Hmm. Well mine is a PPC but only getting about a third of claimed credit. Could it be an older cpu (G4) and older OS (panther)?

In the end, I don't really care - I don't do this for rank or anything. As long as I see SOME credits coming in, it doesn't much matter how many. Mostly, I'm just curious.

Oh, one more question. So was I wrong that macs generally got less than they deserved under the old credit system?

I'm not a Mac person. I owned an Apple II+ and made fun of those silly people that had to use a mouse to do anything. I eventually moved to the PC world, and now look where I am. David Kim's PPC Mac actually mentioned the PPC client in the results. And your system hostid=199825 is a Powerbook6,3 running Darwin 7.9.0. If hostid=301183 is your PPC Mac, then it should be doing better than 199825.

Macs weren't getting less credits under the old system based on what the Rosetta applications are able to produce on a Mac; as they look like they're producing 1/2 to 1/3 of the data their benchmarks claim they're able to; although if another Boinc project's benchmarks results in a higher number of credits/day than you were getting on Rosetta then feeling short changed by the standard Boinc benchmark on Rosetta starts making sense.

---

CPU is 800MHz G4 and 256MB memory. I'm not sure what MenuMeters is but no, memory is not maxed out when running BOINC.

MenuMeters is a great, free little app for the mac. All Mac users should have it, IMO. It lists all the critical stats in the menubar. Get it here.

For what it's worth, this is the memory consumption I see, running *only* BOINC. If you are running other applications too, your memory requirements will go up.


270212: G4 PowerBook 1.25ghz running Tiger (10.4.7) uses 185mb = RAC 90

185730: G4 iBook 1.33ghz running Panther (10.3.9) uses 170mb = RAC 111

---

Reno, NV
Team: SETI.USA

I ran a load test on two identical computers over the weekend (Celeron 2.66GHz, 768MB RAM, XP) using Rosetta. The results are:

PC1: 509.19 credits, RAC = 44.98
PC2: 508.40 credits, RAC = 44.99

I know it was only a short test, and they are running identical hardware/software but comparison of the same hardware and OS is where I would expect the old system to give more consistent credits (assuming the same version of BOINC is used!) as it only relied on benchmark * time.

The new system has passed that test very well - I'm impressed!

---

To follow up on what dcdc did over the weekend. . I checked a couple of my machines. They are cash registers by a company called NCR. . you don't get more identical hardware :)

They are 2ghz celerons with 256mb of RAM. They are also being actively used, about half the day they run our various restaurants.

The following results are from the past 6.5 to 7 days worth of work units for each register. I have about 20 or 30 registers going, these are the ones that have been running the longest.

Machine 274086
146.71 credits/day

Machine 274090
145.39 credits/day

Machine 274095
146.37 credits/day

I'd say the system is very good since the results for identical machines falls within 1% of each other.

---

Anything new on the issue failed wu-s ?

Is there any thoughts on how to grant credit for faild Wu-s?

Anders n

---

Some recent examples:

My P4 1.8 died and was replaced with an X2 4800, and I'm cycling it through the projects and adding that data. All comparisons below are from stock Boinc clients. I've added data for Einstein, VTU, Chess960alpha, Xtremelab, QMC, and Spinhenge. The blue is "avg Claimed Credit/hour/core" and the purple is "avg Granted Credit/hour/core". The very first blue one is the "calculated credit/hour" as figured with the official boinc benchmark.

Thanks for pointing out that Einstien and PrimeGrid are good for credit, i'll see if my team want to switch over to that from Rosetta for a bit to move up the boinc overall :-)

Ta.

Any idea how long they intened to do that for, we may have a small marathon for a bit.


Also do you know which give claimed credit (Simap has just swithed back).
Xtremlabs should not be included, they just crant what you claim and they shouldn't be using a time*benchmark anyway but just a time (due to the fact a 500Mhz PC does the same work as a 4GHz PC).

---

Team mauisun.org

Einstein has been constantly changing their fixed credit system to bring it into parity (or atleast that's what is declared). They have reduced credit atleast twice since the implementation of their newly optimized application. They have their own credit complainers who bemoan each cut in credit. To me it seems as if they are sneaking up on parity instead of taking a big leap towards it.

I attached to Tanpaku and they seem to be (with my small sample so far) giving more credit/time than Primegrid. My best guess is the "saturation percentage" of optimized boinc clients is higher than at Primegrid. Primegrid "averages all claims" in the quorum. Instead of deleting the high/low claims.

If I were to pick a project based upon "granted credit/time", then I'd pick WCG, Einstein, Tanpaku, then Primegrid in that order. Note: I haven't run CPDN or it's offshoots but I believe (though I can't prove) that they'd be on that list somewhere.

---

Einstein has been constantly changing their fixed credit system to bring it into parity (or atleast that's what is declared). They have reduced credit atleast twice since the implementation of their newly optimized application. They have their own credit complainers who bemoan each cut in credit. To me it seems as if they are sneaking up on parity instead of taking a big leap towards it.

I attached to Tanpaku and they seem to be (with my small sample so far) giving more credit/time than Primegrid. My best guess is the "saturation percentage" of optimized boinc clients is higher than at Primegrid. Primegrid "averages all claims" in the quorum. Instead of deleting the high/low claims.

If I were to pick a project based upon "granted credit/time", then I'd pick WCG, Einstein, Tanpaku, then Primegrid in that order. Note: I haven't run CPDN or it's offshoots but I believe (though I can't prove) that they'd be on that list somewhere.

WCG is a great project. And it has become the proving ground that all the slanders/accusations and unadulterated ... said against XtremeSystems are lies. We as a team are having the best dailies of any team bar none. Not only are we the better team under the triple quorum , we are doing it using the standard Boinc Client.

