mersenneforum.org  

Go Back   mersenneforum.org > Prime Search Projects > Riesel Prime Search

Reply
 
Thread Tools
Old 2013-04-23, 16:21   #1
f1pokerspeed
 
Jun 2012

2×53 Posts
Default Help with sieving

Do you need any help with sieving Kosmaj? I'd be happy to throw a core or two at that for a while, just to give me something to do.

Is there a specific range that I could sieve, or just to do, say, n = 3m to 3.25m?
f1pokerspeed is offline   Reply With Quote
Old 2013-04-23, 16:54   #2
amphoria
 
amphoria's Avatar
 
"Dave"
Sep 2005
UK

2×19×73 Posts
Default

Quote:
Originally Posted by f1pokerspeed View Post
Do you need any help with sieving Kosmaj? I'd be happy to throw a core or two at that for a while, just to give me something to do.

Is there a specific range that I could sieve, or just to do, say, n = 3m to 3.25m?
We are using sieve files provided by PrimeGrid who are sieving k=3-9999 for both n-1 and n+1. n=3-6M is currently sieved to 274P (274e15) and they are still sieving.
amphoria is offline   Reply With Quote
Old 2013-04-23, 19:21   #3
f1pokerspeed
 
Jun 2012

10610 Posts
Default

Ok then - I guess PrimeGrid are going to trump anything I can do - ah well.
f1pokerspeed is offline   Reply With Quote
Old 2013-04-26, 03:06   #4
Kosmaj
 
Kosmaj's Avatar
 
Nov 2003

2×1,811 Posts
Default

Hi f1,

Thanks for your kind offer. As the others said, k<10000 are already extensively sieved, but you can help us extra sieving our candidates from our low-weight 10th drive, featuring k>10000.

Let me know if you are still interested.

Thanks!
Kosmaj is offline   Reply With Quote
Old 2013-04-27, 16:07   #5
f1pokerspeed
 
Jun 2012

2·53 Posts
Default

Sounds good to me! I can get started at any time.

EDIT: Having sewn the whole range together, I'm at 2.5M p/sec. So, I'm looking at 10G/hr or 1T/100 hrs of wall clock time. Expecting 1 factor every twenty hours, am I going to be working efficiently at this range?

Last fiddled with by f1pokerspeed on 2013-04-27 at 16:29
f1pokerspeed is offline   Reply With Quote
Old 2013-04-27, 16:39   #6
pinhodecarlos
 
pinhodecarlos's Avatar
 
"Carlos Pinho"
Oct 2011
Milton Keynes, UK

2·3·773 Posts
Default

Quote:
Originally Posted by f1pokerspeed View Post
Sounds good to me! I can get started at any time.

EDIT: Having sewn the whole range together, I'm at 2.5M p/sec. So, I'm looking at 10G/hr or 1T/100 hrs of wall clock time. Expecting 1 factor every twenty hours, am I going to be working efficiently at this range?
Did you receive the latest sieve file? If so try to make a LLR test on a candidate.
pinhodecarlos is offline   Reply With Quote
Old 2013-04-27, 16:51   #7
f1pokerspeed
 
Jun 2012

2·53 Posts
Default

The lowest candidate in the file takes ~1,100s (by quick extrapolation) - so this range looks to be optimal right now. The highest candidate is roughly the same - I'd say it's even oversieved ever so slightly.

It might be easier for me to just pre-sieve some later ranges, say n = 2m -> 3m. If that could be accommodated for, I would be happy to do some work on it.

Last fiddled with by f1pokerspeed on 2013-04-27 at 16:53
f1pokerspeed is offline   Reply With Quote
Old 2013-04-27, 21:33   #8
Kosmaj
 
Kosmaj's Avatar
 
Nov 2003

2·1,811 Posts
Default

Hi f1

Thanks again for your kind offer.
I'm enclosing the latest sieving file with candidates from the 10th drive, starting at n=1312k, sieved to 75T.

We are also preparing another file with high weight Ks for sieving. When it's ready we'll ask you to kindly stop this one and sieve the new one.

If you don't mind, can we establish e-mail connection for easier exchange of files.

Thanks!
Attached Files
File Type: zip rps10abcd.zip (448.4 KB, 46 views)
Kosmaj is offline   Reply With Quote
Old 2013-04-27, 22:07   #9
VBCurtis
 
VBCurtis's Avatar
 
"Curtis"
Feb 2005
Riverside, CA

4,243 Posts
Default

Quote:
Originally Posted by f1pokerspeed View Post
The lowest candidate in the file takes ~1,100s (by quick extrapolation) - so this range looks to be optimal right now. The highest candidate is roughly the same - I'd say it's even oversieved ever so slightly.

It might be easier for me to just pre-sieve some later ranges, say n = 2m -> 3m. If that could be accommodated for, I would be happy to do some work on it.
f1-
Generally speaking, sieve ranges smaller than a factor of two from n-min to n-max are not very efficient. This is an effect of the sieve speed scaling with the square-root of n-range, combined with LLR testing scaling with the square of n.

So, while this sieve from 1300 to 2000 is probably properly sieved, you are correct that not only would sieving 2-3m be of use, but once 75T is reached you could combine the current sieve and run from 1300 to 3000 with more efficiency than either 1300-2000 or 2000-3000. If demand supported it, I'd even go with 2-4 million.

However, that would be quite a large project, and it's not clear that this low-weight drive would have sufficient interest to bother with a new sieve up to the 100T range. With primegrid sieving over 1000 times deeper (10 bits, at almost 2% candidates removed per bit, so 15 to 18% fewer candidates than 100T), it's hard to believe we'll have sufficient interest to justify a massive effort on the 10th drive.

That said, I'll be happy to participate in a future high-weight-k sieve and drive, and hope you will too.
-Curtis

Last fiddled with by VBCurtis on 2013-04-27 at 22:08
VBCurtis is offline   Reply With Quote
Old 2013-04-27, 22:43   #10
pepi37
 
pepi37's Avatar
 
Dec 2011
After milion nines:)

2×5×131 Posts
Default

What is "chance" to find any prime in this low weight K-s?
pepi37 is online now   Reply With Quote
Old 2013-04-27, 23:41   #11
pinhodecarlos
 
pinhodecarlos's Avatar
 
"Carlos Pinho"
Oct 2011
Milton Keynes, UK

2×3×773 Posts
Default

Quote:
Originally Posted by pepi37 View Post
What is "chance" to find any prime in this low weight K-s?
Check second table here (http://www.mersenneforum.org/showpos...64&postcount=2).

Carlos

Last fiddled with by pinhodecarlos on 2013-04-27 at 23:41
pinhodecarlos is offline   Reply With Quote
Reply

Thread Tools


Similar Threads
Thread Thread Starter Forum Replies Last Post
NFS sieving? Dubslow Factoring 8 2012-09-28 06:47
Line sieving vs. lattice sieving JHansen NFSNET Discussion 9 2010-06-09 19:25
10^420 + 1 sieving juno1369 Factoring 20 2010-04-28 01:11
Sieving OmbooHankvald Prime Sierpinski Project 4 2005-06-30 07:51
Sieving robert44444uk Sierpinski/Riesel Base 5 8 2005-04-02 22:30

All times are UTC. The time now is 08:57.

Thu Aug 13 08:57:34 UTC 2020 up 5:33, 0 users, load averages: 1.23, 1.39, 1.60

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2020, Jelsoft Enterprises Ltd.

This forum has received and complied with 0 (zero) government requests for information.

Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation.
A copy of the license is included in the FAQ.