| Author |
Message |
|---|
John Honorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
                 
|
|
Welcome to the Dog Days of Summer Challenge
ATTENTION: The primality program LLR is CPU intensive and the WU's for this Challenge are quite long; so, it is vital to have a stable system (solid hardware) with GOOD COOLING. Invalid WU's will not receive credit.
PrimeGrid’s Challenge series continues with the Dog Days of Summer Challenge. Most dictionaries define the Dog Days as follows:
1 : the period between early July and early September when the hot sultry weather of summer usually occurs in the northern hemisphere
2 : a period of stagnation or inactivity
Please come join us in ushering an end to the hot sweltering days of summer. It's been 1 year 8 months since PrimeGrid's last Woodall prime 3752948*2^3752948−1. Let's break this "prime stagnation" with a new prime. :)
A 5 day Challenge is being offered on PrimeGrid's Woodall Prime Search (LLR) application. The duration is needed for the length of the WU's being tested. This will be another demanding Challenge similar to the PSP LLR Challenge.
Woodall LLR is currently testing n>6.6M which can produce just under a 2M+ digit prime. It will be like running a constant "stress test" on your computer for 5 days straight. Solid, well cleaned and cooled computers will be a necessity. Good hardware will "rule the day".
As for the psychology of the Challenge, patience will be the virtue needed if you wish to participate. :) The recent average WU is 47 hours long (range: 29hrs to 100hrs). So, you can see the importance of having stable hardware.
To participate in the Challenge, please select only the Woodall Prime Search (LLR) project in your PrimeGrid preferences section. The challenge will begin 18 August 2009 18:00 UTC and end 23 August 2009 18:00 UTC. Application builds are available for Linux 32 bit and Windows 32 bit. These applications will be sent to 64 bit clients. As with all LLR application projects, there is no advantage of 64 bit over 32 bit.
ATTENTION: The primality program LLR is CPU intensive; so, it is vital to have a stable system with good cooling. It does not tolerate "even the slightest of errors." Please see this post for more details on how you can "stress test" your computer. WU's will take ~32 hours on fast/newer computers and 75+ hours on slower/older computers. If your computer is highly overclocked, please consider "stress testing" it. Sieving is an excellent alternative for computers that are not able to LLR. :)
Please, please, please make sure your machines are up to the task.
Time zone converter:
The World Clock - Time Zone Converter
NOTE: The countdown clock on the front page uses the host computer time. Therefore, if your computer time is off, so will the countdown clock. For precise timing, use the UTC Time in the data section to the left of the countdown clock.
Scoring Information
Scores will be kept for individuals and teams. Only work units issued AFTER 18 August 2009 18:00 UTC and received BEFORE 23 August 2009 18:00 UTC will be considered for credit. We will use the "prime score" method which is based on k,b,n values (k*b^n-1) to score the challenge. The only difference is that the primary and double checker of a WU will receive the same score.
Therefore, each completed WU will earn a unique score based on its k,b,n values. The higher the n, the higher the score. This is different than BOINC cobblestones! A quorum of 2 is NOT needed to award Challenge score - i.e. no double checker. Therefore, each returned result will earn a Challenge score. Please note that if the result is eventually declared invalid, the score will be removed.
For details on how the score is calculated, please see this thread.
At the Conclusion of the Challenge
We kindly ask users "moving on" to ABORT their WU's instead of DETACHING, RESETTING, or PAUSING.
ABORTING WU's allows them to be recycled immediately; thus a much faster "clean up" to the end of an LLR Challenge. DETACHING, RESETTING, and PAUSING WU's causes them to remain in limbo until they EXPIRE. Therefore, we must wait until WU's expire to send them out to be completed.
Please consider either completing what's in the queue or ABORTING them. Thank you. :)
About Woodall Prime Search
Woodall Numbers (sometimes called Cullen numbers 'of the second kind') are positive integers of the form Wn = n*2^n-1, where n is also a positive integer greater than zero. Woodall numbers that are prime are called Woodall primes (or Cullen primes of the second kind).
The Woodall numbers Wn are primes for the following n:
2, 3, 6, 30, 75, 81, 115, 123, 249, 362, 384, 462, 512, 751, 822, 5312, 7755, 9531, 12379, 15822, 18885, 22971, 23005, 98726, 143018, 151023, 667071, 1195203, 1268979, 1467763, 2013992, 2367906, and 3752948 and composite for all other n less than 6688837.
It is conjectured that there are infinitely many such primes. Currently, PirmeGrid is testing for Woodall primes in the n=6.7M level (M=mega, 10^6). The last 3 Woodall primes found by PrimeGrid are:
2013992*2^2013992−1 (Lasse Mejling Andersen): official announcement | decimal representation | Prime Pages Entry
2367906*2^2367906−1 (Stephen Kohlman): official announcement | decimal representation | Prime Pages Entry
3752948*2^3752948−1 (Matthew J. Thompson): official announcement | decimal representation | Prime Pages Entry
For more information on Woodall numbers, please visit the following sites:
What is LLR?
The Lucas-Lehmer-Riesel (LLR) test is a primality test for numbers of the form N = k*2^n − 1, with 2^n > k. Also, LLR is a program developed by Jean Penné that can run the LLR-tests. It includes the Proth test to perform +1 tests and PRP to test non base 2 numbers. See also:
* Lucas-Lehmer-Riesel test (WIKI)
* Download LLR by Jean Penné
(Edouard Lucas: 1842-1891, Derrick H. Lehmer: 1905-1991, Hans Riesel: born 1929)
About the Dog Days of Summer
* What are the "dog days of summer"?
* Dog Days (WIKI)
____________
|
|
|
|
|
|
Hmpf
Wanted to try the woodall beforehand and experienced a strange behaviour: this client only gets one llrWOO-workunit albeit having two cores.
My cache-setting is: connect to network everey 1.000 days and maintain 0.001 days additional work.
I even bumped it up localy in the boinc-manager to 7.00/1.00 days without any effect. Resetting the project did not help either. I have boinc 6.4.5 on this client and this is a - of course - a 32-bit system.
