LLRnet/PRPnet rally April 4th-11th

[gd_barnes]

Originally Posted by Flatlander

Is it showing stats from the last rally?

Dave Emailed me back to confirm what you noticed. The wrong rally DB is being updated. He's at work and so won't be able to get it fixed for 10-12 hours.

[Rick Reynolds]

Here's the prpnet port I'm running for the rally
server=NPLB_9000:100:2:noprimeleftbehind.net:9000

however in the results page I don't see either me, Neo or our team Aggie The Pew

was wondering if I'm doing something wrong or have I wasted a 1/2 day of crunching on the wrong port?

Thanks Rick

[Rick Reynolds]

Sorry, please ignore my first post. I just checked and we are now on the result page... guess I was just to impatient.

Rick

[AMDave]

Originally Posted by mdettweiler

Hey Dave, looks like the gremlins aren't quite done with you yet...I'm not seeing any data on the rally page, despite the first DB refresh of the rally having already occurred. Any idea what's up?

Apologies for the weird time warp.

There was some bad grammar in my foundation incantation which incarnated some gremlins by accident.

The gremlins have now been vanquished with a magical 'debug' spell.

All's well as the prime hunt crashes forcibly and inexorably through the dark undergrowth of the K-and-N forest.
BAOOOOO! BAOOOOO!
[Stage direction: sound of baying hounds and the thundering of hooves]

Beware you elusive little primes, your time is nigh.

[mdettweiler]

Originally Posted by AMDave

Apologies for the weird time warp.

There was some bad grammar in my foundation incantation which incarnated some gremlins by accident.

The gremlins have now been vanquished with a magical 'debug' spell.

All's well as the prime hunt crashes forcibly and inexorably through the dark undergrowth of the K-and-N forest.
BAOOOOO! BAOOOOO!
[Stage direction: sound of baying hounds and the thundering of hooves]

Beware you elusive little primes, your time is nigh.

Ah, much better. Thanks for getting all that ironed out!

Now to find some primes...perhaps I'll at long last fine one in this rally. (Of course, it would help if I didn't accidentally leave one of my quad's cores idle when I moved them to port 9000...need to fix that tonight.)

Meanwhile, Lennart is posting some impressive hourly stats...600 pairs/hour! That's incredible. Last rally he just managed 400 pairs/hour--though to be fair, the tests were bigger that time, so the amount of underlying resources is probably about the same.

Also yet to be seen is whether our friends from AMD Users and Free-DC (Vaughan, James, and Brian), who so resoundingly dominated the last couple of rallies, will make an appearance. You guys coming?

[Lennart]

Originally Posted by mdettweiler

Ah, much better. Thanks for getting all that ironed out!


Meanwhile, Lennart is posting some impressive hourly stats...600 pairs/hour! That's incredible. Last rally he just managed 400 pairs/hour--though to be fair, the tests were bigger that time, so the amount of underlying resources is probably about the same.

I have upgraded 5 q6600 to 2600k. Two new i7 970 are also added :)

Lennart

[Ralf Recker]

Originally Posted by mdettweiler

Also yet to be seen is whether our friends from AMD Users and Free-DC (Vaughan, James, and Brian), who so resoundingly dominated the last couple of rallies, will make an appearance. You guys coming?

The first one has arrived...

[Rick Reynolds]

Don't suppose there's any 32 bit cuda enabled llr versions running around for this rally or is that even possible here?

Rick

[Ralf Recker]

Originally Posted by Rick Reynolds

Don't suppose there's any 32 bit cuda enabled llr versions running around for this rally or is that even possible here?

Rick

In theory it should work:

Core 2 Q9550 @ 3.4 GHz

885*2^846293-1 is not prime. LLR Res64: 46AF89C695A35A07 Time : 594.089 sec.
[2011-04-05 16:33:14 CEST] G9000: 885*2^846293-1 is not prime. Residue 46AF89C695A35A07

llrCUDA 0.6.0 (270.26beta drivers, Linux 64 bit) on a GTX 470 @ 607 MHz:

885*2^846293-1 is not prime. LLR Res64: 46AF89C695A35A07 Time : 593.996 sec.

The problem is: The CUDA app as fast as a core here on and requires nearly a core to do it's work. So the gain is zero at the cost of a significantly higher power consumption.

[Rick Reynolds]

Originally Posted by Ralf Recker

In theory it should work:

Core 2 Q9550 @ 3.4 GHz

885*2^846293-1 is not prime. LLR Res64: 46AF89C695A35A07 Time : 594.089 sec.
[2011-04-05 16:33:14 CEST] G9000: 885*2^846293-1 is not prime. Residue 46AF89C695A35A07

llrCUDA 0.6.0 (270.26beta drivers, Linux 64 bit) on a GTX 470 @ 607 MHz:

885*2^846293-1 is not prime. LLR Res64: 46AF89C695A35A07 Time : 593.996 sec.

The problem is: The CUDA app as fast as a core here on and requires nearly a core to do it's work. So the gain is zero at the cost of a significantly higher power consumption.

True but I was really looking for someone with a 32 bit version since I can't convert to a 64 bit yet. Thanks for the info however.

that 607 on your 470 is that the core and if so what are you running your shaders at? I have a 460 running at 1512 on the shaders.
Rick

[Ralf Recker]

Originally Posted by Rick Reynolds

True but I was really looking for someone with a 32 bit version since I can't convert to a 64 bit yet. Thanks for the info however.

that 607 on your 470 is that the core and if so what are you running your shaders at? I have a 460 running at 1512 on the shaders.
Rick

The GTX 470 runs at stock clocks: I'm currently testing my tpsieve changes on it. I initially tested them on a GTX 460 (GF 104 chip, Compute Capability 2.1) and now trying out a few changes on a GF 100 chip (Compute Capability 2.0) based GTX 470 card. Clocks are the standard: 607 MHz Core, 1215 MHz Shaders (2x core clocks, as usual) and 1674 MHz memory.