> Edit:
>
> Rethinking what I posted


24 Hours later...

I think he posted something, then wasn't sure if he agreed with it any more...

---

> I think he posted something, then wasn't sure if he agreed with it any more...

Could be. Did it ever happen to you ?
I have spent the whole last night thinking about a radio problem
before posting the answer Here this mornig. It really feels good.
I hope it is the same for him.

System 1:
AMD 2400+
Gentoo linux 2.6
1 GB RAM
Average credit over the last 35 work units: 35.51

System 2:
Intel Pentium-M 1.3 GHz (laptop)
Windows XP SP1
768 MB RAM
Average credit over the last 35 work units: 42.16

Total average credit (70 work units): 38.83

The high average on the Pentium-M surprised me. This hasn't been complained about but I think the benchmark code also does not work very well on the penium-m CPUs. The 1 MB of L2 cache seems to help out a LOT on crunching work units but I don't think it has much bearing on the benchmarks. So the result is off but I think it works out so that the machine requests MORE credit than it should so no one complains about it whereas the linux boxen claim LESS credit so everyone whines :)

---

A member of The Knights Who Say NI!
For rankings, history graphs and more, check out:
My BOINC stats site

>
> The high average on the Pentium-M surprised me. This hasn't been complained
> about but I think the benchmark code also does not work very well on the
> penium-m CPUs. The 1 MB of L2 cache seems to help out a LOT on crunching work
> units but I don't think it has much bearing on the benchmarks. So the result
> is off but I think it works out so that the machine requests MORE credit than
> it should so no one complains about it whereas the linux boxen claim LESS
> credit so everyone whines :)
>

I have a Windows XP machine which consistently requests over 100 credits per WU! I'm a Mac guy so I don't really get into the Windows world much, it's running on machine that we have at work simply for accounting crap (which we can't do on a Mac).

---

-Aardvark

Yes, the theory of "avarage credit" is fine.
The distribution of claimed credits (defined by any measurement system) on the computer population defines an average claimed credit. This fact and the validation system mean that the more processed WUs the more precise fit to this average. (I could explain this in a more detailed way but it is not important.) This average is moving very slowly by some reasons.
So the granted credit depends on this average not the properties of your machine (if you process "enough" WUs).

(You can check this thread:
http://setiweb.ssl.berkeley.edu/forum_thread.php?id=5843)

---