MIT: Light speed now a bottleneck in fastest networks
- — 07 December, 2010 04:05
With global networks carrying complex time-sensitive data, the speed of light is actually becoming a significant source of latency, researchers have found.
While today's fiber optics-based networks can shuttle data around the world at the speed of light -- momentarily slowed only by routing and switching -- the vast geographic distances data has to travel can be a factor of delay, especially when the information itself is generated so quickly by computers and is useful only within a very short time period. At least one industry, finance, is starting to chafe at this limit.
"For high frequency trading, light propagation delays are in many cases are the single largest limiting factor to taking advantage of arbitrage opportunities quickly," said study co-author Alexander Wissner-Gross, who a research affiliate of the MIT Media Laboratory and the founder of the Enernetics research consultancy.
As a result, the researchers recommend high frequency trading firms, as well as any organization that deals in complicated time-sensitive global interactions, take a hard look at where they locate their data centers. In other words, location really does matter.
"By positioning an intermediate node, there is the potential to partially skirt that light-speed delay," Wissner-Gross said.
The MIT work was recently published in the Physical Review E scientific journal. The subject, as well as the paper itself, will be the topic of one talk at the High Frequency Trading World conference, being held this week in New York.
High Frequency Trading (HFT) is the financial industry's short-hand term for how they use computers to analyze the market and rapidly buy and sell shares of stocks and other financial instruments. Such firms rarely refine their trading to only a single exchange, so by necessity they route information around the globe to different exchanges.
"There are many types of transaction that fundamentally depend on sources of information from multiple places around the earth," said study co-author Cameron Freer, who is a mathematics researcher at the University of Hawaii at Mānoa (and was at MIT when the paper was written). Some finance firms, for instance, may trade against momentary price discrepancies between the London and New York exchanges.
Typically it takes about 50 milliseconds to send a message from New York to London. Placing a server in between the two could cut the speed of communication in half, Wissner-Gross reasoned, which may be enough time to take advantage of some momentary pricing discrepancy. The hyper-competitive finance industry will continue to cut transaction times wherever possible, which will make light-speed delays more pronounced in the years to come, the paper argues. Light travels at 299,792,458 meters per second, or about approximately 186,282 miles per second.
"As financial trading moves to higher frequencies , it will be more and more useful to organize both networks and the computers that sit on them in unique geographic ways that take advantage of the geographic positions of various stock exchanges in order to coordinate light speed trading between them," Wissner-Gross said.
The researchers came up with a general formula that an organization can use to derive the best location to sit in between multiple sources of information. They did not take into account delays inherent in networking due to equipment latencies, assuming an organization could factor these numbers into their own equations.
Using this formula, the researchers also triangulated the best locations in the world for locating servers that do high frequency trading across 52 exchanges worldwide. A few of the prime locations include spots in the middle of the ocean, or in small countries with as-of-yet underdeveloped communications infrastructures.
Ultimately, this sort of work may bring about new data center hotspots around the globe, the researchers predict. "Geography may be viewed as a new type of natural resource," Wissner-Gross said.
Freer maintains that this work would have implications for other organizations in addition the hyper-fast financial market.
"Any application where it is important for you to coordinate in multiple locations in a low latency way, and the reason you need those latencies because situations can change rapidly at those locations. Our treatment is quite flexible and deals with a large class of systems that are fluctuating in multiple locations and where coordination is necessary," Freer said.