@CraigyD Map/Reduce should definitely reduce that time as it does processing in parallel; how much reduction depends on how many processors you use and how you're going about the data processing