Yeah, we ended up writing the graphing library ourselves. Due to the nature of the data and frequencies of redrawing the large open source libraries were too CPU intensive.