Rosetta running on fastest cluster in world

and all off a Intel® Pentium® 4 CPU 2.26GHz

Yeah right :jealous: (honest :nono: )



Look at those benchies!

but only one CPU? :confused:

It must be reporting as only one CPU, weird as that may be.

I’m jealous too…

This is a rogue machine…but, I don’t think this is a case of deliberate over-inflated benchies. Why go the to lengths of borging the benchmarks, if you are going to make it so damned obvious?

It seems to me more like the box has blown a gasket. It appears a few days ago that box was reporting as normal with reasonable claimed credit, using the Truxoft calibrating client. Then, there were several units that had computation errors over a period of a few days, which were followed by units with massive claimed credit due to the huge reported benchmarks. Finally it appears that the client has crashed again.

I’ve seen something like this in the past when the Truxoft client gets messed up. The reported benchmarks become corrupted, which is not mormally a problem due to Boinc quorums and validation. The problem on Rosetta is that there is no quorum or validation, so the cruncher gets what they claim, even if it is ludicrously wrong.