• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

A brief message about YoYo@Home

Rudy Toody

Rudy Toody

Diamond Member
This is a thread about YoYo@Home, but there is formula I use to determine the projects that play nice with my boxes.

First, I take back everything I said about caferace watching his back. I am watching his back as he fades into the distance!

Second, We have 3 TeAm Members who are Walking the Dog, Rocking the Baby and going Around the World and the other are tangling their strings! I recommend you check out YoYo@Home for the TAS BOINC Month, because it is very liberal with the credits. See how many tricks you can do.

I used what I call GFR to determine how each project works with my boxes.

GFR (Granted For Requested) = Granted_Credits * 100 / Requested_Credits.

If the number is greater than 100%, then you are doing well. If it is less than 100%, find a project where that number goes up for that box.

On YoYo@Home: for the Phenom 9850BE the GFR is around 155% and for the FX-60s the GFR is around 136%.

On 3x+1: for the FX-60s the GFR is around 122%.

So, you can see it pays for me to move everything to YoYo@Home that is crunching 3x+1. I don't have stats for other projects.

If you decide to check this formula, be sure to use the same box to do so, since Requested_Credits is dependent on the processor, math ability, and memory, etc....

--Fred
 
Thanks RT, but you're giving me a run for my $$ today. I woke up this morning, looked at the sTAtS and went :shocked:. 😉

Yoyo rocks. That being said I'm avoiding running the DPAD project on it, as it works out we compete against the TeAm that way. OGR is a hoot, and Evolution seems to have stabilised. After I get my surgery done tomorrow, I'll get back on a few trhings I'm behind on. The TAS belt, Peters new toy to ship off and maybe some general neffage. 😀

-jim
 
Originally posted by: caferace
Thanks RT, but you're giving me a run for my $$ today. I woke up this morning, looked at the sTAtS and went :shocked:. 😉

Yoyo rocks. That being said I'm avoiding running the DPAD project on it, as it works out we compete against the TeAm that way. OGR is a hoot, and Evolution seems to have stabilised. After I get my surgery done tomorrow, I'll get back on a few trhings I'm behind on. The TAS belt, Peters new toy to ship off and maybe some general neffage. 😀

-jim
Click to expand...

I wish you well on your surgery! In fact, I hope you come out of sedation with a nurse on each arm.

--Fred
 
Originally posted by: Rudy Toody
Originally posted by: caferace
Thanks RT, but you're giving me a run for my $$ today. I woke up this morning, looked at the sTAtS and went :shocked:. 😉

Yoyo rocks. That being said I'm avoiding running the DPAD project on it, as it works out we compete against the TeAm that way. OGR is a hoot, and Evolution seems to have stabilised. After I get my surgery done tomorrow, I'll get back on a few trhings I'm behind on. The TAS belt, Peters new toy to ship off and maybe some general neffage. 😀

-jim
Click to expand...

I wish you well on your surgery! In fact, I hope you come out of sedation with a nurse on each arm.

--Fred
Click to expand...

Ditto! .. and a beer close to your mouth ... 🙂
 
The Yoyo client actually gives me much lower points per hour/day than some other projects on my "Frankenstein" machine at home that is attached to every possible project (but only actually runs a few of them). Using your formula, my Sempron 64 (2.2Ghz) gets a score of about 75% from Yoyo. Of course, a few of the projects I do run aren't any better than that, but I share some CPU time with them because I like the projects and would do it regardless of the points.


For example, I know that the ABC project awards an almost obscenely high number of points compared to most other projects, but the project itself just doesn't interest me so I would only run it if I was in a race for the most possible points regardless of the usefulness (or lack thereof) of the project.


