mersenneforum.org  

Go Back   mersenneforum.org > Prime Search Projects > No Prime Left Behind

Reply
 
Thread Tools
Old 2009-02-07, 05:06   #45
gd_barnes
 
gd_barnes's Avatar
 
May 2007
Kansas; USA

1009410 Posts
Default

Mini,

Thanks for posting.

A couple of items: The Generalized Woodall Search is not new and Les GeneFermiers did not drop off the top 20 since last time. I made the corrections.

Everyone should also check out this post that I just now edited for a total scores change since we first started this thread. We've now moved into 5th place there and are gaining rapidly on 4th place. :-)


Gary

Last fiddled with by gd_barnes on 2009-02-07 at 05:24
gd_barnes is offline   Reply With Quote
Old 2009-02-07, 12:27   #46
Mini-Geek
Account Deleted
 
Mini-Geek's Avatar
 
"Tim Sorbera"
Aug 2006
San Antonio, TX USA

17·251 Posts
Default

Quote:
Originally Posted by gd_barnes View Post
Mini,

Thanks for posting.

A couple of items: The Generalized Woodall Search is not new and Les GeneFermiers did not drop off the top 20 since last time. I made the corrections.
Oh, oops, I must've made a mistake in copying because they were there when I last updated (I based my update on the last spreadsheet I had, which had Les GeneFermiers, then updated all numbers and noticed that Les GeneFermiers was missing), I hope that didn't throw off anything else with the stats or changes. They fell off the top 20 in your last update.
Quote:
Originally Posted by gd_barnes View Post
Everyone should also check out this post that I just now edited for a total scores change since we first started this thread. We've now moved into 5th place there and are gaining rapidly on 4th place. :-)


Gary
Mini-Geek is offline   Reply With Quote
Old 2009-02-22, 06:30   #47
gd_barnes
 
gd_barnes's Avatar
 
May 2007
Kansas; USA

2·72·103 Posts
Default

We've now topped a score of 2000 as our outstanding performance continues! The most remarkable thing about this is that we've done it while also dedicating a large # of resources to finding smaller non-top-5000 primes as well as double-checking and sieving; all at the same time and WITHOUT a rally!

What a great group we have! Thanks to all for your continued efforts to fill in all the holes in primes of all sizes!

scores:
Code:
6856746 Great Internet Mersenne Prime Search by Woltman & Kurowski
 106656 Seventeen or Bust
   7076 PrimeGrid
   6096 Riesel Prime Search
   5195 Riesel Sieve Project
   4515 The Prime Sierpinski Problem
   2154 321search
   2002 PrimeSearch
   1718 Yves Gallot's GFN Search Project
    967 Prime Internet Eisenstein Search
    938 GFN 2^17 Sieving project
    847 12121 Search
    272 15k*2^n-1 search
    252 Mat's Prime Search
    231 Riesel Base 5
    209 GFN 2^16 Sieving project
    153 Conjectures 'R Us
     47 Generalized Woodall Prime Search
     41 GFN 2^15 Sieving project
differences:
Code:
    +83 PrimeSearch
     -2 GFN 2^15 Sieving project
     -3 Generalized Woodall Prime Search
     -5 Conjectures 'R Us
     -8 GFN 2^16 Sieving project
     -9 Mat's Prime Search
    -10 Riesel Base 5
    -15 15k*2^n-1 search
    -32 12121 Search
    -36 GFN 2^17 Sieving project
    -38 Prime Internet Eisenstein Search
    -76 Yves Gallot's GFN Search Project
    -84 321search
   -137 Riesel Prime Search
   -157 PrimeGrid
   -174 The Prime Sierpinski Problem
   -201 Riesel Sieve Project
  -4109 Seventeen or Bust
-264182 Great Internet Mersenne Prime Search by Woltman & Kurowski
A score gain of 220 vs. RPS in the latest 2 week period and a gain of 853 vs. RPS in the last 4 months!! It won't be long until we're in 7th place.

Everyone, be sure and check out this post for the updated total scores change since we started this thread.


