Architectures for Extreme Low Latency
In recent years, a combination of regulatory reforms, market structure evolution and technology innovation have combined to create a new paradigm for the global financial trading markets; one in which the speed at which transactions can be executed is fundamental to the success of market participants.
Put simply, for such a participant – be it an exchange operator, a broker or a trading firm – to win in the new electronic marketplace of complex algorithmic and high-frequency trading (HFT), they have to be faster than their competition. And the margin that separates the winners from the also rans is now measured in microseconds, with nanoseconds not too far off.
Within the IT platforms upon which financial trading systems run, the speed at which data can be processed and exchanged between components directly affects the efficiency and overall speed of those systems. And the phenomenon that limits this speed is known as latency. Minimising latency – to a point where it is close to eradicated – is the design goal for these platforms.