hi,
after running it almost 24 hours, the credits sum is much lower then I expected.
Teemu: can you describe in details how credits are given?

I see:
[May 03 00:24:36 UTC] RC5-72: 4 packets (256.00 stats units) are in out.r72
results in 414.81


[May 02 23:27:09 UTC] RC5-72: 4 packets (254.00 stats units) are in out.r72
results in 411.57

which seems to be: 0.617149 par stats units.


Is is mainly based on CPU time?


All other GPU projects is giving much more credits. I saw other users complaining about this "problem". Do you consider a little increase?

I have looked at some of the other messages as well - agree that credits are all over the place and looks like I'm not the only one getting huge differences. 400 credits for an hour of GPU is not so good - about the same as seti under ati!?

The first couple of WUs were around 400 seconds for 400 credits (ok), but since then it looks pretty sad. I understand some tinkering for the new project is required, but some notes/guidance would be useful.

I wonder if there is some boinc politics regarding credits?

I know with dnetc, that machine was doing 600 000+ in credits with dnetc (which is 10X the current credit).

Some other person feel the credits should be higher too?

Hi,

We are using the new credit system of the latest BOINC server code. It has quite complex algorithms and scales the credit according to statistical host and application performance. It's supposed to make sure credit is granted fairly (eventually as it learns averages).

This credit system is detailed at http://boinc.berkeley.edu/trac/wiki/CreditNew for those that are interested in technical details. It's the third credit system on that page.

We've had some problems that resulted in some results getting either too little (around 400cr) or too much (over 15kcr) for similar work. And yes, there was that "almost zero credit" incident too. :)

Some of our problems could be due to bugs in this new credit system or are result of our estimated and, especially, limit values for work units. I've increased the limit during early days (fixed the 400cr) and lowered it few days ago (should be at least a partial fix for the over 15kcr).

In addition to the standard credit calculation we now also have a partial crediting system in place (still need some edge cases fixed, though) that gives credit for tasks that haven't done the full amount of work in a work unit due to some local failure (CUDA errors, bugs in the dnet client. This partial credit is based on second task that was fully done and validated and is granted relative to the done vs. full work.

-w