Gary

Last fiddled with by gd_barnes on 2009-04-07 at 04:23
gd_barnes is offline   Reply With Quote
Old 2009-04-07, 04:22   #48
gd_barnes
 
gd_barnes's Avatar
 
May 2007
Kansas; USA

2×72×103 Posts
Default

I've held off quite a while seeing if it would continue and it has: Our outstanding performance continues! Once again, thanks to all for your continued efforts to fill in all the holes in primes of all sizes!

Check out the total domination score-wise of our project vs. all others, which includes moving from 8th to 7th place in the last time frame!

scores:
Code:
5955464 Great Internet Mersenne Prime Search by Woltman & Kurowski
  92637 Seventeen or Bust
   7058 PrimeGrid
   5671 Riesel Prime Search
   4512 Riesel Sieve Project
   3922 The Prime Sierpinski Problem
   2098 PrimeSearch
   1871 321search
   1491 Yves Gallot's GFN Search Project
    846 Prime Internet Eisenstein Search
    814 GFN 2^17 Sieving project
    734 12121 Search
    226 15k*2^n-1 search
    219 Mat's Prime Search
    195 Riesel Base 5
    182 GFN 2^16 Sieving project
    133 Conjectures 'R Us
     43 Generalized Woodall Prime Search
     36 GFN 2^15 Sieving project
differences:
Code:
    +96 PrimeSearch
     -4 Generalized Woodall Prime Search
     -5 GFN 2^15 Sieving project
    -18 PrimeGrid
    -20 Conjectures 'R Us
    -27 GFN 2^16 Sieving project
    -33 Mat's Prime Search
    -36 Riesel Base 5
    -46 15k*2^n-1 search
   -113 12121 Search
   -121 Prime Internet Eisenstein Search
   -124 GFN 2^17 Sieving project
   -227 Yves Gallot's GFN Search Project
   -283 321search
   -425 Riesel Prime Search
   -593 The Prime Sierpinski Problem
   -683 Riesel Sieve Project
 -14019 Seventeen or Bust
-901282 Great Internet Mersenne Prime Search by Woltman & Kurowski
The continued huge # of primes found by PrimeGrid is raising the 5000th place prime so high that it is almost impossible for any single project to gain score including PrimeGrid themselves! :surprised

This amounts to a score gain of 521 vs. RPS in the latest 1-1/2 month period and a gain of 1374 vs. RPS in the last nearly 5-1/2 months!! Thanks, PrimeGrid, for the big boost in knocking the score down on other projects!

Check out this post for the updated total scores change since we started this thread.


Gary
gd_barnes is offline   Reply With Quote
Old 2009-04-08, 14:04   #49
henryzz
Just call me Henry
 
henryzz's Avatar
 
"David"
Sep 2007
Cambridge (GMT)

24×353 Posts
Default

brilliant news
i was a little concerned by RPS's new 7th and 8th drives but it seems we are still beating them so far
we need to keep searching relatively low(n<700k) ranges to keep lots of primes and points coming on the table
i think it is less efficient to put lots of resources into searching for really high primes(n>1M) currently
the 8th drive helped massively

Last fiddled with by henryzz on 2009-04-08 at 14:05
henryzz is offline   Reply With Quote
Old 2009-04-09, 00:59   #50
gd_barnes
 
gd_barnes's Avatar
 
May 2007
Kansas; USA

2×72×103 Posts
Default

Quote:
Originally Posted by henryzz View Post
brilliant news
i was a little concerned by RPS's new 7th and 8th drives but it seems we are still beating them so far
we need to keep searching relatively low(n<700k) ranges to keep lots of primes and points coming on the table
i think it is less efficient to put lots of resources into searching for really high primes(n>1M) currently
the 8th drive helped massively
It depends on what you are after. Technically, you will get the same prime SCORE over the long run by testing low or high n-ranges if they are sieved to the same depth. BUT...high n-ranges are almost always sieved deeper, due to the longer testing times so...

It's slightly more efficient for prime SCORE to test higher n-ranges because more factors are eliminated ahead of time and the primes won't drop off the top-5000 list anytime soon.

