Implementing a low cost, low latency parallel platform