Obtaining additional work from WCG works - even with the 1.000/0.001 setting.
What is afoot here i wonder?
____________
 |
|
|
|
|
Hmpf
Wanted to try the woodall beforehand and experienced a strange behaviour: this client only gets one llrWOO-workunit albeit having two cores.
Hi roadrunner,
Have you set BOINC to use 100% of the processors? |
|
|
|
|
|
I use the same profile as for my other client(s).
The Core 2 Duo behaved well, only the Core Duo is awkward in behaviour.
Anyways, what have i assembled today that may help in the next challenge?
Let us see what he is up to... ;)
____________
|
|
|
|
|
|
No Windows Machine, only Mac.
Sorry!
____________
 |
|
|
RytisVolunteer moderator
Project administrator
Send message
Joined: 22 Jun 05
Posts: 2653
ID: 1
Credit: 94,913,623
RAC: 139,244
                   
|
No Windows Machine, only Mac.
Sorry!
Actually, LLR Mac apps just went live. We are not sure if they'll work as intended, so you may test it :)
____________
|
|
|
|
|
|
From the front page:
Available: Woodall Prime Search (LLR) 107
Will this be increased before the challenge? I have about 50 cores that will be running the challenge...
____________
 |
|
|
|
|
From the front page:
Available: Woodall Prime Search (LLR) 107
Will this be increased before the challenge? I have about 50 cores that will be running the challenge...
Yes it will be increased.
Lennart |
|
|
|
|
From the front page:
Available: Woodall Prime Search (LLR) 107
Will this be increased before the challenge? I have about 50 cores that will be running the challenge...
Yes it will be increased.
Lennart
Thanks for the quick answer...hoping to do better this time and not have a host return all junk (going to take the OC offa that one...)
____________
|
|
|
LexsVolunteer developer
Send message
Joined: 16 Mar 08
Posts: 61
ID: 20289
Credit: 49,033,000
RAC: 0
             
|
|
Is there really a Woodall LLR app for Mac ?
Can't see it on the application list.
Would be great.
____________
 |
|
|
RytisVolunteer moderator
Project administrator
Send message
Joined: 22 Jun 05
Posts: 2653
ID: 1
Credit: 94,913,623
RAC: 139,244
|
|
It had problems and had to be pulled. We hope to have one before the challenge, but no guarantees.
____________
|
|
|
|
|
|
The challenge has begun!
I'm getting an error message with my new WUs, and I don't know what to do about it. One Woodall WU downloads and starts crunching. The second WU downloads soon after with no problems. Then, while the computer is starting to crunch the second WU, the first one ends up resetting the computation, with this message: "Task llrWOO_36282041_0 exited with zero status but no 'finished' file." But, it resetted the clock on its crunching. I aborted it, thinking that I may have a bogus finish, and don't want to waste some 80 hours on this thing for nothing. Now my computer downloads another WU to replace the aborted one, and the second WU does the same thing when my computer starts work on the third one. I'm afraid if I abort the second one, this may become just a never-ending loop of zero status, aborted WUs.
Any advice?
____________
 . |
|
|
 Sysadm@NbgVolunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 5 Feb 08
Posts: 1216
ID: 18646
Credit: 858,332,648
RAC: 171,091
                 
|
|
Here all possible cores crunching without any error, at the moment.
Known error is the "write to disk"-setting.
It had to be greater than zero!
But I don´t know, if it fit your problem ...
____________
Sysadm@Nbg
my current lucky number: 113856050^65536 + 1
PSA-PRPNet-Stats-URL: http://u-g-f.de/PRPNet/
 |
|
|
|
|
|
As in pspLLR my laptop with C2D T8300 is again the fastes client broken down by core.
extrapolated runtimes per WU and throughput (overall/per core):
Xeon E5504: ~291000 s/WU; 2.375 WU/d; 0.297 WU/d
C2Q Q9550: ~198500 s/WU; 1.741 WU/d; 0.435 WU/d
C2D T8300: ~153000 s/WU; 1.129 WU/d; 0.565 WU/d
C2D E6550: ~193350 s/WU; 0.891 WU/d; 0.446 WU/d
CD T2600: ~369100 s/WU; 0.468 WU/d; 0.234 WU/d
C2D 6400: ~183000 s/WU; 0.944 WU/d; 0.472 WU/d
Xeon E5405: ~216850 s/WU; 1.594 WU/d; 0.399 WU/d
The Nehalem without Hyperthreading would be as fast as the C2D T8300 i suppose.
If you take the core-frequency in account the C2Q Q9550 is even worse and well behind the Xeon E5405 - both having the same core - and on par with the Xeon E5504 but having only half the cores.
And no, none of my pc's is therm-tripping and all are running full speed.
With ppsLLR - which i run when there are no challenges - the ranking of the cpus is consistent.
____________
|
|
|
|
|
|
I stop my I7 because the CPU go up at 90°C with no OC and a good V8 coolermaster, here the temp is around 38° the 9650 and 9550 series all working well. |
|
|
|
|
|
most likely a very BAD idea to run stuff like that during the DOG DAYS - next time better go for the ice saints' days for a challenge on LLR's.. |
|
|
|
|
|
Tomorrow its going to be 35°C here in Germany. I will probably turn off my computer, because a room temperature above 40 would be bad for me and my computer. |
|
|
|
|
|
Quad Core 6600 64 Bit MS Home 3MB Ram running with 2245 million ops/sec after 24 Houers 40%.
Intel tz   
AMD Phenom 9650 Quad Core 64 Bit MS Home 4MB Ram running with 1885 million ops/sec after 24 Houers 50%.
Intel clearly the loser,
Why?Ram?
Uli
|
|
|
LexsVolunteer developer
Send message
Joined: 16 Mar 08
Posts: 61
ID: 20289
Credit: 49,033,000
RAC: 0
|
|
Yeah it will be hot tomorrow in germany.
I'll use this fan (see avatar), further I limited BOINC CPU usage to 90% to give the cooler a chance to catch up.
My Macs are set to only crunch at nighttime AP26.
