NY, and just from a pure operational perspective, that is so much more secure. It is so much more reliable and, by the way, it’s so much faster. We estimate that by building SABR from CT to Northern Brazil and integrating that into our Seabras-1 cable to NY, we will be about 30% faster than the current routes from CT to NY today so, latency will be improved.
So, Andy, would that then translate into a 30% bump-up in internet speeds, if you talk about latency?
Well, as an end-user you won’t necessarily notice the difference in terms of overall latency. We’re talking here the difference between 210 milliseconds, which it is roughly today, and what we estimate to be 140 milliseconds. That’s a couple of blinks of the eye so, in terms of what you see as a user on a user experience, you would not necessarily notice that difference. What you will see is that content providers, financial institutions, gaming companies – if you’re a heavy gamer you will definitely notice the difference because that is key to the level of service they can provide to end-users, latency.
So, you mention financial institutions as well. Does that mean the like of trading and banking will also experience a boost, in terms of latency?
Well, there’s an opportunity there for the financial institutions that do real time trading and high frequency type stuff. If that is a, and that is not the reason for building SABR, but if there’s a by-product to that then effectively, between the trade and exchanges of CT and NY then there is a new route that is somehow faster and more efficient for those traders, in terms of passing information back and forth. Yeah, that could have an impact on volumes and amounts of trade, etc. But again, that’s not why we’re building it. That might be a by-product of this build but that’s not the rationale for building it.
The real rationale for building this is actually around a more secure connection and cost of access and cost of access for small to medium ISPs. Cost of access for large telcos in the region. Cost of access for not only the largest of the cloud providers but all the up and coming cloud providers as well. When we talk about cost of access here, what we’re talking about is, it’s the difference between being able to buy sufficiently large amounts of capacity at a price that makes sense, and that means that you can do that, which then enables you to plan your network development domestically, in a much, more efficient manner. And actually, provide better levels of service to end customers, like you and me.
So, if you think in terms of small to medium ISP, today what they do is, as all these different cloud services come along and you and I are using laptops, computers, and smart TVs, and smart appliances, web-streaming 24 hours a day. Everybody in the house is trying to do something online at the same time and what that does is that that puts upstream. That puts huge pressure on your service provider because they’re having to do that for you and thousands of other people at the same time, which means they have to buy more international bandwidth. If that bandwidth is overly expensive, as it is today in this region, they can’t do that. They can’t do that without dramatically increasing your pricing and then all the users get upset and everything is too expensive. Then it’s a cycle that’s very difficult to breakout of.
What we aim to do with our SABR project is for any customer on that cable system from CT to Brazil, they will pay a proportional cost based on how much capacity they buy on cable. A proportional cost as in whatever it costs to build, they’ll pay their share of it, of how much capacity they take. We anticipate it costs them somewhere between $120m and $140m to build it so, it’s a significant project but capacity won’t be sold on it at market price. At least ahead of it being completed and built so, anybody that right now, we’re in what we call the development phase and we’re taking to all kinds of different companies, fixed line operators, all kinds of folks around, and then calculating based on that, what their share of the cost is.
I was just going to ask about that specifically. Is your company then going to own the cable because we’ve seen with other cables that there’s often a consortium that gets together and builds these cables? But from you’re telling me it sounds like your company is going to own the cable but you’re then going to pass-on the costs to the users.
Yes so, Seaborn as an independent owner and operator, we will own the cable. We’ll build it, we’ll own it, and we’ll operate it. Customers that come onboard before its built will get to buy capacity on at cost. So, as a very simple example, if the cost is $100.00 and you wanted to buy a tenth of the capacity on that cable, you pay $10.00. You don’t pay whatever the market price is – you pay $10.00 and that’s yours now, forever. I think for 20 to 25 years, you own it. You paid $10.00 for it – it’s yours. You don’t have to use it all today. You may think that my plan says I need that over the next 10 years, but if I buy it today for $10.00, I’m good. I have all the capacity I need. I can grow my business now without worrying about where do I find all this extra capacity? That’s a step-change for this region, we believe.
We saw this when we built Seabras-1 from NY to Brazil, before we went live that was one of the most expensive routes in the world, NY to Brazil, more expensive than here. When we went live what we call it is that route is now the right size price – it’s the right price. It’s not expensive anymore. It’s not as cheap as going from NY to London, but it’s not much more expensive. It’s the right price. Even for those customers that buy on SABR after we go live, the market should have been right sized by then and that’s what the region needs.
Andy, just in terms of timelines here, what are we looking at? So, you said that it’s still very much in the planning phase at the moment?
Yeah so, we’re well down the road in terms of the development phase. I think we would anticipate that by the end of this year we’ll be through that phase. We’ll have finished all of our discussions. We’ll have done a lot of our planning that’s already underway, and we’d be starting the project in Rio. As in the first-phase of that is to put a ship in the water that does a survey of the route. We would anticipate that starting early next year and then basically, it takes about 18 months to build. They take a while a build them but realistically, somewhere around two years, about two years from now the system should be going live, which sounds a long time away but it’s not, but that will go in a heartbeat.
