[Billy]

This WU has the same Mac Core Duo computers. One is running the standard app, the other, the Alex Khan optimized app. The optimized app is underclaiming by about 30%.

Shouldn't the amount claimed be the same? Equal claim for equal work?

[beadman]

I notice the one that claims the higher amount has 2 GB of RAM, and the other only has 1 GB - is it possible the one with less RAM had the -9 error earlier because it had less RAM?

beadman

[QSilver]

Originally Posted by Billy 

This WU has the same Mac Core Duo computers. One is running the standard app, the other, the Alex Khan optimized app. The optimized app is underclaiming by about 30%.

Shouldn't the amount claimed be the same? Equal claim for equal work?

That WU has already been purged from the results DB, but here's my guess: The Mac Core Duo running the standard app used v5.23, which over-claims. At one point during the development of the new science apps to handle multi-beam data, Berkeley thought that they would raise the credit mulitplier; instead they lowered it.

The standard Mac multi-beam app over-claims because it was released during the period when a higher credit multiplier was expected to be in use. When the "final" multi-beam apps were released for other platforms (i.e., v5.27), they had a multiplier that was lower than what had been used historically, and significantly lower than what was in v5.23. When AK re-released his latest and greatest, he adjusted the credit mulitplier to be equal to the S@H standard in v 5.27.

Even though the standard-issue Mac application over-claims by fairly large margin, there has been no word from Berkeley as to when an app with a corrected mulitplier will be available.

QS