I think a programmer familiar with the Rosetta code should first look With the introduction of double-precision floating point in newer video cards, precision should no longer be an issue with GPU processing.

The question is... why Rosetta is not able to increase the CPU power for the project?

R@H is not able to maintain volunteers. The new volunteers incorporated to the project has more CPU power but they spend less time with the project

Hi! I am founder of TomaszPawelTeam!

I would like to reserve so that nobody can never take control of my squad. How I can do this? Even if for some reasons I won't be active for long time?

Sorry, you can't do that. If you're inactive for a long time and someone wants to take over the foundership duties, I think it's fair for that person, particularly if the founder cannot be contacted.

From my own testing using a system with exactly the same components, just changing the memory speed, it has no effect on R@H.

Tested on a Phenom 9850 at 800MHz and 667MHz. There was no drop in the average credit granted between the two.

What does affect R@H pretty significantly however, is the amount of L2/3 cache your CPU has. My 9850 (RIP) only had of 4MB L2/3 cache total. I swapped it out for an Intel Q9450 which has a slower clock speed but a massive 12MB of L2 cache (4MB per core!) and the amount of decoys I was able to generate per work unit shot up resulting in a much higher RAC :)

I'm looking into this. Don't really know what is going on.

You can EMail me the details on the link. Thanks in advance for reporting it.

As for RAC, RAC is based on credit history, and credit was not issued as quickly as results were being reported during that timeframe. This is why the estimate project teraFLOPs dropped during that time as well. But then you see a spike, and then the normal trendline. Please post again in a couple of days if your figures don't look back to normal by then.