While FPGAs are well known for providing deterministic low latency, they are also known to be difficult to program, creating an enormous barrier to flexible and fast change. The Cheetah Framework provides building blocks which can be configured in real time by your host server. No longer is your alg crossing busses and looping into the software domain. Instead, the algorithm can execute entirely in the FPGA-enabled network card; and the server software takes only a supervisory role, monitoring the alg's performance and adapting the hardware alg on the fly.

Cheetah Solutions provides out of the box components such as EMA, MACD, ATR, ADX, Arithmetic, Logic, but also provides the tooling to easily create your own. The Cheetah Framework provides real-time interconnections between such components as well as reporting and auditing functions. Your host server defines the "wiring" between the components as well as the individual configuration of each component.

All of this has been tuned to run FAST - the EMA component calculates 38 million price tick EMAs per second on up to 255 instruments with a latency of 144ns; if you want to look up the current ADX value of an instrument, you can access this from the ADX component in 24ns.

"True low latency will only be acheived through total hardware solutions which garantee deterministic low latency" explains Peter Fall, CEO of Cheetah Solutions. "But if the market moves, you want to be able to tweak an algorithm or change it completely to take advantage of current conditions. Traditional FPGA programming may take weeks to make even a simple change whereas Cheetah Framework provides on-the-fly reconfigurability"