mersenneforum.org

mersenneforum.org (https://www.mersenneforum.org/index.php)
-   GPU to 72 (https://www.mersenneforum.org/forumdisplay.php?f=95)
-   -   The "(W)TF Depth Debate" state of the bunion (https://www.mersenneforum.org/showthread.php?t=18975)

LaurV 2013-12-08 13:03

[QUOTE=chalsall;361361]Are you sure you can't handle more? I'm serious.
I had planned on bringing three other R720s over to you today.[/QUOTE]
On a second thought, the P-1 work reported by the "aliens" accounted like 49.94G from Kracker, including the factors, and 86.58G from you, which is a total of 136.54G, so I am still ahead, with (currently) 1517.3G of TF, hehe, so you can bring more power in. :razz:

Joking apart, I thing you can talk [B][U]nucleon[/U][/B] into it. And better give him the other 3 computers. His firing power is also 80% (or more) GPU, and I remember there was a time when he used to lead the TF tops from far away. He also has like eight times the power of a 580 or so (I think he used to have 580s but upgraded them to titans, or so). He is mostly doing [URL="http://www.gpu72.com/reports/worker/fc9f090d094ad7c7eff10a39caffe3a4/"]80-120G of P-1 per day[/URL], since cudapm1 is out in the wild, so if you can talk him into TF (and I will support you with this), then we are good for a while, we can keep the LL wave behind until Jerry recovers from his "water broken" :smile:

chalsall 2013-12-08 15:40

[QUOTE=LaurV;361461]On a second thought, the P-1 work reported by the "aliens" accounted like 49.94G from Kracker, including the factors, and 86.58G from you, which is a total of 136.54G, so I am still ahead, with (currently) 1517.3G of TF, hehe, so you can bring more power in. :razz:[/QUOTE]

You got it! R720_1, 3 and 5 are now "yours".

And, yes, I've already contacted nucleon.

LaurV 2013-12-10 15:38

I keep clicking on his name regularly on GPU72, hoping to see some green, or at least blue, hehe, but everything is red and doesn't seem to move since few days. He might be in holiday...

Otoh, the "accounting" today, signaled that you are overtaking me. One hour ago, kracker had ~153G pumped (and 6 exponents cleared by finding factors), and Chris had ~306G, (and 5 factors). That would be a total of 459.7G, and an equivalent of 3678G of TF when multiplied by 8. Meantime, my cards were pumping ~900G per day, for a total of 3396G, which means that I am a bit behind. :smile: Also, by the "clearing" measure, I am also behind, with only 3 factors. Also, clocker's computer is at the end of stage2 for 4 expos, so it will spit aout about 16-20G in the following hours... Grrr... I have to think seriously of adding more power into it.

The good part is that I raised in the top, only 6 places to go, hehe, ETA about one week. After that, you can take back the computers, and I will continue to do TF for a while - I found out that the computer is more usable when TF-ing so I can keep it running.

As a curiosity, if we keep this speed (all my CPUs are PM1-ing too, beside of the "aliens") I will reach second place in about... 920 days :razz: (i.e. I have now the "highest" speed, except Oliver, who is still unreachable, behind of the ">999" limit)

kracker 2013-12-10 15:53

I only have half CPU power running for you... you sure you don't want to keep this up? :razz:

chalsall 2013-12-10 16:08

[QUOTE=LaurV;361670]Grrr... I have to think seriously of adding more power into it. [/QUOTE]

Please do! :smile: We haven't had to release any candidates at 73 yet, but it's *really* tight!

[QUOTE=LaurV;361670]After that, you can take back the computers, and I will continue to do TF for a while - I found out that the computer is more usable when TF-ing so I can keep it running.[/QUOTE]

Thanks for offering to continue to do the TFing. I'm happy to keep "trading" for as long as you are producing TF results.

LaurV 2013-12-10 16:21

[QUOTE=chalsall;361674]We haven't had to release any candidates at 73 yet[/QUOTE]
Apropos of that, I am getting about two thirds of the assignments from 72 to 74, but a third of the assignments from only 72 to 73 (?!? no 74!). Are you sure you didn't feed any alcohols to that spider? :razz:

edit: [B][U]correction:[/U][/B] right now I have only about 16 assignments 72-74, everything else coming after them are 72-73 only. Which seems a bit illogical for me, as we want to raise as many as possible exponents to 74.

chalsall 2013-12-10 16:26

[QUOTE=LaurV;361676]Are you sure you didn't feed any alcohols to that spider? :razz:[/QUOTE]

Absolutely. :wink: I'm balancing the assignments between the two, to ensure we have 73's ready to release if we need to. It's a bit of a juggling act, since there's latency between when the assignment is issued, and it's returned.

LaurV 2013-12-10 16:28

ok then, everything under control :D

Edit, P.S.: I found out why we are [URL="http://www.gpu72.com/reports/overall/graph/"]going down[/URL] with the factoring power, that is because you use a "regression line". To go up, you must use a "progression line" :razz:

Mark Rose 2013-12-10 18:24

[QUOTE=chalsall;361674]Please do! :smile: We haven't had to release any candidates at 73 yet, but it's *really* tight![/QUOTE]

I guess I'll give up finding factors and switch to TF'ing to 74 for a bit :P

I should be TF'ing to 74 by tomorrow.

chalsall 2013-12-10 18:28

[QUOTE=Mark Rose;361694]I should be TF'ing to 74 by tomorrow.[/QUOTE]

Coolness. Thanks.

And you'll still be finding factors, just not as many per "wall-clock time". It's about 1% per "bit level", but it takes twice as long for each level. (I'll leave it to others to explain the math.... :wink:)

Mark Rose 2013-12-10 18:55

[QUOTE=chalsall;361695]Coolness. Thanks.

And you'll still be finding factors, just not as many per "wall-clock time". It's about 1% per "bit level", but it takes twice as long for each level. (I'll leave it to others to explain the math.... :wink:)[/QUOTE]

So if I find 1 factor per time x of factoring to 72, I should find 3 per (1+2+4)x factoring to 73, or about 57% fewer factors.


All times are UTC. The time now is 09:42.

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2021, Jelsoft Enterprises Ltd.