It's always more efficient for # of primes to test lower n-ranges as long as those ranges aren't so low on the "top-5000 curve" that the primes are dropping off quickly, hence causing you to gain little over a period of months.

NPLB of course leans towards the latter but there are people of many different tastes and tolerances so we want to keep the higher n-ranges for people who like fewer but larger primes. At the opposite end, several people don't really care about messing with reporting top-5000 primes so they orient towards sieving and testing for smaller primes. Bruce is our best 'dirty work' guy in that regard. Thanks Bruce! Sieving, finding smaller primes, and double-checking don't get the recognition that they deserve.

If we really wanted variety, we'd sieve n=1M-2M and have n>1M available for people also but that's such a large effort that the project hasn't undertaken it just yet. Bruce and I had a brief PM discussion about it and after Karsten chimed in that he had started sieving k=2000-3000 for n=20K-1M, we thought it made sense to go that route. The sieving for n=1M-2M will likely be around the corner, sometime early to mid 2010.

RPS's 7th/8th drives are miniscule compared to any of ours. You're talking 30-40 k's combined on the 2 drives vs. 300 k's on our 10th drive and 100 k's each on our 5th/6th/7th drives. With what appears to be the equivalent of about 32-40 modern quads currently running port 8000 at all times right now, we won't have any problem overtaking them. And that doesn't even include all of the effort that Ian, Brian, Chris, Karsten, etc. are doing on the individual-k effort and their own "side efforts" that have yielded some very nice primes lately. Just a few days ago, Brian had 3 n>700K primes on his side effort within about 2 days of each other.

I estimate that we have about 200-250 cores running for us full time right now including cores dedicated to sieving. Thanks all to a great team!


Gary
gd_barnes is offline   Reply With Quote
Old 2009-04-09, 01:46   #51
IronBits
I ♥ BOINC!
 
IronBits's Avatar
 
Oct 2002
Glendale, AZ. (USA)

111310 Posts
Default

Teams as in plural I think

Quote:
I estimate that we have about 200-250 cores running for us full time right now including cores dedicated to sieving. Thanks all to a great team!
IronBits is offline   Reply With Quote
Old 2009-05-03, 07:30   #52
gd_barnes
 
gd_barnes's Avatar
 
May 2007
Kansas; USA

2·72·103 Posts
Default

This probably should have been done on May 1st to see how the Free-DC boost for the complete month of May helps us but now should be sufficient.

No words are sufficient to describe our performance recently:

scores:
Code:
5369059 Great Internet Mersenne Prime Search by Woltman & Kurowski
  83515 Seventeen or Bust
  14397 PrimeGrid
   5323 Riesel Prime Search
   4068 Riesel Sieve Project
   3535 The Prime Sierpinski Problem
   2409 No Prime Left Behind (formerly: PrimeSearch)
   1687 321search
   1344 Yves Gallot's GFN Search Project
    762 Prime Internet Eisenstein Search
    734 GFN 2^17 Sieving project
    661 12121 Search
    202 15k*2^n-1 search
    197 Mat's Prime Search
    173 Riesel Base 5
    164 GFN 2^16 Sieving project
    117 Conjectures 'R Us
     39 Generalized Woodall Prime Search
     32 GFN 2^15 Sieving project
differences:
Code:
  +7339 PrimeGrid
   +311 No Prime Left Behind (formerly: PrimeSearch)
     -4 Generalized Woodall Prime Search
     -4 GFN 2^15 Sieving project
    -16 Conjectures 'R Us
    -18 GFN 2^16 Sieving project
    -22 Mat's Prime Search
    -22 Riesel Base 5
    -24 15k*2^n-1 search
    -73 12121 Search
    -80 GFN 2^17 Sieving project
    -84 Prime Internet Eisenstein Search
   -147 Yves Gallot's GFN Search Project
   -184 321search
   -348 Riesel Prime Search
   -387 The Prime Sierpinski Problem
   -444 Riesel Sieve Project
  -9122 Seventeen or Bust
-586405 Great Internet Mersenne Prime Search by Woltman & Kurowski
PrimeGrid fround 2 monster primes to distance itself in 3rd place. This was to be expected for a well-run and coordinated BOINC project. Our objective is to compete against non-BOINC projects.

