mersenneforum.org Sieving drive for k=2000-3400 n=50K-1M
 Register FAQ Search Today's Posts Mark Forums Read

2009-06-01, 02:28   #23
Brucifer

Dec 2005

313 Posts

Quote:
 Originally Posted by MyDogBuster Thats because I have a hard time dividing without taking off my shoes.

Ha Ha! LOL Same here, only I also have to use my fingers as well as the toes.

 2009-06-01, 05:42 #24 Lennart     "Lennart" Jun 2007 100011000002 Posts Reserving 10T-10.9T Lennart ETA Jun 3 Last fiddled with by Lennart on 2009-06-01 at 06:02
 2009-06-01, 23:23 #25 gd_barnes     May 2007 Kansas; USA 5×2,017 Posts Thanks Lennart for the excellent boost on sieving! :-) A couple of notes for everyone here: First, depending on how quickly we progress on sieving here, I might suggest that we sieve it up to P=25T; IF that can be done by June 20th. Otherwise we'll stick with P=20T. P=25T is going to be slightly more efficient as for total sieving plus LLRing time up to n=600K but I don't want to delay the start of the new drive any more than the 20th. Second, Lennart has finished LLRing k=3000-3400 for n=20K-50K. Previously Karsten had already LLR'd k=2000-3000 for n=20K-40K (it's not in the sieve file right now) and he is in the process of LLRing n=40K-50K. Therefore I am suggesting that Max remove the n=20K-50K range from the sieve file and repost it on our noprimeleftbehind.net server. Once that is done, I'll post it here and you can choose whether to stop your sieve, replace your sieve file with the smaller one, and then restart it. Gary Last fiddled with by gd_barnes on 2009-06-01 at 23:23
2009-06-01, 23:25   #26
gd_barnes

May 2007
Kansas; USA

5×2,017 Posts

Quote:
 Originally Posted by MyDogBuster Thats because I have a hard time dividing without taking off my shoes.

Funny! :-) But aren't you using a full quad? If so, shouldn't it be P=125G vs. 250G per core?

I just realized that both of us were likely wrong there; unless you're using only 2 cores.

(lmao)

2009-06-02, 00:20   #27
MyDogBuster

May 2008
Wilmington, DE

1011000100102 Posts

Quote:
 Funny! :-) But aren't you using a full quad? If so, shouldn't it be P=125G vs. 250G per core? I just realized that both of us were likely wrong there; unless you're using only 2 cores.
Now I have to multiplication too? I'm definately having a brain cramp on this one. 500G divided by 4 = 125G times 4 = 500G. GOT IT.

Last fiddled with by MyDogBuster on 2009-06-02 at 00:23 Reason: Re-edited using 2 calculators on 2 machines

 2009-06-02, 00:27 #28 Brucifer     Dec 2005 313 Posts Reserving 10900G - 12000G
 2009-06-02, 01:10 #29 MyDogBuster     May 2008 Wilmington, DE B1216 Posts Reserving 12000G - 12400G 12000G / 6 * .3456 + 10955 -11546.2 comes to 100G per core Last fiddled with by MyDogBuster on 2009-06-02 at 01:35
2009-06-02, 01:18   #30
gd_barnes

May 2007
Kansas; USA

276516 Posts

Quote:
 Originally Posted by MyDogBuster Reserving 12000G - 12400G 12000G / 6 * .3456 + 10955 -75 comes to 100G per core
It looks like you're doing some fuzzy math there Ian. (Note to Max: You're not the only one I pick on for fuzzy math. lol) I'm not sure where you are getting these figures from but:

12000G / 6 * .3456 + 10955 -75 = 691,210,880 or 0.691G

A completely nonsensical answer so some of your figures must be incorrect. For instance, where did the .3456, 10955, and -75 figures come from? The 10955 and -75 have almost no effect on the total answer since they are so small and are additions or subtractions from something in the billions. I even tried them assuming they were billions and couldn't come close to what you came up with.

Plug exactly what you put there into Excel or any calculator and you'll get what I came up with.

Not that this matters a whip.

Last fiddled with by gd_barnes on 2009-06-02 at 01:23

2009-06-02, 01:51   #31
MyDogBuster

May 2008
Wilmington, DE

2·13·109 Posts

Quote:
 Plug exactly what you put there into Excel or any calculator and you'll get what I came up with.
Calculator or Excel, thats cheating. LMAO

 2009-06-02, 10:50 #32 Svenie25     Aug 2008 Good old Germany 8D16 Posts Would also take a range. 12400G-12650G, 125G per core. Hopefully I didnĀ“t make a mistak in the math. ;) Edit says: Would be nice to have the link for the sievefile. ;) Last fiddled with by Svenie25 on 2009-06-02 at 11:02
 2009-06-02, 13:14 #33 gd_barnes     May 2007 Kansas; USA 5·2,017 Posts Lennart, do you still have the sieve file posted on your web page? If so, could you give Sven a link to it? It will be a day or so before Max gets the file posted on the noprimeleftbehind.net server and has it set up so everyone has access to it. I tried it and it gave me a "permission denied" message. Once he gets that taken care of, he will have removed all n<50K from the file, which should speed sieving up by ~1-2%. (weepee!) :-) Sven, if Lennart doesn't still have it on his web page, I'll Email the file to you later today. Gary

 Similar Threads Thread Thread Starter Forum Replies Last Post gd_barnes No Prime Left Behind 61 2013-01-30 16:08 gd_barnes No Prime Left Behind 96 2012-02-19 03:53 gd_barnes No Prime Left Behind 11 2009-06-12 21:28 gd_barnes No Prime Left Behind 160 2009-05-10 00:50 gd_barnes No Prime Left Behind 118 2009-01-17 16:05

All times are UTC. The time now is 22:29.

Sun Mar 29 22:29:36 UTC 2020 up 4 days, 20:02, 3 users, load averages: 1.39, 1.45, 1.42