The apps that we're running right now control it pretty well using a sneaky trick. We can't get much computer information from java in an unsigned app, so we measure how long each sub-part of a work unit has taken and sleep an appropriate amount of time. For example, if we are trying to use 25% and a sub-work takes 100ms, we sleep for 300ms. This effectively gives us 25% usage of the available CPU. It will be no higher than that anyway - it might be lower if something else is using substantial CPU time.
As far as the fast versus slow nodes, we normalize everything based on the average computer that completes work each day. So a fast computer may earn 2X or more what a slow computer earns.
As far as the fast versus slow nodes, we normalize everything based on the average computer that completes work each day. So a fast computer may earn 2X or more what a slow computer earns.