John Forum moderator Project scientist Send message Joined: Feb 21 06 Posts: 1560 ID: 2449 Credit: 357,929 RAC: 0
Welcome to PrimeGrid's Birthday/Summer Solstice Challenge
While the Challenge is still quite some time away, we wanted to prepare everyone for the intensity of doing LLR testing on such large numbers. PSP LLR is currently testing n>7M which can produce a 2M+ digit prime.
This Challenge will definitely live up to its definition...it will truly be challenging. It will be like running a constant "stress test" on your computer for 9 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 average WU is 50 hours long (range: 28hrs to 100hrs). So, you can see the importance of having stable hardware.
------------------------------------
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 ~30-35 hours on fast/newer computers and 50-90 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 return valid LLR results.
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.
____________
robin Send message Joined: May 2 09 Posts: 64 ID: 39483 Credit: 1,856 RAC: 0
I have just finished a 3*2^n+1 Proth LLR test on my single (only one computer)
GenuineIntel
Intel(R) Pentium(R) D CPU 3.00GHz [x86 Family 15 Model 6 Stepping 5]
Number of processors 2
Operating System Microsoft Windows Vista
Home Basic x86 Editon, Service Pack 1, (06.00.6001.00)
It took 54,5 hours and i do not even know which n it was and thus the number of digits of the number.No information at all was given to me.
I kept the computer running 100% on both processors 24h/day even when i was personnally using the computer.
The problems i have:
The air-fan cooling system seems to be good but is halas too noisy when the two processors are running 100% with LLR or AP26 tests .After 2 hours or more you get stressed with the noise. I do not think, though, that processors or mother board are going to die.
I do not know if the processors are 64 or 32 bits and if vista Home Basic allows 64 bits and what factor of time i would gain if i could pass from 32 to 64 bits on LLR or AP26 tests.
I do not know if there is a way to make work a long 321-LLR test on both processors to gain time-
Most of the Proth LLR tests i get are only 20 minutes long.Compared with the 54 hours of the 321 test of a number near 2 milliion digits (i imagine); indicates that these proth numbers might be very small, far from 2 million digits.
If a number is k times longer (in number of digits) what is the LLR time function of k?
Thank you for your answers if any.
Sorry for my bad English.
Core Duo T2600 => 9h 50' for 10 % => projected 98 hours 90/85°C but got only one WU, no idea why
Core 2 Duo T8300 => 8h 12' for 17 % => projected 48 hours 96/89°C but had to downclock from 2.4 GHz to 2 GHz.
I think i have to skip this challenge since in June the ambient temperature would be 3 to 4 °C higher.
DoES Volunteer tester Send message Joined: Oct 11 08 Posts: 289 ID: 30382 Credit: 1,883,604 RAC: 0
Uarg!
Core Duo T2600 => 9h 50' for 10 % => projected 98 hours 90/85°C but got only one WU, no idea why
Core 2 Duo T8300 => 8h 12' for 17 % => projected 48 hours 96/89°C but had to downclock from 2.4 GHz to 2 GHz.
I think i have to skip this challenge since in June the ambient temperature would be 3 to 4 °C higher.
Hi Roadrunner --- I will be very pious here --- " It servest thou self right for over clocking" (joke) --- I'm rather neutral to the new look page you commented on -- I do like the filters on the task page though -- filtered for errors and found 7 for the last 3 days (3 I knew about--hardware failures) --3 from the same host (another celeron dying??) and 1 unknown???-- suppose not bad for 17 boxes
The ambient temp in my computer room is 37 deg centigrade--- if I see smoke from a heatsink -- I just pour a few drops of beer from my can on the CPU fan (works for me)
@robin: "I have just finished a 3*2^n+1 Proth LLR test"
there are 321 LLR and Proth LLR, the one you finished is a 321 LLR (see here http://www.primegrid.com/result.php?resultid=104606509). The number needs do be double checked first. When this is done, there should be some notice on the Workunit-Info (http://www.primegrid.com/workunit.php?wuid=70760011) telling you something like "3*2^123456+1 is not prime" (at least it used to be like that).
The Proth LLR numbers are in fact much smaller. In the Proth LLR challenge last year, one could finish such a WU in 2 minutes. Now my computer takes 6 to 9 minutes. The Woodall LLR runtime increased from 28 hours to 40 hours. But I don't know the estimated runtime function for LLRing.
