The new even faster applications are now used for all the production runs. Fermi cards are also supported for Linux and Windows by the CUDA3.0 versions.

GDF

"Even faster" than which ones? Did you apply the 30% increase (introduced what? 1 month ago?) to all applications or is it a brand new 30% increase?

And, about the "full atoms", wasn't they supposed to be finished and substituted by the 6.72 versione?

I am a bit confused, could you make it clear please? :)

If the application is faster and more work is being done, why do credits seem to go down at the same time? Unless I just got a very small WU, this seems to happen. If there is a progress in computing, the project should become more attractive and more rewarding at the same time.

---

@Venturini
Old lines of WUs need to be completed on the old application so for a while there were both the old 6.03/6.04 and the new 6.72/6.73. There was also a version specific to fermi for a while. They have all been rolled into 6.07/6.08 and there is a new beta about to be released. Yes, it can get confusing but overall it seems to work OK.

@Mad Matt
When the apps get faster then tasks return faster so you finish more in a day and your PPD gets better automagically.
The example I assume you are looking at is a TONI_GA type which does typically run faster than the other types. GPUGrid is persuing multiple lines of research and some take longer to calculate. That along with the fast return time bonus is why the points for all WUs are not always exactly the same.

---

Thanks - Steve