The extent of this boom is revealed in new figures from analyst firm TeleGeography, showing that between the beginning of 2006 and the end of 2010, demand for international bandwidth grew seven-fold.
Network operators have matched this demand with a flurry of activity. Old networks have been upgraded well beyond any capacity levels that were initially envisaged, while 49 brand new undersea systems were launched, costing around $6.5 billion between them.
Capacity demand has grown fastest on links to Africa, Latin America and the Middle East, these regions between them experiencing growth in demand of more than 80% per year over those five years. Mature markets have demonstrated hunger too, with international bandwidth usage increasing 63% per year in Europe and 54% in the US and Canada over the period. Supply has more or less mirrored this demand, growing 53% overall in 2010, representing an additional 16.1Tbps of new international capacity.
Pricing for capacity has naturally headed south, declining at an annual rate of around 25%, varying by capacity, location and timeframe. Pricing volatility has not at least tipped over into outright oversupply, believes TeleGeography. Its research proves that while new cable development can cause some short-term price volatility, there is no pervasive supply glut to trouble the market in the same way that oversupply trashed transatlantic pricing a decade ago. Building and upgrading terrestrial and submarine networks remain vital, it concludes, to prevent demand from overwhelming supply.
Pricing on higher capacity circuits, especially 10G wavelengths, has declined more rapidly than on smaller circuits. Wavelength transactions represent a bigger segment of the overall global market as demand continues to shift from smaller circuits to high capacity circuits, says TeleGeography.
Circuit pricing differs significantly by region, dependent on available supply, competition, distance and cost. Transatlantic, trans-Pacific and 10G wavelengths across the terrestrial US are among the cheapest routes worldwide in terms of price per mile, while Miami to São Paulo and Hong Kong to Tokyo are among the more expensive.
The cost of building new cables varies, driven by factors like system length and the number of landings. Africa’s new cables have come at the highest cost of recent years: the Africa Coast to Europe (ACE) and West African Cable System (WACS) cables costing $700 million and $600 million respectively.
So where’s the next wave of growth coming from? “I think we’re going to see a shift away from the places that have had lots of new capacity, like Africa, Middle East and Asia and more to south America where there’s been no new major cable for some time,” said TeleGeography research director Alan Mauldin. “There’s been no announcement yet, but something’s due.”
Mauldin says that the boosting of capacity on existing cables has to a degree pushed back the need for new cables on most routes: “Many are being retro-fitted for 40G,” he explains. “But we will need some new systems. You think the cable boom is over, and then it’s not.”
He does believe that once the current wave of new African builds is launched, there will be no new cables in that theatre for some time: “There will be some African countries with at least three major cables landing there, like Kenya, South Africa and Nigeria,” he says.
But Mauldin believes that the inevitable interplay between supply and demand is not the only factor driving the creation of new systems: “There’s the need for diversity and different routes, and also a demand for reduced latency too,” he says.
Latency has always been important in what makes a network popular and successful, but the advent of high frequency securities and commodity trading has intensified interest in low-latency systems dramatically. High-frequency traders set great store by access to the lowest latency bandwidth available between strategic financial centres, and will pay significant premiums for circuits to match. “Carriers have moved quickly to identify and exploit low-latency network paths, optimising and guaranteeing path routing and even constructing entirely new systems to improve latency performance,” says Mauldin. “The most prominent example of this is Hibernia Atlantic’s Project Express transatlantic cable system, designed specifically to optimise latency performance.”
Another notable trend, according to Mauldin, is a rethink of what has seemed at times almost slavish reliance on some of the world’s more troubled capacity bottlenecks. “Egypt’s well-documented problems with regulation, as well as unrest there, are leading carriers to think the previously unthinkable and consider terrestrial routes through places of geopolitical risk,” observes Mauldin.
He points to the various submarine cable outages of the last few years that have dramatically reduced connectivity to large geographic areas, many occurring in the same bottlenecked regions over and over again.
“Carriers continue to seek methods of mitigating risk – whether from dragging anchors, earthquakes or geopolitical factors – by improving the reliability of infrastructure,” he says. “Recurring submarine cable faults between Taiwan and the Philippines, for example, are leading carriers to build a cable that bypasses the area. The resilience of Japan’s international connectivity following the massive earthquake in March 2011 shows how multiple cables and diverse routing options can limit harm to communications when disaster strikes.”
Damaged Middle East cables and delayed completion of several new submarine cable systems connected to Egypt have helped to fuel interest, he says, in the JADI and RCN terrestrial cables, linking the Middle East to Europe via Turkey.
“Terrestrial routes avoid the dangers that plague subsea systems,” says Mauldin. “But they are not exactly free of risk, just different sorts of risk. The more options you have, the better.”
By no means all the capacity action of the last five years is submarine. Carriers are also upgrading their terrestrial networks. In Europe, 79% of the carriers surveyed by TeleGeography said that they planned to light additional terrestrial wavelengths this year, as did nearly 89% of American carriers.
But the international subsea routes make bigger headlines, indicative as they are of changing trends in international traffic volumes as well as direction: “We see most growth in the submarine links connecting Asia-Pacific, the Middle-East to India, and the US to Latin America,” said Mervyn Kelly, EMEA marketing director with equipment vendor and cable upgrade pioneer Ciena. “A key driver for this growing demand in developing countries is the rise in broadband access. For example, in key South American markets there is a huge push to increase internet access for education and business purposes. This puts strain on the connection to North America, and explains the heavy investments in this area.”
Being able to cost-effectively scale bandwidth is a priority for carriers looking to meet growing capacity demand, says Kelly, especially given the cost of installing submarine cables. “According to a survey by Ciena, almost a third of carriers perceive the cost of deploying new cables as their number one challenge, followed by scaling the business with demand, and the capacity constraints of existing cables,” he says.
Kelly claims Ciena’s 40G/100G technology for submarine networks addresses this by allowing carriers to seamlessly upgrade their infrastructure to 40G/100G with only the addition of new terminal equipment, extending the life of existing cable plants. “Another key factor is the ability to survive multiple failures using mesh network protection, minimising the risk of network disruptions,” he points out.
The feverish levels of cable construction and refurbishment at least illustrate that wholesale international bandwidth, while not the most lucrative or most glamorous segment of the telecoms market, does still serve a vital role.
“As high-bandwidth applications increase in popularity, carriers benefit from increasing traffic growth,” says Mauldin. “Burgeoning demand has outstripped price declines, implying growing revenue for wholesale bandwidth. Wholesale margins tend to be very narrow, and profitability depends on operational scale and low unit costs. However, the industry does face significant challenges in the coming years.”
