In a comment on last week's article about the introduction of Spread Network's Ethernet wave services, a reader asked whether there was much aerial fiber on Spread's new route and whether that added risk to its ultra-low latency proposition. My first thought was that a bunch of aerial fiber might help explain the stealth of their buildout, the directness of the route, and lower costs as well. But as speculation goes it was obviously guesswork, so yesterday I asked them straight up and a Spread Networks' representative sent me this response:
Less than 1% of the fiber mileage of the Spread Networks route between New York and Chicago is aerial. Aerial installation was used in circumstances where underground alternatives presented the greatest risk of fiber cuts due to placement of nearby facilities.
So there you have it. Less than 1% translates to 8 miles or less of the total route being aerial fiber, with those spans coming in areas where underground fiber cuts were also a substantial risk. (Probably somewhere in NJ, where else?) Doesn't sound like they cut that corner, so much for that one!
I'm curious though, if the answer had been otherwise, would it have really mattered? I mean, this is just one route, a big financial player would obviously have it backed up with another diverse route - probably one with very low latency as well. Even if one route were the fastest but riskier in terms of fiber cuts, you'd still have a speed advantage 99% of the time. It's not like they wouldn't use it, traders are used to flying without a net.