KillBait said:
One of the major changes with Enchanced has been credit is now given based on how many Flops is takes to crunch through the unit, making optimized versions of the boinc client is pretty much redundant as the benchmarks are no longer used.
Rather the opposite... Under time based credits, the optimized apps did the work faster and claimed
less credits per WU than the lumbering stock app (unless you used trux's correctinc BOINC). Because of the averaging of cliamed credit to granted credit, and most users running the stock app, the overall credit granted was high.
The optimized apps though got, and get, through more WUs per day and thus get more credits per day.
I'm still crunching Beta project units and they claim about 110 points each, the caveat to this is that the beta version of 5.12 for windows im running claims too much credit and it should be about 50 - 60 per unit.
On the corrected figures i should get ~320 per CPU per 24hrs, so your figures do sound a little off, but then i dont know if they again changed the ratio of flops/credit before releasing it to the main project.
Overall, despite the fact that the playing field is much more level and everyone is in the same boat, I think Berkeley have made a disasterous mistake in not maintaing closer parity of pre-enhanced and enhanced credits.
We had just about finished the discussion/whinging about the fairness and unfairness of the benchmark credit system, and everyone who was going to leave BOINC/SETI had done so.
Now it's changed again, the same comments/whinges are back with avengance and another round of retirements seems inevitable.
Once again Berkeley have completely missed a considerable part of the incentive for running the science projects.