Credit adjustment

Bernd Machenschalk
Bernd Machenschalk
Moderator
Administrator
Joined: 15 Oct 04
Posts: 3,906
Credit: 191,063,801
RAC: 49,949
Topic 193821

Hi!

What people are experiencing here wrt credits is a combination of many things.

1 the average credit was reduced by 25% for S5R4 compared to S5R3 as announced.
2 the claimed credit is currently independent of the granted credit, and found to be much lower atm.
3 the workunits generated at first by the S5R4 workunit generator all at the upper end of the runtime variation for a given frequency band. They should average out over time, but at the beginning we got all "bad" ones.

I intend to fix 2 in the next few days, it's an unwanted side-effect of our efforts to fix the "duration correction factor". I hope to be able to fix 3 in a longer term in the workunit generator. For now I bumped up the credit (for newly generated workunits) manually by 15% (>= 13.9%) based on the following database queries (we gave 28% more credit than SETI at the end of S5R3):

mysql> select avg(granted_credit/cpu_time) from result where name like '%S5R3%' and granted_credit>0 and cpu_time>1000;
+------------------------------+
| avg(granted_credit/cpu_time) |
+------------------------------+
|           0.0095896610820667 | 
+------------------------------+
1 row in set (1.36 sec)

mysql> select avg(granted_credit/cpu_time) from result where name like '%S5R4%' and granted_credit>0 and cpu_time>1000;
+------------------------------+
| avg(granted_credit/cpu_time) |
+------------------------------+
| 0.0061389230693111 |
+------------------------------+
1 row in set (1.42 sec)

mysql> select 0.0061 / 0.0095;
+-----------------+
| 0.0061 / 0.0095 |
+-----------------+
| 0.642105 |
+-----------------+
1 row in set (0.00 sec)

mysql> select 1/1.28;
+--------+
| 1/1.28 |
+--------+
| 0.7812 |
+--------+
1 row in set (0.00 sec)

mysql> select 0.7812 - 0.642105;
+-------------------+
| 0.7812 - 0.642105 |
+-------------------+
| 0.139095 |
+-------------------+
1 row in set (0.00 sec)


We'll keep watching these numbers and adjust credits as needed.

BM

BM

RandyC
RandyC
Joined: 18 Jan 05
Posts: 2,583
Credit: 107,402,225
RAC: 22,837

Credit adjustment

Quote:

Hi!

What people are experiencing here wrt credits is a combination of many things.

1 the average credit was reduced by 25% for S5R4 compared to S5R3 as announced.
2 the claimed credit is currently independent of the granted credit, and found to be much lower atm.
3 the workunits generated at first by the S5R4 workunit generator all at the upper end of the runtime variation for a given frequency band. They should average out over time, but at the beginning we got all "bad" ones.

We'll keep watching these numbers and adjust credits as needed.

BM

Thanks Bernd, appreciate the news updates.

Seti Classic Final Total: 11446 WU.

Brian Silvers
Brian Silvers
Joined: 26 Aug 05
Posts: 772
Credit: 282,700
RAC: 0

@ Bernd My Pentium 4

@ Bernd

My Pentium 4 2.4GHz system just reported in a task.

With the "old" way, the cr/hr calculation is as follows:

65.5923 * 2.92 = 191.53
191.53 * 3600 / 98616.8 = 6.99 (or approximately 7) cr/hr

Increasing by 15% gives:

7 * 1.15 = 8.05

Unless SETI's reduction, which hasn't happened yet, manages to reduce the amount I can get with the optimized application by more than 50% (currently getting 17/hr), then SETI has a "loophole advantage" with the optimized applications.

I would hope that after the disparity between the Linux and Windows apps that you have talked about takes place, further reductions here would not be implemented to reachieve "parity" if such a loophole is present.

Additionally, see my post in the "Thoughts on credits" thread about the implied work level of two identical systems processing 2000 tasks each... here

David @ TPS
David @ TPS
Joined: 4 Jan 06
Posts: 5
Credit: 5,245,634
RAC: 0

Just an FYI, none of my boxes

Message 83809 in response to message 83808

Just an FYI, none of my boxes are accepting R4 work and I am running out of R3's, so the backups of Rosetta and Seti are taking over full time on them.