This is the opportunity situation for the Boinc Based WCG as posted today in Free DC Stats:

Rank Team Score Average Daily Gain Days to Overtake
13 Rochester Institute of Technology 2,302,874 9,201 103,220 2.81
12 IBM 2,420,016 13,042 99,378 4.09
11 Slashdot Users 2,568,584 13,702 98,718 5.63
10 L'Alliance Francophone 2,586,274 18,349 94,072 6.09
9 MyOnlineTeam 2,626,115 13,659 98,762 6.21
8 IBM ITCenter Mainz 2,853,882 17,921 94,500 8.90
7 SWARH 3,389,877 37,189 75,232 18.30
6 EIC Limited 4,093,203 0 112,421 18.50
5 UserFriendly.Org 4,206,716 36,506 75,915 28.89
4 IBM 26RA LoveTeam 5,083,137 45,000 67,421 45.53
3 Gay US 6,395,079 36,334 76,087 57.59
2 Clubic 8,646,620 44,200 68,221 97.23
1 Easynews 15,738,014 86,695 25,726 533.51

So what is going to be the excuse now of those who belittle XtremeSystems? We are a small team, the project has a triple quorum and we are using the Standard Client. And yet, by next week we are going to be among the top 10. And now we are producing better dailies than teams 2 , 3 , 10 times bigger than us.

I hope the moderators do not erase this one. People that read the Rosetta Boards need to know that those who said XS couldn't produce big under a quorum system or using a standard clients and thus we needed to "cheat" [their words nor mine] are being proven wrong in a project they highly recommend.

Also do you know which give claimed credit (Simap has just swithed back).

The change back for SIMAP is just temporary. It seems that they have figured out a way that does not rely on claimed credits or the benchmark function (like we still do even under the new system). Perhaps we should look into how they are doing it, and see it it would make sense to use here.

---

Reno, NV
Team: SETI.USA

Also do you know which give claimed credit (Simap has just swithed back).

The change back for SIMAP is just temporary. It seems that they have figured out a way that does not rely on claimed credits or the benchmark function (like we still do even under the new system). Perhaps we should look into how they are doing it, and see it it would make sense to use here.

Maybe they (and we [R@H] and other projects) could report back to BOINC the way of giving credit so it can be published on their site so other projects can get ideas when they setup or even need/want to find one suitable.
[Something for Rom's Q&A maybe :-)]

---

Team mauisun.org

I haven't run WCG in many many months. So the puters I had attached were:

my AMD64 3700, Mobile AMD64 3700, AMD64 2800, and P4 1.8. I completed 53 WUs and heres what it says:

Avg. Points Per Hour of Run Time 73.21890

This is across all machines which would have claimed an average of 11.1825 credit/hour (13.33, 14.14, 9.61, and 5.37 claimed credit/hour). This is why WCG is tops on the list for those who crave credits.

Avg. Points Per Hour of Run Time 73.21890

WCG points or Boinc points Tony, cos average machines get between 10-12 points an hour there at present. Which is average for Boinc platforms.

And people go there first for project and second to get away from snide jibes at others that is a forte with this place.

---

Not all Czech`s bounce but I`d like to try with Barbar ;-)

Make no mistake This IS the TEDDIES TEAM.

Avg. Points Per Hour of Run Time 73.21890

WCG points or Boinc points Tony, cos average machines get between 10-12 points an hour there at present. Which is average for Boinc platforms.

And people go there first for project and second to get away from snide jibes at others that is a forte with this place.

To be honest Carl, I didn't know there were two different systems. I see this when I click on "my Grid".

Totals:
Total Run Time (y:d:h:m:s) (Rank) 0:008:10:17:51 (#151,017)
Points Generated (Rank) 14,812 (#112,472)
Results Returned (Rank) 53 (#111,290)

Averages:
Avg. Run Time Per Calendar Day (y:d:h:m:s) 0:000:00:55:41
Avg. Run Time Per Result (y:d:h:m:s) 0:000:03:49:01
Avg. Points Per Hour of Run Time 73.21890
Avg. Points Per Calendar Day 67.94495
Avg. Points Per Result 279.47170
Avg. Results Per Calendar Day 0.24312

Miscellaneous:
Last Result Returned (UTC) 08/02/2006 10:12:35 [1,087+ hour(s) ago]
Device Installations 6

To be honest Carl, I didn't know there were two different systems. I see this when I click on "my Grid".
....

Right. Now compare that to what shows on BOINCstats. Quite a difference, eh? Internally, WCG takes the BOINC number, and multiplies it by ~7 to try to make the number competitive with the credits claimed by the non-BOINC windown-only client (UD). 85%+ of the WCG users are on the non-BOINC client.

However, what gets reported externally to BOINC stat sites is the normal BOINC credits. Those sites do not see the UD generated credits.

You can read all about it here.

So, no, people do not go to WCG for credits. In my experience, they were slightly less than most other projects.

---

Reno, NV
Team: SETI.USA

Sure enough, my sig (below) shows 2115. Divide 14,812/2115 and we get 7.0033097 as the multiplier. Learn something new today. thanks

Now if I could only see each individual wu and it's data, I'd be all set.

Click "My Grid" " Device Statistics or Result`s stats (can`t remember which)" and all the statistics are available for each unit..That is run time, claimed and given credit etc.,

Stat`s are updating at present, give it 1/2 hour!

---

Not all Czech`s bounce but I`d like to try with Barbar ;-)

Make no mistake This IS the TEDDIES TEAM.

How long before RACs have no old-method credits in them?

---

Reno, NV
Team: SETI.USA

How long before RACs have no old-method credits in them?

Never. However after 2 to 4 weeks the effect of the old method credits should be negligable.

---

BOINC WIKI

BOINCing since 2002/12/8