The PS3 is already causing burn blister when you put your hand in the air exhaust, very scary.
(BTW: the new PS3 slim will not run linux, so no more crunching, except for Folding@Home natively.)
____________
|
|
|
|
|
Quad Core 6600 64 Bit MS Home 3MB Ram running with 2245 million ops/sec after 24 Houers 40%.
Intel tz
AMD Phenom 9650 Quad Core 64 Bit MS Home 4MB Ram running with 1885 million ops/sec after 24 Houers 50%.
Intel clearly the loser,
Why?Ram?
Uli
Answer is an badly optimized app, nothing to do with Intel or AMD.
Every additional core in the calculation makes worse the processing time.
Up to 50% performance drop with my V8 or 30% with the Q6600@3400 in comparison to a T9300@2500.
Intel C2D T9300@2500 Vista Home Premium 64 bit 26:50h 65.6% |
|
|
|
|
Answer is an badly optimized app, nothing to do with Intel or AMD.
Every additional core in the calculation makes worse the processing time.
Up to 50% performance drop with my V8 or 30% with the Q6600@3400 in comparison to a T9300@2500
optimized? maybe it could be tuned to stress memory not as much as it does now. curently it's just an ugly hog for most systems.
|
|
|
|
|
|
I did a test run of 4 woodalls last week on my OCed AMD PII 940 3.56ghz
I completed tasks at around 29 hours. this may be a slow moving challenge for sure |
|
|
|
|
I did a test run of 4 woodalls last week on my OCed AMD PII 940 3.56ghz
I completed tasks at around 29 hours. this may be a slow moving challenge for sure
K10 design has a big advantage on this app.. |
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
|
The first result has been returned. Congratulations!!!
Rank Name Team Score
1 Jeff17 BOINCstats 12420.65
____________
|
|
|
|
|
optimized? maybe it could be tuned to stress memory not as much as it does now. curently it's just an ugly hog for most systems.
This can't result from the memory.
My C2D E6550 is faster than the C2Q Q9550 which in turn is only a tad faster than the Xeon E5405.
The C2D E6550 has only 3GB single-channel DDR2-RAM whereas the C2Q Q9550 has 4GB dual-channel DDR2-RAM and the Xeon E5405 has 2GB dual-channel DDR2-FB-RAM with only 666 MHz.
The faster RAM as well as the faster clock-speed of the C2Q Q9550 clearly does not materialize in any way.
No the outcome from my point of view is very odd - as is with the pspLLR-WUs. |
|
|
|
|
Quad Core 6600 64 Bit MS Home 3MB Ram running with 2245 million ops/sec after 24 Houers 40%.
Intel tz
AMD Phenom 9650 Quad Core 64 Bit MS Home 4MB Ram running with 1885 million ops/sec after 24 Houers 50%.
Intel clearly the loser,
Why?Ram?
Uli
It might be. For comparison, my Q6600 @2.4GHz with 64bit Vista Home Premium and 4GB RAM hit 50% at about 22 hours with tasks on all four cores. By the way its measured fps is 2340.68 million ops/sec and my computer RAM is PC2-6400.
____________
~It only takes one bottle cap moving at 23,000 mph to ruin your whole day~
 |
|
|
pschoeferVolunteer developer
Volunteer tester
Send message
Joined: 20 Sep 05
Posts: 685
ID: 845
Credit: 2,886,413,454
RAC: 139,386
                           
|
|
My Q9450@3.2GHz (RAM: 2*2GiB+ 2*1GiB DDR2-800 CL4, OS: WinXP x64) finished it's first four workunits after less than 36 hours, that's about 8 hours or 25% longer than a single WU would take... at least it looks like it can do 12 WUs during the challenge, not only 10 or 11. :)
____________
  |
|
|
|
|
|
Got my first validation from the challenge here. Will the credits be capped at 1000 for the WUs?? (I was the one claiming 1140...)
120496084 109174 18 Aug 2009 18:03:40 UTC 20 Aug 2009 12:43:48 UTC Completed and validated 135,904.70 917.79 1,000.00
120496085 91120 18 Aug 2009 18:03:32 UTC 20 Aug 2009 11:01:50 UTC Completed and validated 144,443.77 1,140.49 1,000.00
____________
|
|
|
RytisVolunteer moderator
Project administrator
Send message
Joined: 22 Jun 05
Posts: 2653
ID: 1
Credit: 94,913,623
RAC: 139,244
|
|
Credit cap lifted.
____________
|
|
|
|
|
Credit cap lifted.
Retroactive??
____________
|
|
|
|
|
...
No the outcome from my point of view is very odd - as is with the pspLLR-WUs.
That's right - but we have to consider that not all WU running with the same "default runtime"
Example: same host (yorkfield@3,33 GHz), parallel running WU:
resultid=120506655 ~29h
resultid=120493827 ~38h
(both already validated)
____________
Member of Crunching Family
http://crunching-family.at/ |
|
|
|
|
|
Thin I blew my AC electricity lost 6 wu mislead, arghh |
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
|
Just over 48 hours remain!
Now's the time to start preparing for the end of the Challenge. For those cores that take longer than 48 hours to complete a WU, it's best to set "no new work" now. Make sure you have updated your clients to get those completed WU's reported and counted.
While this will be a small cleanup in terms of WU numbers, it will be very long in terms of time. If everyone can remember to either complete their cache or ABORT their WU's, then we'll be able to finalize earlier than later.
If you have the resources, please consider helping in the "clean-up" process after the end. However, we fully understand the need/desire to "move on" as this has been a VERY demanding Challenge.
Good Luck to everyone as we finish these last 48 hours...
At the Conclusion of the Challenge
We kindly ask users "moving on" to ABORT their WU's instead of DETACHING, RESETTING, or PAUSING.
ABORTING WU's allows them to be recycled immediately; thus a much faster "clean up" to the end of an LLR Challenge. DETACHING, RESETTING, and PAUSING WU's causes them to remain in limbo until they EXPIRE. Therefore, we must wait until WU's expire to send them out to be completed.
