Saenger, what follows is Eric K's first post to a topic which is being discussed. I think they're working on it.

tony

THX Tony!
That's good news!

I put together a little spreadsheet of Claimed Credit values and sorted it by CPU type. Two thirds of this is from Rosetta users and the other third by Seti and Einstein users. I opened the most recent threads at the time and just started working down the list of posters and adding their values to the database. NOTE: there is also listed users of "Optimized Boinc Versions" in the list such at 5.5.0, 5.3.6, and some 5.2.13 crunch3r's in there, but if you stick with 5.4.9 and other "non optimized clients" you'll get an idea of what Claimed Credit was "supposed" to be with the benchmark system. The different boinc versions are shown with different colors. If anyone wants this just email me and I'll ship it to you. yes the first two column are links so someone can check my figures.

After gathering data from all my projects, I put together the following comparison. The number of data points is small so take it for what it is. I'll be continuing to add datapoints until I get a more rounded picture. All hosts/credits are from standard Boinc Clients (no optimized apps either in the comparison). Note: there is a comparison between 5.15 and Simon's app near the bottom. "Go.... Simon!!!"

I basically think Eric has the "load store adjustment" (fpops multiplier) about right, maybe just a snidge low. We'll see what happens when he adjusts it upwards as he's stated that he'll be doing. Einstein seems to be granting 50% more than they should (I.E they should cut the multiplier by 1/3rd). Primegrid uses only a Quorum of 2 and averages those two claims, so "optimized" boinc core client users are increasing my granted credit there (not that I'm glad about it).

Some interesting figures.

---

Grant
Darwin NT

Looks like things could get worse when comparing to Einstein. The new beta version there is cutting crunch times to 2/3 of present standard app.
It means that my Pent M now gets 27 to 28 credits/hr.

Andy

Looks like things could get worse when comparing to Einstein. The new beta version there is cutting crunch times to 2/3 of present standard app.
It means that my Pent M now gets 27 to 28 credits/hr.

Note that this credit is for the beta only. Bernd has posted that the credit scoring will be 'adjusted' when it goes live, "to stay conformant to the other projects".

I've updated my charts (see old one below/above) to include Seti beta with the new multiplier, and to include both old and new Rosetta credits. I've heard Einstein is dropping their credit some, but my data for einstein is not newer than 7 August.

NOTE: GC is claimed credit, GC is granted credit, GWC is granted work credit (the new rosetta term, for their fixed credit system)

Hello Tony,
[nitpicker mode] it's Einstein, not Einstien [/nitpicker mode]

But thanks for the work done!

Hello Tony,
[nitpicker mode] it's Einstein, not Einstien [/nitpicker mode]

But thanks for the work done!

Thanks, I was probably Sober, when I typed that.

LOL

Yes please I'd like a copy of the spreadsheet to have a deeper look at. Where do I get your email address from?

Chris

Chris, contact me through an.arbitrator (at) gmail dot com. John will send me your email and I'll ship it to you.

It looks like QMC will be moving to serverside credits in October, and they will be benchmarking against Einstein.

---

Evening All

I am still collecting Data for Cross Project Calibration... At this point more projects are becoming a bit more Serious about Fair Credit across the Projects.

about Mid week, I will finalize what I have collected and post soem information or links to web pages for everyone to see. I am fortunate that I have both Intel and AMD's to work with... Between what Tony has posted at various locations and what I will post it will give a broader picture...

Regards

Pappa

---

Please consider a Donation to the Seti Project.

OK folks, it's been sometime since my last update. Many projects have been added and more results have been added to the projects I've previously collected data on.
My P4 1.8 has died and beeen ressurected as the AMD X2 4800 you now see. All data is from standard Boinc clients. This time I've included the data from optimized
applications to the charts/tables below. The individual xls files and charts are available for download/inspection upon request. There is one xls file for each project
and one worksheet for each host.

I'll draw no conclusions here and simply present what I'm seeing, leaving it up to you to fish out what you may see. (I.E which are giving more/less than what the
benchmark would indicate, and which are credit rich/richer)

This data is constantly changing and being updated. I wanted to wait till I was done, but quite frankly the more I do, the more I see needs to be done (like
breaking project app versions down), and I'm beginning to think it'll never end.

---