Obviously, we have quite a few subsea cables connecting to SA right now. We’ve got more than what we ever had. We’ve got SEACOM WACs, we’ve got EASSY along the East-Coast of Africa so, will your cable then presumably also add a level of redundancy to local networks here because we’ve had issues before where a cable would get cut in Egypt and suddenly the entire network across the country would completely come to a grinding halt? So, presumably that kind of thing wouldn’t happen that much anymore, once your cable is up and running.
When we talked to service providers in the region, not just in CT and SA, but throughout the Continent, reliability is what gets them really excited about it, when we talk about it. Cables getting cut in Egypt or embargoed or other failures throughout Europe or other countries in Africa because cables, they tend to vest in up and down both coasts. The reliance on the services that are provided outside of the country you’re in, is always a concern for operators. So, the fact that this, other than one landed in Brazil where it comes up on the beach, and it immediately leaves the beach again – everything else is in the water so inherently more reliant, much more reliant. But yes, that’s it. When we talk to folks throughout the region that is what gets them excited that the realisation that they don’t have to go through Egypt. They don’t have to be reliant on 20-plus different landings or data centres to get through to get to where they want to go. That’s a huge issue for them and one that’s been around for a very long time.
Yes, and once your cable is up and running do you think that, we’re then reaching saturation point in SA, when it comes to these subsea cables because then we would have been connected to the Americas, as well as Europe, the Middle East-Asia? Are we potentially, reaching that point where we will reach some sort of saturation with cables or do you think that we’ll still need more?
I think it’s a matter of timing. Our cable won’t be a huge cable. We’re envisioning it being two fibre pairs, 12 to 20 terabits on each fibre pairs so, to be honest, 30 to 40 terabits into SA today is a huge amount of capacity. But once you start thinking down the road, in terms of 4G and 5G rolling out fully. That can disappear very quickly in terms of international cable capacity. I think the challenge for the region, (the entire Continent) is not one of are there enough submarine cables for international capacity or could more be built? There could always be more built. I think the challenge is more about pushing that penetration in terms of accessibility to this bandwidth beyond the coastline into the countries. That’s where demand will grow and the more and more networks that get built further inland, and with better penetration and a better quality of service – absolutely, there’ll have to be more cables. Just the services that would become available and, as I say, the rollout of 4G and 5G and all of this will certainly help. The data of volume on a per capita basis is just tremendous, as long as its affordable and that’s why cables why SABR are important because it will make bandwidth affordable.
What is your take on the whole last mile infrastructure in a county like SA? Obviously, SA probably has the most developed last mile infrastructure on the Continent but for many years there’s been a lot of criticism around it that it’s still not good enough. Do you think that it will ever be good enough? Do you think it will ever get to a stage where it will be able to compliment these new subsea cables sufficiently?
Well, I think, and I’ll put my hand up and be honest. Domestic infrastructure, I know in previous lives we built an awful lot of it. It’s not our area of expertise but I think what you will see is not dissimilar to what we’ve seen in some [inaudible 0:17:02.3] industry where rather than, I think there will come a point where move more and more towards wireless last mile, as opposed to wire line, and more than likely, as that changes from 2G and 3G in a country today, I think that the move straight to 5G. As that becomes more efficient and more economically and the capabilities in the reach of that will have improved. But I think that’s where we’ll see the most dramatic improvements in the last mile exits. I don’t think it will be fibre to the home. I don’t think it would be increased cable as in [inaudible 0:17:50.6] utilisation. It has to be wireless and, by the way that’s not just an African Continent issue. That’s common across all the Continents.
So, these 5G technologies and future generation networks will be able to completely replace overland fibre cables? We’re talking about fibre cables that connect mobile towers and that type of thing.
Well, like today, you can have your Smart TV and connect it to either your fibre to the home, depending on where you live. Or your local cable companies connect and you can stream content to your Smart TV, and your kids can be sat right next door to you with their smartphone or their iPad or tablet, doing exactly the same thing over the wireless connection. Their service is exactly the same as yours. As those wireless networks evolve and they become more efficient I think the ability to reach deeper into the community with those, through either handheld devices or wireless set talk boxes or wireless Smart TVs even. Then that will become a natural progression and that will replace either non-existent local infrastructure in some cases, ageing infrastructure in others, and that’s the same – that is what exists in the US today, it won’t be any different but yes, I think that’s where it’s going.
Just as a last question, Andy. Just looking at the demand for data in SA and Africa. Obviously, you guys are doing this because you feel that there will probably be a further explosion and demand. Is that a fair statement to make?
Well, I think we see growth, ongoing growth. If you go back at available day you can see the compound annual growth to the region is one of the largest. Now, admittedly, in some places it’s starting from a small starting point right now so, when you measure it against some of the more developed countries, in terms of growth by penetration, and access to capacity bandwidth. Those growth rates can sometimes look explosive but we see very good growth rates in terms of historically, but also, looking forward, in terms of forecasted growth rates, they look very good. But I think these supplements the existing network into the region because it’s diverse, because it’s more direct, and because it will be much more efficient from a cost perspective. They see it as supplement in what’s already there.
Right, Andy Bax, thanks a lot for chatting to me today.
It’s a pleasure talking to you and thanks for reaching out.
Cyril Ramaphosa: The Audio Biography
Listen to the story of Cyril Ramaphosa's rise to presidential power, narrated by our very own Alec Hogg.