Please consider either COMPLETING what's in the queue or ABORTING them. Thank you. :)
____________
|
|
|
|
|
|
Seems as if even my Xeon E5405 can't make 8 WUs during the challenge, just one hour ago I suspended one WU to get the three make the deadline. And let us see if the E5504 could at least finish 3 more in addition to the 8 WUs he is completing. Would be a tough one on those two. My other clients all are able two finish 2 WUs per core except the too slow Core Duo T2600.
I consider to help clean up with the Xeon E5504 and the C2D 6400 if it would be any help. At home you could find my clients back at the ppsLLR-project as soon as the challenge is through. |
|
|
|
|
|
The first comparison
Intel Q 64 Bit 61 hours.
AMD Q 64 Bit 49 hours.
I do not believe this!!!!
Dont worry be happy.
Uli
|
|
|
|
|
|
I also notice LLR to perform better on AMD processors.
Opteron 1352 @ 2.1 GHz (linux 64bit): 48h
Intel Q9450 @ 2.66 GHz (winxp 32bit): 45h
both machines running 4 Woodalls at a time. The Opteron processor is faster considering the lower frequency (45h scaled down to 2.1GHz would be 57h, 48h is a speedup by 18%).
I don't know how the i7 performes here.
I hope, that I can finish another 10 WUs till the end of the challenge.
PS: heavy rain in Germany, temperature dropped to 20°C :) |
|
|
|
|
|
Mi record is 32 hours on a QX9650 W64 the i7 is far away behind
Mi old q6600 take 60 hours with only 1 Go ram
4 go ram 1066 is the best quickly results here. |
|
|
|
|
|
I7 920 No OverClocking
1 WU = 45 Hours (all 8 Cores had Work)
8 WU's = 59 Hours |
|
|
|
|
Intel Q 64 Bit 61 hours.
AMD Q 64 Bit 49 hours.
I do not believe this!!!!
my Q9450 @2.66 with PC8500 memory needs about 38 hours, the PHENOM II @3 GHz and PC6400 memory is just over 33. |
|
|
Scott BrownVolunteer moderator
Project administrator
Volunteer tester
Project scientist
Send message
Joined: 17 Oct 05
Posts: 2380
ID: 1178
Credit: 17,880,839,073
RAC: 8,995,761
                                               
|
|
E8400 Win XP Pro 32-bit - 33 to 33.5 hours running two at a time.
____________
141941*2^4299438-1 is prime!
 |
|
|
|
|
|
My runtimes in seconds, all with Red Hat Enterprise Linux 5 except for the C2D 6400, that one with Fedora 10.
The Xeon E5504 was run with 6 WUs at a time, with 8 the runtime would be in the region of ~297000s
Three additional WUs would come in around 1400 UTC on the Xeon E5405.
I could save my words saying that the C2Q Q9550 in comparison to the C2D E6550 is a huge disappointment.
I am now running only two WUs on the Quad Core in order to see if 2 WUs result in 130000s runtime as the T8300 would suggest.
C2Q Q9550
197255
196902
198126
197170
194443
194715
194478
194376
C2D T8300
152877
152777
155101
155481
C2D E6550
193578
194033
195889
196099
CD T2600
367571
366138
C2D 6400
182504
182771
177834
178690
Xeon E5504
257443
259130
259612
259543
263239
263677
241183
241795
Xeon E5405
215476
215937
216240
216133 |
|
|
|
|
|
2h 40' for 7.5%, that would be 1280s/1%.
That proves it for me, i now crunch away the remaining 8 WUs this client drew at the beginning of the challenge to help in the clean-up. |
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
|
The Dog Days are over! (well, at least the Challenge is.)
That concludes an extremely grueling Challenge. We hope everyone can now wipe the sweat from your brows and move on to more cooling endeavors. :)
The clean up process begins. If you have the resources, please consider helping out. However, we fully understand the need/desire to "move on" as this has been a VERY demanding Challenge.
Thank you to everyone who participated.
REMINDER: We kindly ask users "moving on" to ABORT their WU's instead of DETACHING, RESETTING, or PAUSING.
____________
|
|
|
|
|
|
Two of my AMD quad computers gave their lives to this challenge.......LOL
RIP
|
|
|
|
|
Two of my AMD quad computers gave their lives to this challenge.......LOL
RIP
power supply ?
Lennart |
|
|
|
|
Two of my AMD quad computers gave their lives to this challenge.......LOL
RIP
power supply ?
Lennart
Changed PSU on last machine that died....still doesn't work. I think maybe caps on the mobo.
Haven't even looked at the first one that died yet.
|
|
|
|
|
|
Care to offer Motherboard type & model #. About a year ago I lost an Asus M3A & a Phenom 9850 during a less demanding challenge. Hole burnt in the motherboard just to the west of the cpu. "Caps." I should really start the rebuilding of fallen computers. Sorry about your losses.
Mike |
|
|
|
|
Two of my AMD quad computers gave their lives to this challenge.......LOL
RIP
Now, that is dedication...
Sorry for your loss...
____________
|
|
|
LexsVolunteer developer
Send message
Joined: 16 Mar 08
Posts: 61
ID: 20289
Credit: 49,033,000
RAC: 0
|
|
It may sound funny, but my keyboard also kind of died during the challenge.
When I press some keys especially backspace, it will print a "f" sometimes.
Very annoying. (No, the f-key is not hanging.)
uck. ;-)
____________
|
|
|
|
|
Care to offer Motherboard type & model #. About a year ago I lost an Asus M3A & a Phenom 9850 during a less demanding challenge. Hole burnt in the motherboard just to the west of the cpu. "Caps." I should really start the rebuilding of fallen computers. Sorry about your losses.
Mike
ASUS M2AVM and AMD 9950.
9950 transplanted to another mobo and works as well as ever, which still means HOT.
The other puter that died is also ASUS M2AVM with either 9950 or 9850. Haven't checked it out yet.
I think these 125+ watt CPUs are toooo much for the M2AVM when running LLRs. |
|
|
|
|
|
My 2 main PC kept losing time due to the wonderful texas glitchy power grid. I lost power for a second or 2 3 times during the challenge. unfortunately these times occurred while I was at work.