However, since I am a stats junkie at heart, I'm going to have to let "Frankie" do some work for other projects now and then just to see what kind of scores I get from them with your formula. 🙂
 
I have checked a few projects using RudyToody's formula:

100.00 ______ Seti@home
100.00 ______ CPDN
114.23 ______ MalariaControl.net
137,04 ______ POEM@Home
100.00 ______ Einstein@home

This shows clearly, that the project which really use FLOPs for a measure give the credits due, while the credit for some other projects are dependent on the other cruncher's computers (MalariaControl) or POEM@Home, where you get a set number of credits for each WU (9.45 or 15.09 or 49.35 or 69.05). If a fast computers delivers the result faster (i.e. crunches shorter time and thus claims less credits) then you get a better efficiency, and, conversely, a slow computer crunches longer then expected, claims more credits, gets the fixed amount and thus a worse efficiency.

YoYo@home did not play nice with the computer which I test the projects on: one WU, which completed OK, claimed 10.27 credits, got none, one other, which crunched for 82,350.46 seconds, claimed 189.30 credits but got none - because of a "client error". I'll try YoYo at one more comp before I decide whether to continue or not. I really like the idea of a wrapper for DPAD and for evolution@home.

(posted also at an other forum) 😉
 
Originally posted by: petrusbroder
YoYo@home did not play nice with the computer which I test the projects on: one WU, which completed OK, claimed 10.27 credits, got none, one other, which crunched for 82,350.46 seconds, claimed 189.30 credits but got none - because of a "client error". I'll try YoYo at one more comp before I decide whether to continue or not. I really like the idea of a wrapper for DPAD and for evolution@home.
Click to expand...

Just an FYI on the Evolution WU's which is Petrus errored out on above. If an Evo WU shows a progress of greater that 100% do not abort it. I made that mistake at the beginnings of my Yoyo career 🙂 and discovered that if one is patient they will finish soon and reward handsomely.

Yes, it's confusing.

-jim

(home, recuperating nicely with zero complications. No nurses involved except my live-in (dba: Ms. Caferace))

 
I've got all my machines on Yoyo but seeing so many bad WUs really worries me. Some of the machines I'm not able to check consistently so it'd be a pain if they got a bad WU and never left it 🙁

 
caferace said:
Yoyo rocks. That being said I'm avoiding running the DPAD project on it, as it works out we compete against the TeAm that way.
Click to expand...
hate to bump an old thread, but i was wondering if the bolded statement still holds true. that is, if i install the yoyo@home project and run Muon (DPAD) through BOINC, is it still impossible to earn credit for the TeAm? must i install DPAD and its standalone (non-BOINC) client in order to earn credit for the TeAm? or can points be earned for the TeAm whiile crunching Muon (DPAD) on BOINC (via yoyo@home) these days?

TIA,
Eric
 
We are the number one team on the original DPAD non-boinc project.

Yoyo puts a wrapper around the non-boinc app and gives us boinc credits for crunching it. However, the DPAD credits on the non-boinc project go to Yoyo. Therefore, we threaten our number one position when we crunch the boinc version.
 
thanks Rudy...so i'll have to run the standalone DPAD client if i want to earn points for the TeAm. is it a PITA to manage more than one type of client (i.e. both a BOINC client and a non-BOINC client)? also, will BOINC statistics websites (BOINCstats, BOINC Combined Statistics, BOINC all Project Stats, etc.) log the work/progress statistics of the non-BOINC DPAD client? i would imagine they would, considering that they monitor team stats in addition to general user stats...but i'm hoping someone could confirm this.

TIA,
Eric
 
to the top again...

...i originally bumped this old thread to get some Muon1 DPAD questions answered. but now i actually have questions pertaining to the OP.

Fred, your GFR formula makes sense to me. but where do we find the values that need to be plugged into the formula? i see that i can find "claimed credit" and "granted credit" values on a per task basis for Einstein@Home for instance. but how many tasks must i sample in order to get an accurate GFR? for example, just about every Global Correlations task and every Gravitational Wave task claims ~75 credits and grants ~251 credits. but (251*100)/75 = 335% seems like a ridiculously high average project efficiency. what gives?

also, projects like SETI@Home and Milkyway@Home don't seem to have a "credit claimed" column, so how am i supposed to calculate GFR for these projects? is there another place where i can find "claimed credit" for these projects?

TIA,
Eric
 
Sunny129 said:
to the top again...