RandyC
RandyC
Joined: 18 Jan 05
Posts: 2,583
Credit: 107,402,225
RAC: 22,837

RE: Just an FYI, none of my

Message 83810 in response to message 83809

Quote:

Just an FYI, none of my boxes are accepting R4 work and I am running out of R3's, so the backups of Rosetta and Seti are taking over full time on them.

If you have been using the Power or Beta apps, then you will need to adjust or delete your app_info.xml file and restart BOINC. See this thread.

Seti Classic Final Total: 11446 WU.

David @ TPS
David @ TPS
Joined: 4 Jan 06
Posts: 5
Credit: 5,245,634
RAC: 0

Thank you, sir.

Message 83811 in response to message 83810

Thank you, sir.

MAGIC Quantum Mechanic
MAGIC Quantum M...
Joined: 18 Jan 05
Posts: 1,288
Credit: 404,803,183
RAC: 83,692

5 Aug 2008 10:18:07 UTC 7

5 Aug 2008 10:18:07 UTC

7 Aug 2008 21:16:51 UTC

Over Success Done

78,994.63 claimed credit 65.83 and granted credit 192.24

I have turned in 8 so far and all the granted credits go up around that amount.

 

Ed1934158
Ed1934158
Joined: 10 Nov 04
Posts: 62
Credit: 14,481,483
RAC: 0

I'm not sure how is this 25%

I'm not sure how is this 25% decrease. True I get 25% less credit per unit, but it takes me twice as long to complete a work unit than it did for S5R3 units with optimized client.

Gary Roberts
Gary Roberts
Moderator
Joined: 9 Feb 05
Posts: 5,032
Credit: 34,590,832,272
RAC: 34,325,294

RE: I'm not sure how is

Message 83814 in response to message 83813

Quote:
I'm not sure how is this 25% decrease....

Try reading this message which explains very good reasons why a certain average reduction might translate to a big reduction for some but a very much smaller reduction for others.

If you were using the best optimised app for S5R3, you will be using the same for S5R4 (ie no further performance gain). Anyone using the Windows stock app for S5R3 will now be using a better optimised app for S5R4 and will see a performance gain which will cushion the fall due to the average credit reduction.

Cheers,
Gary.

Bikeman (Heinz-Bernd Eggenstein)
Bikeman (Heinz-...
Moderator
Joined: 28 Aug 06
Posts: 3,515
Credit: 446,243,302
RAC: 54,780

RE: I'm not sure how is

Message 83815 in response to message 83813

Quote:
I'm not sure how is this 25% decrease. True I get 25% less credit per unit, but it takes me twice as long to complete a work unit than it did for S5R3 units with optimized client.

There's no way to arrange the credits so that everybody sees the same credit reduction. Note that most Windows users were still using the rather slow stock app at the end of the S5R3 run, which is now replaced by a much faster one. On the other hand, if you were already using an optimized app at the end of the S5R3 run, the stock app in S5R4 is essentially the same as the power app before.

If credit was reduced in a way that the optimized app users would see only a 25% credit reduction, actually most Win users would see a credit boost (!!) because the performance gain for the stock Windows app is far more than 25 %. Instead of closing the gap in comparison to other projects, the gap would actually widen a bit!

So to bring the *overall* credit rate down by a factor of 1.25 or 1.28 or whatever, credits have to be adjusted in a way that power app users will see a greater reduction, and stock app users will see less of a reduction (relatively).

Sounds unfair at first, but if you think about it, it's perfectly fair. Everybody is now using "optimized" apps because they were made the stock app. and of course credit cannot and should not be awarded depending on what app you happened to use in the previous run :-).

CU
Bikeman

EDIT: Gary was faster and more economical in making a reference :-). I hope by rephrasing everything a bit I made it a bit clearer.

Ed1934158
Ed1934158
Joined: 10 Nov 04
Posts: 62
Credit: 14,481,483
RAC: 0

Ok, thanks, I understand now.

Ok, thanks, I understand now.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.