Don't forget to check out this post for the updated total scores change since we started this thread.


Gary
gd_barnes is offline   Reply With Quote
Old 2009-06-04, 00:31   #53
gd_barnes
 
gd_barnes's Avatar
 
May 2007
Kansas; USA

2·72·103 Posts
Default

Well, here we are a month later after the Free-DC month-long rally on our project. The boost was outstanding and with no other projects finding huge primes, we easily topped all projects on score gain.

Our score was over 2600 for quite a while. Unfortunately we started dropping rapidly towards the end of the month after Free-DC cleared out and started a new project plus all of our main drives are now searching n>660K. With a large part of our resources now on sieving at the moment, June promises to be a very challenging month for us primes-wise until the sieving effort is done. We will be doing well if we can avoid losing more than 200 points in score this month.

The good news is that the new drives from the huge sieving effort will be huge and we should get a large infusion of primes once we start them.

scores:
Code:
4589297 Great Internet Mersenne Prime Search by Woltman & Kurowski
  71386 Seventeen or Bust
  12807 PrimeGrid
   4809 Riesel Prime Search
   3477 Riesel Sieve Project
   3020 The Prime Sierpinski Problem
   2539 No Prime Left Behind (formerly: PrimeSearch)
   1442 321search
   1151 Yves Gallot's GFN Search Project
    655 Prime Internet Eisenstein Search
    627 GFN 2^17 Sieving project
    564 12121 Search
    168 Mat's Prime Search
    157 15k*2^n-1 search
    144 Riesel Base 5
    143 GFN 2^16 Sieving project
     99 Conjectures 'R Us
     65 Free-DC's Prime Search
     33 Generalized Woodall Prime Search
     28 GFN 2^15 Sieving project
differences:
Code:
   +130 No Prime Left Behind (formerly: PrimeSearch)
    +65 Free-DC's Prime Search
     -4 GFN 2^15 Sieving project
     -6 Generalized Woodall Prime Search
    -18 Conjectures 'R Us
    -21 GFN 2^16 Sieving project
    -29 Mat's Prime Search
    -29 Riesel Base 5
    -45 15k*2^n-1 search
    -97 12121 Search
   -107 Prime Internet Eisenstein Search
   -107 GFN 2^17 Sieving project
   -193 Yves Gallot's GFN Search Project
   -245 321search
   -514 Riesel Prime Search
   -515 The Prime Sierpinski Problem
   -591 Riesel Sieve Project
  -1590 PrimeGrid
 -12129 Seventeen or Bust
-779762 Great Internet Mersenne Prime Search by Woltman & Kurowski
Updated total scores change have now been updated in this post.


Gary
gd_barnes is offline   Reply With Quote
Old 2009-12-15, 18:17   #54
henryzz
Just call me Henry
 
henryzz's Avatar
 
"David"
Sep 2007
Cambridge (GMT)

24×353 Posts
Default

maybe it is time for an update
henryzz is offline   Reply With Quote
Reply

Thread Tools


Similar Threads
Thread Thread Starter Forum Replies Last Post
Best msieve poly scores VBCurtis Msieve 160 2020-03-13 12:19
4000 < k < 5000 otutusaus Riesel Prime Data Collecting (k*2^n-1) 5 2012-03-07 20:01
5000 < k < 6000 justinsane Riesel Prime Data Collecting (k*2^n-1) 26 2010-12-31 12:27
As Math Scores Lag, a New Push for the Basics -- New York Times ewmayer Soap Box 14 2006-11-17 09:40
Top-5000 List edorajh Riesel Prime Search 17 2006-03-28 21:57

All times are UTC. The time now is 22:49.

Sat Apr 4 22:49:43 UTC 2020 up 10 days, 20:22, 0 users, load averages: 1.01, 1.16, 1.32

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.