variation in credit allocation

Message boards : Number crunching : variation in credit allocation
Message board moderation

To post messages, you must log in.

Previous · 1 · 2

AuthorMessage
Tomáš Brada
Project administrator
Volunteer developer
Avatar

Send message
Joined: 3 Feb 19
Posts: 505
Credit: 407,379
RAC: 525
Message 3565 - Posted: 3 Sep 2019, 7:49:10 UTC

8cd52af Increase credit multipliers.

You are not asking for changes. None of you posts ask for increase in credits. You are only complaining. Only word Please was posted by me. You should learn some manners and try using the magic words.
That your tasks take 6 to 9 or more hours to finish is because there are problems on your computer. My cpu is only about 25% better per thread so you should be seeing runtimes of 3 hours. You can recompile the app if you think error is in there.
ID: 3565 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
LCB001

Send message
Joined: 16 Feb 19
Posts: 2
Credit: 3,309,365
RAC: 22,025
Message 3566 - Posted: 3 Sep 2019, 19:13:40 UTC - in response to Message 3565.  


You are not asking for changes. None of you posts ask for increase in credits. You are only complaining. Only word Please was posted by me. You should learn some manners and try using the magic words.

I don't normally post in the forums but your arrogance is astounding.

We are providing you with free computing power for your project and you complain because we don't use please when asking for a better credit rate.

Your volunteers will treat you the way you treat them.

And you can go stuff your magic words.
ID: 3566 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Sphynx

Send message
Joined: 17 Feb 19
Posts: 7
Credit: 1,019,594
RAC: 0
Message 3567 - Posted: 4 Sep 2019, 2:04:34 UTC

Normally I have 2 goals when a project comes out. Minimum 100 hours of run time to get a star at wuprop and minimum 1,000,000 in credit. It is not unusual for me to shoot for 25, 50 or ever 100 million on a project, but it is dependent on fair credit for work done.

You'll notice that I stopped running your project after reaching 1 million and it was because of the low credit. It took me over 59,000 cpu hours to reach that total. You can see a better indicator that many are doing the same over at Free-DC stats page https://stats.free-dc.org/proj/pad
Note that most of the top producers are stopping at just over 1 million and that team totals are pitiful. Unless credit is increased this trend will probably continue and you will be chasing new people as long as your project is in operation. The solution is easy, though there probably won't be a "please" attached to it.
ID: 3567 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Natalia Makarova
Project scientist
Avatar

Send message
Joined: 8 Feb 19
Posts: 279
Credit: 0
RAC: 0
Message 3568 - Posted: 4 Sep 2019, 5:15:27 UTC
Last modified: 4 Sep 2019, 5:17:44 UTC

Messrs!

You are aggressive.

I note that Tomáš Brada is a voluntary developer.

Let's ask Tomáš Brada: what kind of funding does he receive?
Who pays for his electricity? Who pays for his internet? Who pays for his work?

There are no more enthusiasts in the world?
Enthusiasts in Russia - Progger, Demis, enthusiast in Italy - S. Tognon, enthusiast in Slovakia - Tomáš Brada, enthusiast in Austria - XAVER.
It's all?

Yes, Tomáš Brada may be mistaken with credits.
We all make mistakes!
But this pressure does not resolve the issue.
ID: 3568 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Natalia Makarova
Project scientist
Avatar

Send message
Joined: 8 Feb 19
Posts: 279
Credit: 0
RAC: 0
Message 3569 - Posted: 4 Sep 2019, 5:30:46 UTC - in response to Message 3557.  

I believe this project is the lowest paying of any BOINC project. ODLK and ODLK1 pay very poorly in comparison to other projects but they pay far better than this project.

Bryan
I got your PM.
Thank you for your detailed clarification, but they are useless to me.
I cannot increase credits in ODLK and ODLK1 projects.
This is the prerogative of administrators.
ID: 3569 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Tomáš Brada
Project administrator
Volunteer developer
Avatar

Send message
Joined: 3 Feb 19
Posts: 505
Credit: 407,379
RAC: 525
Message 3570 - Posted: 4 Sep 2019, 13:25:26 UTC

Please give me feedback on the increased credits. Increase even more?
ID: 3570 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile HK-Steve

Send message
Joined: 19 Aug 19
Posts: 1
Credit: 442,767
RAC: 0
Message 3571 - Posted: 5 Sep 2019, 12:54:28 UTC - in response to Message 3570.  

Please give me feedback on the increased credits. Increase even more?


Any increase in points is a good thing. Honestly I think it should be higher again.
How much higher I will give you some more suggestions after I have a think, and read of the CreditNew is defined by Berkley link that you posted.

I have completed over 5200 tasks so far, and will continue to support Your Project.
All the best.
ID: 3571 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Tomáš Brada
Project administrator
Volunteer developer
Avatar

Send message
Joined: 3 Feb 19
Posts: 505
Credit: 407,379
RAC: 525
Message 3572 - Posted: 5 Sep 2019, 17:24:34 UTC - in response to Message 3571.  

and read of the CreditNew is defined by Berkley link that you posted.

