South Africa and the rest of the continent will in the not-too-distant future be connected by a whopping submarine cable, possibly measuring between 100 and 120 Terabits per second (Tbit/s).
That eclipses anything else currently linked to SA by quite a margin, prompting the question: do we need more of the international bandwidth these cables deliver?
If you permit a broad stroke attempt at it, the answer is interesting. An obvious place to start is seeing just how much the above-mentioned submarine cable, called Equiano, would represent among our current capacity.
According to the site Submarine Cable Map, there are six submarine cables of consequence landing on SA’s shores. Combined, they offer a capacity of around 42.3Tbit/s. To get a sense of scale, let’s compare that to the capacities of two other areas. Around 10 cables between North America and Europe total 317Tbit/s. India has about 12 cables, not counting regional links such as to Sri Lanka, that collectively hold 127Tbit/s capacity.
Africa as a whole is relatively sparse in comparison. Other than the cable networks that extend to SA, there are a few on the west coast and a concentration along the northern coast and the Red Sea. There are also two recent massive lines from South America of between 30 and 40Tbit/s each. Still, Africa’s total capacity is quite small and a large chunk of that capacity is in the new transatlantic cables not yet distributed along the coastline.
The situation becomes more acute if you focus on a different metric: the amount of capacity inside the cables that are lit and active. Of SA’s 42.3Tbit/s capacity, only between 2 and 6Tbit/s is being used. In 2016, the EASSy consortium rebuffed the arrival of the Africa-1 submarine cable, reasoning that the former has only lit 500Gbit/s of its own 10Tbit/s cable and that only around half of that was actually being used.
Granted, 2016 is ages away in technology years, but more current data backs this up. ICASA’s latest State of the ICT Sector report estimated 2018’s international data traffic into SA was just over 500Gbit/s and overall international traffic came to around 840Gbit/s. It doesn’t sound like we’re jamming up those cables...
Is Africa’s demand for bandwidth lacking? Maybe. But it can also be reasoned that the cost of data itself is an issue. Africa has some of the world’s biggest mobile- and smartphone-using populations. Mobile infrastructure has definitely spread connectivity far and wide, but data costs remain a painful barrier. It was once argued that a lack of connectivity restricted Africa’s relationship with the digital world, then it was a lack of devices. Both those have, at face value at least, been resolved. So where is the appetite?
Is Africa's demand for bandwidth lacking? Maybe. But it can also be reasoned that the cost of data itself is an issue.
To be fair, the above numbers don’t look at actual usage in other parts of the world. But it’s unlikely to be frugal. If we accept the view that data costs are holding the numbers back, then a glut of submarine capacity will be a good thing. It can have a direct impact on the cost of data and create opportunities for companies to outplay each other with cheaper data deals.
This is all academic because the glut is already on its way. Near-future projects such as Equiano, Africa-1 (50Tbit/s), SAEx1 (72Tbit/s) and the very ambitious South Asia Express cable system could more than double some of the numbers I threw around here. That is a gross oversimplification, but it’s nonetheless easy to imagine a dramatic jump in capacity causing big ripples in the African telecommunications pond. If there is any fact to be gleaned here, it’s that Africa’s usage of international data appears to be lagging well behind.
Let’s stop beating around the bush. Africa clearly doesn’t consume its weight in data, even though it represents around 11% of the world’s Internet users through only 40% penetration (Internet World Stats). South Africa is chomping at the bit: after a slight slump in 2016, demand for international bandwidth jumped 62% in 2017 and 42% in 2018. That 840Gbit/s annual consumption I mentioned earlier is the result of a growth spurt, not stagnation.
It will take a few years until Equiano as well as another proposed massive cable, Simba, will materialise in any real sense. But some of the other projects are imminent, in transition or already landed. This flood of bandwidth might finally bring down data costs where the average monthly income towards data is 5.5%, 3.5% higher than what the UN Broadband Commission considers affordable in developing economies.
Much of this, in my opinion, can be blamed on the greed, profiteering and gouging perpetrated by mobile networks onto the public. That being said, questions are already raised about the financial viability of the new submarine cable projects.
If African economies don’t start growing to raise incomes, the bandwidth glut might come to nothing. One of those major submarine cables land in Angola, a country with one of the lowest per-capita incomes in the world.
Still, do we need more submarine cables and more international data? Frankly, we need all the help we can get.
Share