PrimeGrid: Year of the Snake Challenge Annoucement

[RSS]

With the recent change to a larger short range for GFN and also the change of the year to the year of the Snake, PrimeGrid would like to invite everyone to participate in the first challenge of 2013!A 13 day challenge is being offered on the Generalized Fermat project with short work units. This is only the second challenge of its kind we're offering, and because the work units will grow too big for a challenge after this it's mighty likely there won't be another one. So have you always dreamed of finding a gigantic prime (top 15 for sure, maybe even top 10!), then this is your chance! Join us from January 4 2013, 18:00 UTC till January 17 2013, 18:00 UTC and see what's going on.For more information on this challenge, please read this thread on our forums (http://www.primegrid.com/forum_thread.php?id=4760) and feel free to leave a comment or ask any questions you might have.

More... (http://www.primegrid.com/forum_thread.php?id=4761)

[Mike029]

Well, what do you guys think? Wanna give it a go? Pick up some badges along the way? Perhaps get a record Prime in the process?

[Duke of Buckingham]

We need to start the year in our best shape. I say let's go on that challenge.

[Mike029]

Yub yub, Checked the poll and looks like we are in. I've placed in on the calendar.

From the PrimeGrid Forums

Taken from the PrimeGrid forums. http://www.primegrid.com/forum_thread.php?id=4760

A Cautionary Reminder

ATTENTION: The primality programs Genefx64 (CPU) and GeneferCUDA (GPU) are both computationally intensive; so, it is vital to have a stable system with good cooling. They do not tolerate "even the slightest of errors." Please see this post for more details on how you can "stress test" your CPU, and please see this post for tips on running GFN on your GPU successfully.

As with all number crunching, excessive heat can potentially cause permanent hardware failure. Please ensure your cooling system is sufficient.

If you have never run GeneferCUDA before, please read these tips. GeneferCUDA stresses a GPU more than any other program I've seen and is incredibly intolerant of overclocking. It is strongly suggested that you run at least one (Short) Work Unit before the challenge to make sure your computer can handle it. Note that "Overclocking" includes GPUs that come overclocked from the factory, and lowering clock rates to the reference speeds is sometimes necessary. In the case of one particular GPU, the GTX 550 Ti, lowering the memory clock below the reference speed is often necessary.

WU's will take ~12 hours on a GTX 580 GPU and about 2 to 3 days on the fastest CPU core. A GTX 680 will be slower than a GTX 580 for this application. On my computer (GTX 460 & C2D Q6600), which is more of an "average" computer and certainly not the fastest, a Work Unit takes about 24 hours on the GPU and 6 to 8 days on the CPU.

If your CPU is highly overclocked, please consider "stress testing" it. Overclocking your GPU not recommended at all for GeneferCUDA. Sieving is an excellent alternative for computers that are not able to run Genefer. :)

Please, please, please make sure your machines are up to the task.

[Al]

They are not kidding. Run a wu or two before the start of the challenge and while you're around. The last time I ran them my 570s were pushing 98c at 100% fan when I finally shut it down. I have better air flow now, so that may have changed a bit for me. In any case, watch them while they're running.

[Fire$torm]

Well good thing its winter. The boxes in my garage will have little trouble keeping cool. OK, looks like I'm in.

@Carlos: I'll be back on NFS after this PG challenge.

[zombie67]

I have a 580, a 590, and a 560 Ti dual card, and I will put them all on this challenge. But none of my CPUs, for this one.

[STE\/E]

I'm Badge Hunting this Year :rolleyes: so since I already have over 16 Million Credits for the GFN's & the Ruby Badge I'll sit it out, looking at the BOINCstats Upcoming Team Challenges there's waaaaaaaay to many to try to run them all & most conflict with other Challenges. I think some people are getting Challenge Happy by Posting so many to keep their Pet Project in the Forefront especially for the SETI Project ...

[zombie67]

So....what are the best block size settings for 560 Ti and 580?

[Al]

So....what are the best block size settings for 560 Ti and 580?

Z - here's a post regarding block size. The gist, for the most part, is leave it at 0, the default.

http://www.primegrid.com/forum_thread.php?id=3982

I did have some screen lag on my personal system, so I lowered the Block Size to 6 (default 0=7 on Windows.) Seemed to take care of the lag. I'm running a few wus on each of my cards right now at stock clocks and voltages, with the memory downclocked from 1900 to 1200. Seems to be running pretty smooth and cool so far.