Please note that we are NOT using credit new (as explained by my multiple posts).
ID: 3572 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Tomáš Brada
Project administrator
Volunteer developer
Avatar

Send message
Joined: 3 Feb 19
Posts: 505
Credit: 407,379
RAC: 525
Message 3573 - Posted: 5 Sep 2019, 20:12:02 UTC

Also I do not want to attract much more crunchers to this project. It is running off a Banana Pi and single 5400 disk and it's purspose is to develop this application (tot5/padls total) and my other endeavours. One day maybe Ice00 or Progger imports the developed app into their server.
ID: 3573 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Dirk Broer

Send message
Joined: 21 Feb 19
Posts: 7
Credit: 350,607
RAC: 3,151
Message 3576 - Posted: 6 Sep 2019, 0:32:10 UTC - in response to Message 3573.  

Also I do not want to attract much more crunchers to this project. It is running off a Banana Pi and single 5400 disk


Amazing. Which model Banana Pi deserves your trust? A M2 Ultra, a M3, a A64?
ID: 3576 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Natalia Makarova
Project scientist
Avatar

Send message
Joined: 8 Feb 19
Posts: 279
Credit: 0
RAC: 0
Message 3577 - Posted: 6 Sep 2019, 2:16:17 UTC - in response to Message 3573.  
Last modified: 6 Sep 2019, 2:29:28 UTC

One day maybe Ice00 or Progger imports the developed app into their server.

Do you have hope?

This will never happen!

ice00 wrote to me on July 14th
I could probably be able to add it in ODLK1 only if Tomáš Brada will
help me in setting it.

Two months have passed. Nothing done.

Progger has been ignoring me for over a year. He will not run my experiments.
ice00 at one time imported a project by Progger.
In my opinion, so far, tasks for the ODLK1 project are generated by Progger.

I have been waiting for the start of my experiments in ODLK and ODLK1 projects for more than a year.
This did not happen!

After my despair ... if they want to run my experiments, they will have to do it without me.
They seem to want this. I will not bother them anymore ...
ID: 3577 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
mikey
Avatar

Send message
Joined: 28 Aug 19
Posts: 7
Credit: 578,871
RAC: 0
Message 3578 - Posted: 6 Sep 2019, 4:02:01 UTC - in response to Message 3573.  

Also I do not want to attract much more crunchers to this project. It is running off a Banana Pi and single 5400 disk and it's purspose is to develop this application (tot5/padls total) and my other endeavours. One day maybe Ice00 or Progger imports the developed app into their server.


You could have said that in the beginning instead of this long drawn out thread!!

Oh and my dad used to demand 'magic words' when I was a kid, when I turned 18 and moved away I vowed to NEVER mandate anyone ever use them again. Now that I'm 66 years old that's one of the few vows I've kept, well that and my marriage vows that have now gotten to 36 years!!
ID: 3578 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
mmonnin

Send message
Joined: 16 Feb 19
Posts: 20
Credit: 1,456,814
RAC: 0
Message 3580 - Posted: 7 Sep 2019, 1:20:22 UTC

I finally reached my goal of 1mil credits within the past week and now they go up? :( Keeping credit low will stifle interest if you wish to keep the users low.

The project gives lower credit than credit new. I've never had to run a project over 50k hours to reach the 1mil goal. 3/4 of that 1mil was done on modern Zen CPUs. Just as a reference.
ID: 3580 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Michael Goetz
Avatar

Send message
Joined: 18 Feb 19
Posts: 22
Credit: 147,831
RAC: 861
Message 3585 - Posted: 9 Sep 2019, 20:57:05 UTC

Ok, I'm new here, but I'm more than a little familiar with how BOINC credit works (or doesn't).

Data point: I'm running on just the full cores (no hyperthreading) on a 1st gen Core i3.

It's actually a little faster, but let's call it a 2 GHz CPU. According to WUProp, I have about 48 core-hours done on this CPU. Let's call that 2 days.

On a modern CPU, it's not unreasonable for a typical application to get 1 FLOP per clock cycle. The definition of BOINC credit ("cobblestones") is:

"1/200 day of CPU time on a reference computer that does 1,000 MFLOPS based on the Whetstone benchmark"

(See https://boinc.berkeley.edu/wiki/Computation_credit)

A 1-GFLOP computer should earn 200 credits per day. This 2-GFLOP computer, running for two days, should have earned 4*200=800 credits. In fact, It earned. 546 credits. That's about 32% lower than the BOINC standard. If you recently increased the credits, then they were substantially below standard before the increase.

All of these tasks were run on September 8th and 9th. Even though the run times for the all the tasks were similar, most of the tasks were 38 credits and two were 79 credits. If all of the tasks were 79 credits, it would total 869, which is pretty close to the correct value. (800 is a nice round number, but the computer's actually a bit faster, so 869 is pretty close to correct.) If all the tasks were 38 credits, it would total 418 -- about half of what it should be.

For what it's worth, some projects (like PrimeGrid) give out more credit because their apps are highly optimized and compute more than 1 FLOP per clock cycle. With the right CPUs, some of our apps can execute up to 16 FLOPS per clock cycle by using AVX-512 and FMA3 instructions. So, of course, the credit per hour is correspondingly higher. But for a normal app, it's fair to assume one FLOP per clock cycle.

