In order to have the c1mcs model run, we're running with only 3 RFM channels between c1ioo and c1mcs at the moment.  This leaves the model at around 45 microseconds, and at least lets us damp.

Alex and I still need to track down why the RFM read calls are taking so much time to execute.