I estimated that I could get 21 WU's ran on my rigs. I wound up with 11. maybe this challenge would have been better during the winter months |
|
|
|
|
The Dog Days are over! (well, at least the Challenge is.)
That concludes an extremely grueling Challenge. We hope everyone can now wipe the sweat from your brows and move on to more cooling endeavors. :)
The clean up process begins. If you have the resources, please consider helping out. However, we fully understand the need/desire to "move on" as this has been a VERY demanding Challenge.
Thank you to everyone who participated.
REMINDER: We kindly ask users "moving on" to ABORT their WU's instead of DETACHING, RESETTING, or PAUSING.
Is the project continuing to generate new Woodall LLRs or are we only processing the clean up WUs from the challenge.
|
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
Is the project continuing to generate new Woodall LLRs or are we only processing the clean up WUs from the challenge.
Yes, new WU's are being generated. However, as Challenge WU's are returned and enter the buffer, they are the first to be sent out. As long as there's "clean up" work, no new work goes out.
Clean up Progress
Currently 1079 WU's are outstanding from the Challenge. Of those, 535 are in the top 35 positions...affecting the outcome in almost all of those positions.
The timeline for the "clean up":
- 06 September 2009 18:00 UTC, the first round of "clean up" will be complete as last of the Challenge WU's will expire
- 20 September 2009 18:00 UTC, the second round of "clean up" will complete
- 02 October 2009 18:00 UTC, Harvest Festival Challenge
We hope to be finalized by the next Challenge. :)
Thanks to everyone helping out with the clean up.
____________
|
|
|
|
|
|
Is there an explanation of the clean up process posted?
If not, please explain the process.
Thank you. |
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
Is there an explanation of the clean up process posted?
If not, please explain the process.
Thank you.
"Clean up" is a rough term we use for the process of completing all the pending Challenge WU's...or at least enough of them so that the remaining has no effect on the outcome of the rankings.
For LLR Challenges, points are awarded with the completion of each task. However, two matching tasks are needed to complete the WU. Therefore, the actual outcome of the task, valid/invalid, is not known until the WU is complete.
If the task is valid, then the challenge points for that task remain. If the task is invalid, then the challenge points are deducted. So at the conclusion of an LLR Challenge, the point totals are the potential outcome. However, scores can decrease if tasks are later determined invalid.
Therefore, the goal of the "clean up" process is to complete enough of the pending tasks so that the remaining tasks do not alter the rankings.
For example, looking at individual rankings:
30. 418906.31
31. 410827.73
These two users will need all of their pendings complete since the difference in their score is less than the challenge points of 1 task. As long as that task remains outstanding, the Challenge will not be finalized.
Expiring tasks add significant time to this "clean up" process. That's why we always add the following request to each LLR Challenge: REMINDER: We kindly ask users "moving on" to ABORT their WU's instead of DETACHING, RESETTING, or PAUSING. Aborting allows tasks to be immediately recycled and sent out. For the rest, we have to wait until the tasks expire before they are sent out again.
For Woodall tasks of this size, the deadline is 14 days. Usually by the end of the first expiration period, most of the Challenge rankings are finalized. However, typically there's a few straggling tasks which don't get completed until the second expiration period.
Once the outcome of the rankings can not be altered by the remaining tasks, the Challenge is finalized.
____________
|
|
|
|
|
|
Thank you for the explanation.
|
|
|
|
|
(...)
Yes, new WU's are being generated. However, as Challenge WU's are returned and enter the buffer, they are the first to be sent out. As long as there's "clean up" work, no new work goes out.
(...)
Hm
Seems like there are currently no Challenge WU's.
Put two clients back on wooLLR.
Got one unsent (and therefore new), three new with an wingman only an hour old and three old from before the challenge with lots of "error while computing".
An "outstanding challenge wu"-counter would be nice so that one could judge when to put extra power to the clean-up - would that be hard to implement?
____________
|
|
|
pan2000Volunteer tester
Send message
Joined: 23 Jan 07
Posts: 26
ID: 5226
Credit: 328,291,757
RAC: 0
     
|
|
The same behaviour here. 9 of 10 WU´s in Progress are not Challenge relevant. |
|
|
VatoVolunteer tester
Send message
Joined: 2 Feb 08
Posts: 841
ID: 18447
Credit: 643,539,128
RAC: 559,599
               
|
|
Getting cleanup WUs is dependent on them expiring in the first place.
I always keep a chunk of my machines on the challenge project to cleanup, either until the challenge is finalised, or the next challenge starts.
I suspect it would be impossible to implement, but it'd be nice to have another tickbox in the subproject preferences (a bit like the "Send work from any subproject if selected projects have no work" tickbox) with semantics of "Send challenge cleanup work from any subproject if available", which the PG admins could control what subprojects appeared there (if any)
____________
 |
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
An "outstanding challenge wu"-counter would be nice so that one could judge when to put extra power to the clean-up - would that be hard to implement?
The buffer on the front page is set at 30. Anytime that number is over 30, there are returned WU's waiting to go out again. Typically, returned WU's get snatched up right away. However, there are times when there are plenty of returned WU's...right after the conclusion of the Challenge when users are aborting and again when the first round of expirations begin.
Right now there appears to be plenty of resources for the "clean up". The time to check back is when the first WU's for the Challenge start to expire...01 Sept 18:00 UTC. Between then and 06 Sept 18:00 UTC, WU's from the Challenge will be expiring for the first time.
Clean up Progress
Currently 639 WU's are outstanding from the Challenge. Of those, 305 are in the top 35 positions...affecting the outcome in almost all of those positions.
Thanks to everyone for helping out with the clean up.
History
27 Aug: 1079 WU's outstanding; 535 in the top 35 positions
29 Aug: 639 WU's outstanding; 305 in the top 35 positions
____________
|
|
|
pan2000Volunteer tester
Send message
Joined: 23 Jan 07
Posts: 26
ID: 5226
Credit: 328,291,757
RAC: 0
|
|
OK, 10 Cores are waiting for them... |
|
|
|
|
|
Any updates on the cleanup?