...i originally bumped this old thread to get some Muon1 DPAD questions answered. but now i actually have questions pertaining to the OP.

Fred, your GFR formula makes sense to me. but where do we find the values that need to be plugged into the formula? i see that i can find "claimed credit" and "granted credit" values on a per task basis for Einstein@Home for instance. but how many tasks must i sample in order to get an accurate GFR? for example, just about every Global Correlations task and every Gravitational Wave task claims ~75 credits and grants ~251 credits. but (251*100)/75 = 335% seems like a ridiculously high average project efficiency. what gives?

also, projects like SETI@Home and Milkyway@Home don't seem to have a "credit claimed" column, so how am i supposed to calculate GFR for these projects? is there another place where i can find "claimed credit" for these projects?

TIA,
Eric
Click to expand...
This only works if there are values for the claimed credit. I you receive more than claimed it is GOOD! If you receive less than claimed it is BAD!

If the project doesn't show claimed, then you have to decide using some other method.

I came up with this because I was tired of claiming 40 credits and receiving 25. I decided that I would trust their estimate of the credits for each WU and that is what I expect to receive. If I receive more, then that is the icing on the cinnamon roll.

These numbers could be different for every computer so, if your computer is less than 100, go somewhere else.
 
Last edited:
Rudy Toody said:
This only works if there are values for the claimed credit. I you receive more than claimed it is GOOD! If you receive less than claimed it is BAD!

If the project doesn't show claimed, then you have to decide using some other method.

I came up with this because I was tired of claiming 40 credits and receiving 25. I decided that I would trust their estimate of the credits for each WU and that is what I expect to receive. If I receive more, then that is the icing on the cinnamon roll.

These numbers could be different for every computer so, if your computer is less than 100, go somewhere else.
Click to expand...
well, it is what it is i guess. Einstein@Home is currently the only project for which i can calculate GFR's:

Einstein@Home S5GC1HF (Global Correlations search) = 335%
Einstein@Home S6Bucket (Gravitational Wave search) = 335%
Einstein@Home BRP3 (Binary Radio Pulsar search) = 152%

i'd like to know how Petrusbroder was able to find claimed/requested credits for SETI@Home, as i also crunch S@H and would like to calculate my GFR for the project. i'd like to do the same for Milkyway@Home as well, but that's another project for which i cannot find claimed/requested credit values. i could have sworn that i used to see them in the past, but i don't remember where...perhaps it was on the "pending credit" pages of each project's website - the only problem is that i can't verify this at the moment b/c S@H is down and all of my MW@H tasks have been getting verified instantly lately, so they never show up with a claimed credit value under "pending."

*EDIT* - i just realized that S@H's data driven web page servers are still up and running, despite work production still being at a stand still. so in a quest to find out why nothing was showing up on my S@H pending credit page, i discovered a thread on the S@H forums discussing the gradual disappearance of everyone's "pending credit" list. long story short, a volunteer developer/tester stated that "claimed/requested credit" is no longer being used in the calculation of actual/granted credit, and so it is now a waste of the scheduler's time and resources to calculate claimed/requested credit values. unfortunately for those of us looking to calculate our own GFR for the entire S@H project and/or any of its individual applications, it is no longer possible since claimed/requested credits will no longer be calculated by the scheduler and displayed in the "pending credit" list. note that i'm not claiming that everyone's pending credit is all of the sudden getting validated in an attempt to clean the database...rather you can still see your pending credit by going to the tasks page and filtering the pending tasks - its just that a claimed credit value isn't displayed on this page (or any other page on the S@H server from here on out i guess). it also just occurred to me that Petrusbroder was able to calculate his GFR for S@H b/c he originally did so back in 2008 - before they changed the formula for "granted credit" to no longer depend on a "claimed/requested credit" calculation.

*EDIT #2* - from what i understand, Milkyway@Home's "pending credit" page is going the way of the dinosaur, just like it is for S@H (for the same reason). so in addition to S@H, i may not be able to calculate my GFR for MW@H anymore either...:'(
 
Last edited:
You must log in or register to reply here.

TRENDING THREADS

Back
Top