It is not possible to speed up the WUs on multicore/hyperthreading machines, as all the programs are single threaded (and in fact, BOINC might get problems with multi threaded programs). You could only run multiple WUs at the same time, but this would definitely not help with your heat problems.
The Pentium D processors (like the Pentium 4, especially on 3 GHz) are producing quite a lot heat. High CPU temperature is usually not doing any harm. But if the temperature in your computer case gets too high, that can be serious. Hard drives and voltage control chips (don't know the proper english name for those, small chips around the CPU) are very sensitive to temperature.
You should consider investing in a better CPU cooler and multiple (!) silent case fans. I have a Core2Quad running 24/7 with LLR right next to my bed and I have no problems sleeping. I am using 3 fans (Scythe SFlex 800rpm 120x120x25) which you can barely hear, one of them as a CPU cooler, 2 as case coolers (1 in, 1 out).
My mainboard chipset (P965) is specified for 1066 only , but running at 1333 MHz due to the newer CPU. My computer is permanently running fully loaded since 5 months and I also use it for gaming (ATi 4850). No problems so far.
@all: as I am already LLRing all the time, this challenge should be no problem for me :)
(...)
Hi Roadrunner --- I will be very pious here --- " It servest thou self right for over clocking" (joke) --- I'm rather neutral to the new look page you commented on -- I do like the filters on the task page though
(...)
Nothing against the filters, but that is function not look.
The "cross-project stats" no makes the host list very long with 5 lines per host now whereas before it was 2 lines.
Will try pspLLR on my other hosts soon to see how they cope with the additional heat, seems like only 101424 and 77818 do the challenge since they are at work where air conditioning is applied.
DoES Volunteer tester Send message Joined: Oct 11 08 Posts: 289 ID: 30382 Credit: 1,883,604 RAC: 0
Nothing against the filters, but that is function not look.
The "cross-project stats" no makes the host list very long with 5 lines per host now whereas before it was 2 lines.
Hi Roadrunner--- Yeah -- the 2 given stats pages are of little interest to me (I had a look) -- I use 2 others if I get curious about where I'm at--- To me it is just extra lines of script-- but hey- whatever floats your boat--- I will continue to check "stats & stones" and "Boinc all projects Stats" as I have always done
As for the psychology of the Challenge, patience will be the virtue needed if you wish to participate. :) The average WU is 50 hours long (range: 28hrs to 100hrs). So, you can see the importance of having stable hardware.
do we have to face that checkpointig problem for linux again, or has it been fixed now?
I have just finished a 3*2^n+1 Proth LLR test on my single (only one computer)
GenuineIntel
Intel(R) Pentium(R) D CPU 3.00GHz [x86 Family 15 Model 6 Stepping 5]
The air-fan cooling system seems to be good but is halas too noisy when the two processors are running 100% with LLR or AP26 tests .After 2 hours or more you get stressed with the noise. I do not think, though, that processors or mother board are going to die.
when did you clean that fans and heatsinks last time?
I do not know if the processors are 64 or 32 bits and if vista Home Basic allows 64 bits and what factor of time i would gain if i could pass from 32 to 64 bits on LLR or AP26 tests.
yes, it's 64bit capable - no, no advantage on LLRs - yes, sieving is much faster with a 64-bit OS no matter if win64 or linux64.
i'm, still running one of those THINGS too, it finished the psp_sr2sieve wu's in about 46 minutes running ubuntu64 9.04
Will this challenge start at the normal 18:00 UTC or at 5:45 UTC so that it ends at the exact time of summer solstice, like last year's winter solstice challenge?
____________
John Forum moderator Project scientist Send message Joined: Feb 21 06 Posts: 1560 ID: 2449 Credit: 357,929 RAC: 0
Will this challenge start at the normal 18:00 UTC or at 5:45 UTC so that it ends at the exact time of summer solstice, like last year's winter solstice challenge?
05:45 UTC will be the start and finish of the Challenge to coincide with the Summer Solstice on 21 June 2009 05:45 UTC.
This thread has been renamed to "Birthday/ Summer Solstice Pre-Challenge" and the official Challenge thread has been created here: PrimeGrid's Birthday/Summer Solstice Challenge.
____________
