Microwaves thrash fibre on speed... if you like two-nines uptime

The Register ^ | 27 February 2013 | Dan Olds

[ShadowAce]

“Time? They ain’t making any more of it, so you gotta get in and get out, or you’ll get your f*****g face ripped off,” screamed an old Chicago Merc floor trader at my grad school class during a field trip.

I nodded, like the rest of the class, figuring he knew what he was talking about and also not wanting to provoke this human ball of fury. But he was wrong. The kind folks who make their living providing services to financial types have been manufacturing time for decades now. And they’ve hit a new milestone.

Low-latency comms biz McKay Brothers - which hooks up financial centres - announced this month that, following an upgrade, its microwave-based network can now roundtrip a signal between suburban Chicago and New Jersey in only 8.23 milliseconds. The previous record, set by Spread Networks back in October, was an utterly lame 12.98ms over a fibre-optic connection.

Translating this to human terms, traveling the 816 miles between Aurora, IL and Newark, NJ in 8.23ms is equivalent to speeding along at a little over 356.9 million miles per hour. If you aren’t in a big hurry, you could use the Spread Networks fibre and travel at a leisurely 226.3 million mph.

McKay claims that microwave transmission is faster than fibre can ever be, which is true to a point. It’s all about something called the refractive index, which is essentially how well (or poorly) light passes through a material.

In other words, it’s how much a material slows light down versus light travelling unhindered through a vacuum. Water has a refractive index of about 1.33; a brick would be much higher. Air has a refractive index of 1.0003, while fibre optic cable is about 1.5 or so. So the maximum theoretical speed of microwave transmission is a lot higher than you can get via fibre.

But there are other considerations such as how much data you can transmit over a given period of time (the bandwidth), and the latency cost of signal amplification and relaying when covering long distances.

The biggest consideration, however, could very well be the reliability of the transmission mechanism.

“The network was down 1 per cent of the time during trading hours in December and January,” said Bob Meade, co-founder of McKay Brothers. “So our clients had the fastest connection 99 per cent of the time.”

So one per cent downtime during trading hours over a two-month period? Ouch. That isn’t all that reliable, particularly when compared to the aforementioned Spread Network (they’re the ones with the tortoise-like speed of 227,000,000 mph), which claims a 99.999 per cent availability.

McKay’s marketing partner, a company called Quincy Data, addresses this issue prominently on its website: “Better to be fast 99 per cent of the time than slow 99.999 per cent of the time.”

Now that’s world-class vendor spin. ®

[ShadowAce]

[ShadowAce]

[cableguymn]

darn it. My trade as .01 miliseconds late...

Here I am, all tooled up for fiber (not cheap) and the come out with this...

[Gen-X-Dad]

Wall St. insiders build super fast communication links to tie there supercomputing HFT network together that plugs directly into the backbone of the regional exchanges (i.e the server room).

Why? So they can front run the non-insiders to pick their pockets 1000’s of times a second. This is the definition of a rigged casino, and now you know why places like Goldman Sachs go months without losing a penny in the markets.

Just remember the exchanges sell them access to the server room wanting to skim the HFT trades and the SEC blesses all of this while the FED provides the house money for the rigged casino to fleece the non-insiders.

[beef]

I agree. They need to build a delay into every trade. It goes in, sits in limbo for say 15 minutes, and then continues on it’s way. Once in limbo, no cancellation is possible. Can someone tell me what is wrong with that?