____________
|
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
|
Clean up Progress
Currently 444 WU's are outstanding from the Challenge. Of those, 218 are in the top 35 positions...affecting the outcome in 24 of those positions.
Thanks to everyone for helping out with the clean up.
History
27 Aug: 1079 WU's outstanding; 535 in the top 35 positions
29 Aug: 639 WU's outstanding; 305 in the top 35 positions
01 Sept: 444 WU's outstanding; 218 in the top 35 positions
____________
|
|
|
|
|
|
Again, 6 WUs, two from the challenge, four new with a wingman, one new with none.
42 WUs available says the frontpage. But at least two pieces... ^^
____________
|
|
|
|
|
|
Any update on the progress? |
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
Any update on the progress?
Clean up Progress
Currently 298 WU's are outstanding from the Challenge. Of those, 145 are in the top 35 positions...affecting the outcome in 18 of those positions.
Thanks to everyone for helping out with the clean up.
History
05 Sept: 298 WU's outstanding; 145 in the top 35 positions
01 Sept: 444 WU's outstanding; 218 in the top 35 positions
29 Aug: 639 WU's outstanding; 305 in the top 35 positions
27 Aug: 1079 WU's outstanding; 535 in the top 35 positions
____________
|
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
|
Clean up Progress
Currently 180 WU's are outstanding from the Challenge. Of those, 83 are in the top 35 positions...affecting the outcome in 13 of those positions.
Thanks to everyone for helping out with the clean up.
History
09 Sept: 180 WU's outstanding; 83 in the top 35 positions
05 Sept: 298 WU's outstanding; 145 in the top 35 positions
01 Sept: 444 WU's outstanding; 218 in the top 35 positions
29 Aug: 639 WU's outstanding; 305 in the top 35 positions
27 Aug: 1079 WU's outstanding; 535 in the top 35 positions
____________
|
|
|
|
|
|
Why results on AMD and Intel are different? If my wingman have AMD (like me), then WU is ok. But, if my wingman have Intel, then WU is "inconclusive". It's not fair. This same situation have two other members of my team. Please, check it!
Robert
http://www.primegrid.com/workunit.php?wuid=82458780
http://www.primegrid.com/workunit.php?wuid=79926158
|
|
|
pschoeferVolunteer developer
Volunteer tester
Send message
Joined: 20 Sep 05
Posts: 685
ID: 845
Credit: 2,886,413,454
RAC: 139,386
|
Why results on AMD and Intel are different? If my wingman have AMD (like me), then WU is ok. But, if my wingman have Intel, then WU is "inconclusive". It's not fair. This same situation have two other members of my team. Please, check it!
If both systems are stable, AMD and Intel are producing the same results (like 82459255, 83416614 or 81265605). If your workunits are marked as invalid, you should check whether your system is really stable.
____________
|
|
|
|
|
you should check whether your system is really stable. My AMD Q9650 really stable in AP26 and other project. Only Woodall's WU was invalid. Why? It's my fault or application is not stable? |
|
|
pschoeferVolunteer developer
Volunteer tester
Send message
Joined: 20 Sep 05
Posts: 685
ID: 845
Credit: 2,886,413,454
RAC: 139,386
|
you should check whether your system is really stable. My AMD Q9650 really stable in AP26 and other project. Only Woodall's WU was invalid. Why? It's my fault or application is not stable?
LLR is much more CPU intensive as most other applications. See the first post of this thread:
ATTENTION: The primality program LLR is CPU intensive; so, it is vital to have a stable system with good cooling. It does not tolerate "even the slightest of errors." Please see this post for more details on how you can "stress test" your computer. WU's will take ~32 hours on fast/newer computers and 75+ hours on slower/older computers. If your computer is highly overclocked, please consider "stress testing" it. Sieving is an excellent alternative for computers that are not able to LLR. :)
Please, please, please make sure your machines are up to the task.
____________
|
|
|
|
|
|
I know, i know... It's clear!
Well, I will run this host on other LLR applications and we will back to this discussion. |
|
|
|
|
|
Back in the days there was a similar problem at einstein@home, but there mixed OSes yielded different results: 2 Windows => valid, 1 Windows + 1 Linux => inconclusive + resent 1 Windows => both Windowses valid Linux invalid. Since there were more Windows clients than Linux this was a showstopper.
No idea what the reason or the solution was. |
|
|
RytisVolunteer moderator
Project administrator
Send message
Joined: 22 Jun 05
Posts: 2653
ID: 1
Credit: 94,913,623
RAC: 139,244
|
|
It's not at all the same problem.
____________
|
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
|
Clean up Progress
Currently 141 WU's are outstanding from the Challenge. Of those, 64 are in the top 35 positions...affecting the outcome in 10 of those positions.
Thanks to everyone for helping out with the clean up.
History
11 Sept: 141 WU's outstanding; 64 in the top 35 positions
09 Sept: 180 WU's outstanding; 83 in the top 35 positions
05 Sept: 298 WU's outstanding; 145 in the top 35 positions
01 Sept: 444 WU's outstanding; 218 in the top 35 positions
29 Aug: 639 WU's outstanding; 305 in the top 35 positions
27 Aug: 1079 WU's outstanding; 535 in the top 35 positions
____________
|
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
|
Clean up Progress
Currently 79 WU's are outstanding from the Challenge. Of those, 41 are in the top 35 positions...affecting the outcome in 4 of those positions.
Thanks to everyone for helping out with the clean up.
History
16 Sept: 79 WU's outstanding; 41 in the top 35 positions
11 Sept: 141 WU's outstanding; 64 in the top 35 positions
09 Sept: 180 WU's outstanding; 83 in the top 35 positions
05 Sept: 298 WU's outstanding; 145 in the top 35 positions
01 Sept: 444 WU's outstanding; 218 in the top 35 positions
29 Aug: 639 WU's outstanding; 305 in the top 35 positions
27 Aug: 1079 WU's outstanding; 535 in the top 35 positions
____________
|
|
|
|
|
|
I'm back...
