During the dot-com boom, so many undersea cable delivering the Internet traversed the bottom of the ocean between the U.S. and Europe that bandwidth prices plummeted and providers of submarine cables filed for bankruptcy. But those cables may soon no longer be enough to satisfy the global demand for bandwidth between the two continents, according to research out today from TeleGeography. The research firm estimates that bandwidth requirements will grow 33 percent between 2008 and 2015, and trans-Atlantic capacity will be exhausted by 2014.
The report also notes that the wave of bankruptcies caused by the oversupply of trans-Atlantic fiber during the boom artificially lowered the cost of providing bandwidth on those cables because many of the pipe providers were able to erase their cable construction debts. That’s good for the current customers who now pay lower prices for transporting their bits, but it means current prices don’t take into account the construction cost of the cables. So future customers will likely see some price increases on wholesale bandwidth as pipe operators add more capacity, and find themselves paying for expensive optical infrastructure. It’s a good thing that this undersea cable buildout is expected to be cheaper than the last one. From the report:
Consequently, current wholesale rates of approximately $14,000 per month for a 10 Gbps wavelength reflect only the incremental cost of the optical equipment needed to provision the circuit, but not the cost of cable construction. “Trans-Atlantic cable operators and wholesale buyers are facing a slow-motion crisis,” said TeleGeography analyst Alan Mauldin. “The cost of circuits on a new cable built today would be far higher than prices prevailing in the market.”
TeleGeography argues that while 2014 is still five years off, lengthy cable financing and construction cycles mean that carriers must confront this challenge soon. Cable operators may be able to extend their existing infrastructure by using 40 Gbps transmission line rates, but these technologies are unproven on a commercial long-haul submarine cable, and will only postpone the need to lay new fiber.