Setting credit fairly in a BOINC project is a tough, almost impossible, and definitely thankless task. Hopefully you'll find the data provided above to be helpful
Want to find one of the largest known primes? Try PrimeGrid. Or help cure disease at WCG.

ID: 3585 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Tomáš Brada
Project administrator
Volunteer developer
Avatar

Send message
Joined: 3 Feb 19
Posts: 505
Credit: 407,379
RAC: 525
Message 3587 - Posted: 9 Sep 2019, 21:34:38 UTC - in response to Message 3585.  

Dear Mr. Goetz. Thank you for the insight.
The procedure you describe is very close to how I benchmarked the application on my computer. Except that I ran much shorter workunits and used gflops estimate of my CPU-thread from a benchmark site. Moreover I confirmed the accuracy using job_times.php, but only at the beginning with only handful of computers.
Your computer is running windows, and the windows binary is slightly less effective for unknown reasons (same -O compiler level).
The application is not executing any floating point operation. So measuring the size of jobs in FLOP is awkward.
Thanks also for bringing the credit dichotomy into attention. It means I need to increase the "CF" credit coefficient.
Alternative to credit multipliers is credit_new, which is already received a load of hate here.
ID: 3587 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Michael Goetz
Avatar

Send message
Joined: 18 Feb 19
Posts: 22
Credit: 147,831
RAC: 861
Message 3588 - Posted: 9 Sep 2019, 22:21:14 UTC

As a follow up, I thought I'd post the following.

I didn't set the original credit at PrimeGrid, but Jim and I have been running PrimeGrid for about 7 years now and we've tried to keep the credit internally consistent. When AVX CPUs first came out, I realized the BOINC credit system couldn't be used at PrimeGrid anymore. Using the stock credit system, people buying new CPUs with AVX would get half the credit that people with older CPUs receive. That would attract people with slow computers to PrimeGrid and drive away people with the fastest computers, which is exactly the opposite of what we wanted! So we built a better system, where we could model the expected computation time of every task, and assign credit based on that model. It's not perfect, but it's pretty good -- and it's good enough to keep the users from complaining about it. It wasn't easy to do, and it's only feasible if your tasks are predictable. Not every project can do credit that way.

But I never compared the credit we give out to the "reference standard Cobblestone", like I did just now for your tasks. I thought that might be an interesting exercise. I have 9 computers running our new PPS-DIV application. I picked a representative task for each computer, and compared the actual credit (generated by our modeling system) against the the theoretical credit based on the Cobblestone formula. This is what I got for the actual vs. theoretical credit, based on the run times for the task and the characteristics (GHz and CPU family) for the 9 computers:

Anything above 100% means I got more credit than the Cobblestone formula suggests, under 100% means I got less than the formula predicts.

132%
116%
137% -- this is the computer I'm using here
125%
92%
106%
87%
102%
80%

Even at PrimeGrid, with a credit system as highly tuned to our actual applications as I know how to make it, the credit is still varying between 80% and 137% of the theoretical correct amount. It varies according to which computer is running the tasks. Even the last 5 computers on that list, which are all identical Xeon(R) Gold 6140 CPUs, vary from 80% to 106%.

It's unrealistic to expect any project to do better than that.

If all the tasks I ran here had received 79-ish credits, it would have come out to very close to the ideal 100%. At that credit rate, I think you're as close to perfect as you can get.

Yes, it will be less credit than at PrimeGrid, but that's because our apps can make use of the SIMD instructions to do multiple computations in parallel. That produces more credit -- and also more heat, uses more electricity, and requires better cooling solutions. For a "normal" app, I think at the higher credit rate you've got it set correctly. The results are always going to vary from computer to computer, so there's no "perfect" answer.
Want to find one of the largest known primes? Try PrimeGrid. Or help cure disease at WCG.

ID: 3588 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
mikey
Avatar

Send message
Joined: 28 Aug 19
Posts: 7
Credit: 578,871
RAC: 0
Message 3620 - Posted: 22 Sep 2019, 23:40:58 UTC

After the recent increase in credits and some time for the Project itself to tweak the app I have started crunching here again and am okay with the current amount of credits being awarded!!
I am now completing workunits in the 3+ hours range and getting @75 credits for each one of them, that is much better than the 9+ hours I was crunching them in and getting 57 credits for each one previously.
Is this the highest paying project no it's not, do I wish it was more..of course what cruncher wouldn't..but it doesn't need to be so I am okay with the current valuation and length of wokunits and will continue crunching here.

Thank you for your thought fullness in this matter!!
ID: 3620 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Natalia Makarova
Project scientist
Avatar

Send message
Joined: 8 Feb 19
Posts: 279
Credit: 0
RAC: 0
Message 3621 - Posted: 23 Sep 2019, 5:53:20 UTC

Thank you all for participating in the project!
The project gives excellent results.

Look here
https://boinc.tbrada.eu/forum_thread.php?id=3048
ID: 3621 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Previous · 1 · 2

Message boards : Number crunching : variation in credit allocation

©2020 Tomáš Brada