My host AMD Q9650 is really stable. From 10.09 to now this host worked 24/7 on LLR. All WU is valid (some still pending).
http://www.primegrid.com/results.php?hostid=118279&offset=0&show_names=1&state=0&appid=0
But, in last race... all of Woodall's WU was invalid (inconclusive).
So, I think... Woodall's application have some mistakes for AMD Phenom processors. |
|
|
Sysadm@NbgVolunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 5 Feb 08
Posts: 1216
ID: 18646
Credit: 858,332,648
RAC: 171,091
|
So, I think... Woodall's application have some mistakes for AMD Phenom processors.
I do not think so ...
My AMD Phenom does not report any mistake!
klick
____________
Sysadm@Nbg
my current lucky number: 113856050^65536 + 1
PSA-PRPNet-Stats-URL: http://u-g-f.de/PRPNet/
|
|
|
|
|
|
Do not tell jokes!
You are working on Linux, but i'm working on Win64. It is not the same application. |
|
|
|
|
|
Look here, i think you ran into the same problem.
Try pspLLR and report if there are occassionally errors too.
I know of my Xeon-Host - which hast FB-DIMM with integrated temperature sensors - that he is constantly thermal tripping in both of his RAM-Modules. WoodallLLR and pspLLR are very RAM-intensive. I have ~85-90°C FB-DIMM temperature and only ~60-63°C CPU-Core temperature when running on of these WUs.
____________
|
|
|
Sysadm@NbgVolunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 5 Feb 08
Posts: 1216
ID: 18646
Credit: 858,332,648
RAC: 171,091
|
Do not tell jokes!
Am I joking ???
You are working on Linux, but i'm working on Win64. It is not the same application.
Then i woud like to invite you to use a stable operating system like me - and not to use the (whatever it is) from Billy_Redmond
Then you will not have erroring apps ...
PS: irony off, perhaps you do not realized it!!!
____________
Sysadm@Nbg
my current lucky number: 113856050^65536 + 1
PSA-PRPNet-Stats-URL: http://u-g-f.de/PRPNet/
|
|
|
|
|
|
OK, sorry. Now irony is off. ;)
I don't want Woodall's credit, I want resolve this problem.
I'm studying roadrunner_gs post... |
|
|
|
|
Look here, i think you ran into the same problem.
Try pspLLR and report if there are occassionally errors too.
I know of my Xeon-Host - which hast FB-DIMM with integrated temperature sensors - that he is constantly thermal tripping in both of his RAM-Modules. WoodallLLR and pspLLR are very RAM-intensive. I have ~85-90°C FB-DIMM temperature and only ~60-63°C CPU-Core temperature when running on of these WUs. Thanks for important advice. Well...
Host AMD Q9650, no O/C.
http://www.primegrid.com/results.php?hostid=118279&offset=0&show_names=1&state=0&appid=3
In August, Woodall LLR...
3x error = Reason: Out Of Memory (C++ Exception) (0xe06d7363) at address ...
2x inconclusive and invalid
1x still pending, but wingmans go to errors
Now, in September, miscellaneous LLR, 7 day x24h, all right...
Max temp. (afternoon, Speedfan) core 42C, cpu 56C. In the night, 40C/50C.
NO ERRORS!
valid= 2x 321, 4x Cullen, 1x PSP, 7x SGS
still pending = 4x
PSP LLR? OK, I'm going to testing this. My quad standard: 3/4 CPU for PSP and 1/4 CPU for 2x GPU R4770. How long? |
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
|
Clean up Progress
Currently 43 WU's are outstanding from the Challenge. Of those, 21 are in the top 35 positions...affecting the outcome in 3 of those positions.
Thanks to everyone for helping out with the clean up.
History
20 Sept: 43 WU's outstanding; 21 in the top 35 positions
16 Sept: 79 WU's outstanding; 41 in the top 35 positions
11 Sept: 141 WU's outstanding; 64 in the top 35 positions
09 Sept: 180 WU's outstanding; 83 in the top 35 positions
05 Sept: 298 WU's outstanding; 145 in the top 35 positions
01 Sept: 444 WU's outstanding; 218 in the top 35 positions
29 Aug: 639 WU's outstanding; 305 in the top 35 positions
27 Aug: 1079 WU's outstanding; 535 in the top 35 positions
____________
|
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
|
Clean up Progress
Currently 25 WU's are outstanding from the Challenge. Of those, 11 are in the top 35 positions...affecting the outcome in 3 of those positions.
Thanks to everyone for helping out with the clean up.
History
26 Sept: 25 WU's outstanding; 11 in the top 35 positions
20 Sept: 43 WU's outstanding; 21 in the top 35 positions
16 Sept: 79 WU's outstanding; 41 in the top 35 positions
11 Sept: 141 WU's outstanding; 64 in the top 35 positions
09 Sept: 180 WU's outstanding; 83 in the top 35 positions
05 Sept: 298 WU's outstanding; 145 in the top 35 positions
01 Sept: 444 WU's outstanding; 218 in the top 35 positions
29 Aug: 639 WU's outstanding; 305 in the top 35 positions
27 Aug: 1079 WU's outstanding; 535 in the top 35 positions
____________
|
|
|
|
|
|
I'm back... again...
http://www.primegrid.com/results.php?hostid=118279&offset=0&show_names=0&state=0&appid=8
6x LLR PSP = 2x OK, 4x still pending.
6x LLR Woodall = 6x 0 credits.
So, my host AMD Q9650, no O/C, is OK. But Woodall application for AMD Phenom (Win64) is not OK. You can say whatever you want! I will boycott this application. And I recommend a boycott of this application to the members of my team. Of course, only in the case of AMD Phenom processors.
I prefer to count the other subprojects (PG). I don't want to wait two months for validation! This is a neverending story or waiting for Godot. |
|
|
|
|
I'm back... again...
http://www.primegrid.com/results.php?hostid=118279&offset=0&show_names=0&state=0&appid=8
6x LLR PSP = 2x OK, 4x still pending.
