Stay on Top of Enterprise Technology Trends
Get updates impacting your industry from our GigaOm Research Community
Two research groups in the last few weeks have shown off 100 terabit per second speeds delivered over fiber connections according to a nice article in New Scientist (s enl). The advancement in speeds is both unexpected and necessary as we rely on broadband as the interconnect for our increasingly digital lives. We’re about to live in a 100 gigabit world when it comes to our backhaul connections and the next step will be the terabit age to provide the backbone speeds as more and more cities get hooked up with gigabit connections.
For those who think those speeds won’t ever be necessary, or who buy into the current ISP logic of broadband scarcity and caps, just wait and see. Thanks to efforts such as Google’s (s goog) project in Kansas City, Kan. or Xavier Niel of Free, people are thinking about ways to build out faster infrastructure for lower costs, I’m hoping that not only will cities and consumers be able to afford these fast connections, but building them out will be economical. Then, such widespread adoption of gigabit connections will render the need for terabit connections and eventually 100 terabit backhaul.
So how do we get there? At the Optical Fiber Conference last month, NEC said it transferred 101.7 terabits per second through 165 kilometres of fibre. To make that happen, NEC used 370 separate lasers. A researcher at Japan’s National Institute of Information and Communications Technology in Tokyo created a fiber with multiple light-guiding cores, as opposed to the one used today, with each core delivering 15.6 terabits per second for a total of 109 terabits per second. Think of it as bonding network channels with light instead of copper. The distance this connection was able to sustain at those speeds was not mentioned, however.
Given that our long-haul networks don’t need 100 terabit speeds just yet, and because when it comes to fiber sustaining such high speeds over long distances is a key challenge, the New Scientist article quotes Ting Wang of NEC saying its likely that we’ll see these innovations picked up in data centers first where shorter distances and an immediate need for speed make the investment in glass worthwhile.