[Al]

I got through the first set of wus on all my nvidia cards.

The 570's completed 1 wu in 41,000 sec and haven't validated yet

The 560ti completed one wu in 42,000 sec for 24,988 credits - the Stderr output is below:
==================
Stderr output

<core_client_version>7.0.42</core_client_version>
<![CDATA[
<stderr_txt>
genefercuda 2.3.0-0 (Windows x86 CUDA 3.2)
Copyright 2001-2003, Yves Gallot
Copyright 2009, Mark Rodenkirch, David Underbakke
Copyright 2010-2012, Shoichiro Yamada, Ken Brazier
Copyright 2011-2012, Iain Bethune, Michael Goetz, Ronald Schneider

Command line: projects/www.primegrid.com/primegrid_genefer_2_3_0_0_1.07_windows_intelx86__c uda32_13.exe -boinc -q 37932^1048576+1 --device 0

Priority change succeeded.
GPU=GeForce GTX 560 Ti
Global memory=1341849600 Shared memory/block=49152 Registers/block=32768 Warp size=32
Max threads/block=1024
Max thread dim=1024 1024 64
Max grid=65535 65535 65535
CC=2.0
Clock=1464 MHz
# of MP=14
Using project preferences to override SHIFT; using 6 instead of 7

Starting initialization...
maxErr during b^N initialization = 0.0000 (0.380 seconds).
Testing 37932^1048576+1...
Estimated total run time for 37932^1048576+1 is 11:44:27
37932^1048576+1 is complete. (4801436 digits) (err = 0.0013) (time = 11:40:34) 21:29:10
21:29:10 (3216): called boinc_finish

</stderr_txt>
]]>

=====================

All of the genefers I have run have failed at least once with my wingmen, some two and three times. In looking at the Stderr outputs for their work units, most seem to be caused by overclocking. Don't overclock and you should be fine. The temps I was concerned about have been a non issue this time around for me. None of my gpus are running over 70c with 70% fan.

[Fire$torm]

Great info. Thanks Al.

[Al]

The other error I'm seeing is related to insufficient cuda version. Anything from 301.42 should be okay.

I've now completed 8, only 1 has validated so far. Wingman errors are rampant. Good thing this is a long challenge.

[Duke of Buckingham]

My system doesn't download tasks from that sub-project. The GPU was expected, I don't understand is why he doesn't download CPU but you are doing a good work.

GO TEAM GO

[Al]

There is no quorum of 2 during the challenge, so validations should not take long...I missed that. :)

Edit: 1 of the 570 wus validated at 25,000 credits.

[Wanabeelee]

I'm giving it a go, but not sure if I'll be able to help much with an estimated 35,000 hours per work unit on cpu and my duel gtx 9800's will not run the project. adding what I can where I can.

[Al]

I'm giving it a go, but not sure if I'll be able to help much with an estimated 35,000 hours per work unit on cpu and my duel gtx 9800's will not run the project. adding what I can where I can.

Let's hope the scheduler calculated that wrong. :eek:

[zombie67]

There is no quorum of 2 during the challenge, so validations should not take long...I missed that. :)

Edit: 1 of the 570 wus validated at 25,000 credits.

To be clear, the quorum *is* required for the actual BOINC credits that get published to the stats sites. You will not see credits until your wingman reports, and it validates with yours.

The quorum is NOT required to get credits in the challenge. When you return a task, you get the challenge credits awarded. It does not matter what your wing man does.

[Al]

To be clear, the quorum *is* required for the actual BOINC credits that get published to the stats sites. You will not see credits until your wingman reports, and it validates with yours.

The quorum is NOT required to get credits in the challenge. When you return a task, you get the challenge credits awarded. It does not matter what your wing man does.

Thanks for clearing that up Z.

[Wanabeelee]

Let's hope the scheduler calculated that wrong. :eek:

I hope so too... its running on a i7 960 at 3.8gh with 12gigs of ram.

4 hours in and at 2% so maybe 8 days for 8 work units... ouch..

[Mike029]

I hope so too... its running on a i7 960 at 3.8gh with 12gigs of ram.

4 hours in and at 2% so maybe 8 days for 8 work units... ouch..

Hoping the stones will be nice. ;)

[artemis8]

Looks like on my one wu I got about 160 points per hour.