6x LLR Woodall = 6x 0 credits.
So, my host AMD Q9650, no O/C, is OK. But Woodall application for AMD Phenom (Win64) is not OK. You can say whatever you want! I will boycott this application. And I recommend a boycott of this application to the members of my team. Of course, only in the case of AMD Phenom processors.
I prefer to count the other subprojects (PG). I don't want to wait two months for validation! This is a neverending story or waiting for Godot.
You most likely need to cool your CPU better. You can try setting BOINC to use 90% of your CPU and try running some woodall WUs again. (Your CPU won't get as hot.) There is nothing wrong with the woodall application, I ran a few WUs during the challenge on my AMD laptop. The laptop ran at 90C+ and all the workunits were valid. You can download a program to monitor your CPU temperature: http://www.almico.com/speedfan.php.
____________
 |
|
|
|
|
|
I am somewhat speechless here, let us get to the bottom-line: If you are running into this problem on woodallLLR does mean you have a problem with woodallLLR. Not having run into the same problem while using pspLLR up to now only means you haven't run into this problem while using pspLLR yet. Statistically spoken this does not prove anything. You must must crunch a lot more to get a significant empirical base. Furthermore i do not find a lot more persons using woodallLLR complaining about failing WUs so this could be an individual fate taken in account the mass of crunched WUs so far. In conclusion i could only ask whether you have taken the chance to put in some other RAM or CPU in your computer to single out those points? You could also fetch prime95 program and run the CPU- or RAM-Test for hours on end - if you get an error there in either of the tests this is a hint to problems in your computer - but if not this is no significant prove to not having problems in your computer.
____________
|
|
|
|
|
|
It's not necessary the application that's wrong, but the combination PSP LLR / Woodall LLR running on an AMD Phenom is suspicious.
I have a Phenom 9350e / stock speed 2.0 GHz.
In the PrimeGrid's Birthday/Summer Solstice Challenge running PSP LLR's I had the same troubles (running @ 2.11GHz).
However I had no heattroubles at all (50C), I had 8 invalids out of 14.
After the challenge I helped cleaning up and to test without OC and had 4 valids out of 4 (not really statistical proved, but anyway).
A lot of crunchers running on Intel with massive OC don't had the invalids with much higher temperatures.
This challenge running the Woodall's without OC I had 7 valids out of 7.
____________
 |
|
|
|
|
You most likely need to cool your CPU better
Better? I have Scythe Ninja. Look at: http://www.primegrid.com/forum_thread.php?id=1440&nowrap=true#18090
"Max temp. (afternoon, Speedfan) core 42C, cpu 56C. In the night, 40C/50C".
The laptop ran at 90C+ and all the workunits were valid On my laptop too at the same temp. |
|
|
|
|
I am somewhat speechless here, let us get to the bottom-line: If you are running into this problem on woodallLLR does mean you have a problem with woodallLLR. Not having run into the same problem while using pspLLR up to now only means you haven't run into this problem while using pspLLR yet. Statistically spoken this does not prove anything
You're right! Statistically of course.
But... expired over a month and still not able to verify many of the tasks from the race. In that two of my. Perhaps the problem is not such an individual. Maybe.
I think that the problem is in cooperation... Woodall application <--> Boinc library <--> SysOp Win7/64 (7.1000) <--> AMD Phenom processors. Perhaps the problem will solve itself when you upgrade one of the elements. The application has already been upgraded (2009.09.23).
Now, I finish with Woodall and concentrate on other projects PG.
Thanks for the help! |
|
|
|
|
|
It really doesn't make sense to me that we are about to start another challenge and have not yet finalized the last one.
Will we end up with two challenges unresolved at the same time?
|
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
It really doesn't make sense to me that we are about to start another challenge and have not yet finalized the last one.
Will we end up with two challenges unresolved at the same time?
Actually...yesterday the last 3 WU's were run manually to compare residues with the pending. 1 of the pending was determined invalid. We have to manually adjust the rankings. As soon as that's done, the Challenge will "officially" be final.
____________
|
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
|
The results are Final!
With 4 hours to spare before the next Challenge begins, the results are official. (whew) What an incredible amount of work completed! In the 5 day Challenge, over 4200 WU's (8400 tasks) were returned resulting in over 52M challenge points awarded.
Again, we were unable to find a "significant" prime during a Challenge. However, we are certainly much closer to finding the next one...just that it would have been nice to find it during the Challenge. :)
60 teams and 315 individuals participated in the Challenge.
The Challenge Points stats will be updated soon. Congratulations to SETI.Germany as the top team and j2satx as the top individual. Remember, the top 35 places earn points for this Challenge.
Thank you again to everyone. We appreciate your participation and hope you had fun!
Stats: Participants | Teams
Top 10 Teams
1 SETI.Germany 10859709.70
2 BOINCstats 7260535.12
3 US Navy 4311969.97
4 Team 2ch 3610340.83
5 PrimeSearchTeam 3379341.50
6 The Knights Who Say Ni! 3201221.58
7 L'Alliance Francophone 2100419.63
8 BOINC@Poland 1634287.34
9 Crunching Family 1613458.34
10 Czech National Team 1419878.20
Top 10 Individuals
1 j2satx US Navy 2368968.11
2 UL1 SETI.Germany 1933988.47
3 andmore SETI.Germany 1876126.88
4 lennart SM5YMT PrimeSearchTeam 1860967.30
5 fe Chalmers University of Technology 1343151.78
6 Lumiukko PrimeSearchTeam 1285538.96
7 Dogg US Navy 1238129.02
8 John Galt 007 The Knights Who Say Ni! 1125620.81
9 yamaga@news Team 2ch 1099850.03
10 Siegfried Niklas Crunching Family 1029930.81
____________
|
|
|
|
|
|
Yeah! It's great, but... two of my racing WU is still awaiting for validation.
http://www.primegrid.com/workunit.php?wuid=82458807
http://www.primegrid.com/workunit.php?wuid=82459771
Thus, the final doesn't mean the end of the struggle with that race. |
|
|
