mersenneforum.org 10+ table
 Register FAQ Search Today's Posts Mark Forums Read

 2005-01-18, 13:59 #1 garo     Aug 2002 Termonfeckin, IE 32·307 Posts 10+ table Code: Size Base Index Mod Diff Ratio 242 10 323 + 323 0.749 295 10 332 + 332 0.888 300 10 346 + 346 0.867 318 10 347 + 347 0.916 300 10 353 + 353 0.849 292 10 356 + 356 0.82 332 10 358 + 358 0.927 329 10 359 + 359 0.916 289 10 365 + 292 0.989 /5q 283 10 373 + 373 0.758 305 10 374 + 340 0.897 /11 292 10 377 + 348 0.839 /13 278 10 379 + 379 0.733 324 10 382 + 382 0.848 366 10 383 + 383 0.955 348 10 386 + 386 0.901 265 10 392 + 336 0.788 /7 335 10 398 + 398 0.841 306 10 400 + 320 0.956 /5q 265 10 670L 231 10 710L 212 10 710M Last fiddled with by Batalov on 2021-01-06 at 01:24 Reason: 10,371+ is done!
2006-08-24, 05:21   #2
bdodson

Jun 2005
lehigh.edu

40016 Posts
p67 factor (not snfs ...)

Quote:
 Originally Posted by garo Part 3 of 3 Code: Base Index Size 11M(45digits) 43M(50digits) 110M(55digits) 260M(60digits) Decimal 10 381+ C214 0(0.267423) 0(0.0522979) 165(0.00921839) 0(0.00148122) 1009342896093598497779073565555297793634912097024613160705050252018362252646940962166733304631155103166829723922295200013852066602135544397087747985877524133291986126801290042969097513730326998650850605170907339801
We can drop further testing on this one --- the factor is out of ecm range
at 67-digits,

p67 = 4444349792156709907895752551798631908946180608768737946280238078881

Ooops, guess not! Lucky curve had sigma = 834412411, and the
curve order is again way-smooth at

[ <2, 2>, <3, 1>, <131, 1>, <124847, 1>, <1244459, 1>, <1785599, 1>,
<3000931, 1>, <4032877, 1>, <27225659, 1>, <29985143, 1>,
<87373729, 1>, <11805290281, 1> ]

So B1=260M wasn't used so much (B1=87.4M would have done),
and the large factor is under 200*(2nd largest); i.e., large step2
wasn't much used either. A new record 16-months of hard searching
since the previous (p66) record.

Bruce

2006-08-24, 09:49   #3
R.D. Silverman

Nov 2003

22·5·373 Posts

Quote:
 Originally Posted by bdodson We can drop further testing on this one --- the factor is out of ecm range at 67-digits, p67 = 4444349792156709907895752551798631908946180608768737946280238078881 Ooops, guess not! Lucky curve had sigma = 834412411, and the curve order is again way-smooth at [ <2, 2>, <3, 1>, <131, 1>, <124847, 1>, <1244459, 1>, <1785599, 1>, <3000931, 1>, <4032877, 1>, <27225659, 1>, <29985143, 1>, <87373729, 1>, <11805290281, 1> ] So B1=260M wasn't used so much (B1=87.4M would have done), and the large factor is under 200*(2nd largest); i.e., large step2 wasn't much used either. A new record 16-months of hard searching since the previous (p66) record. Bruce
Terrific!!!

 2006-08-24, 23:41 #4 akruppa     "Nancy" Aug 2002 Alexandria 1001101000112 Posts Woohoo!! Bruce stikes again! It was about time we saw an ECM p60+ this year. What took you so long? Thanks Bruce! Alex Edit: This warrants a Last fiddled with by akruppa on 2006-08-24 at 23:42
 2006-11-04, 11:29 #5 smh     "Sander" Oct 2002 52.345322,5.52471 29·41 Posts Finally, after lots of problems, i managed to finish the factorization of 10,372+ Code: N=27631128541915805055082181453641534739220599640437919826011911720853571851003653210276038758402805108684942992674414184678333002887846210747417 ( 143 digits) Divisors found: r1=23140616853203983900922551785166946660605063239337678349130406077337 (pp68) r2=1194053240550935343606131291791034479414833114386116350011658511441777011841 (pp76)
2006-11-09, 03:13   #6
geoff

Mar 2003
New Zealand

13×89 Posts

Quote:
 Originally Posted by smh Finally, after lots of problems, i managed to finish the factorization of 10,372+
Where did the problems occur? The one persistant problem I have is with the matrix not converging in matsolve, which seems to strike in about 50% of the larger GGNFS jobs.

2006-11-09, 14:25   #7
smh

"Sander"
Oct 2002
52.345322,5.52471

29×41 Posts

Quote:
 Originally Posted by geoff Where did the problems occur? The one persistant problem I have is with the matrix not converging in matsolve, which seems to strike in about 50% of the larger GGNFS jobs.
Matbuild kept on crashing with memory reallocation errors. I processed the relations several times, with a different number of relations, different factor bases and on different hardware (all had 4 - 12 GB of RAM) with no luck.

Someone form the GGNFS mailing list sent me a few programs to remove singletons. This apparently decreased memory usage so i was able to build the matrix.

I lost a week of matrix solving time because someone rebooted the server matsolve was running on.

When the matrix finally finished, sqrt produced trivial factorizations (with errors in screen output) on all dependencies.

I was sent a new version of sqrt, but that didn't work either.

I decided to reprocess the relations a second time, this time with a bit less relations (i had plenty the first time). After removing the singletons, the matrix build fine again.

matsolve took another 12 days or so and this time the factors were found on the 9th!! dependency.

2006-11-09, 17:53   #8
R.D. Silverman

Nov 2003

22×5×373 Posts

Quote:
 Originally Posted by smh Matbuild kept on crashing with memory reallocation errors. I processed the relations several times, with a different number of relations, different factor bases and on different hardware (all had 4 - 12 GB of RAM) with no luck. Someone form the GGNFS mailing list sent me a few programs to remove singletons. This apparently decreased memory usage so i was able to build the matrix. I lost a week of matrix solving time because someone rebooted the server matsolve was running on. When the matrix finally finished, sqrt produced trivial factorizations (with errors in screen output) on all dependencies. I was sent a new version of sqrt, but that didn't work either. I decided to reprocess the relations a second time, this time with a bit less relations (i had plenty the first time). After removing the singletons, the matrix build fine again. matsolve took another 12 days or so and this time the factors were found on the 9th!! dependency.

Hi,

Welcome to the woefull and wonderful wacky world of sieve post-processing.

We've all had our share of problems....

2006-11-09, 20:20   #9
smh

"Sander"
Oct 2002
52.345322,5.52471

100101001012 Posts

Quote:
 Originally Posted by R.D. Silverman Welcome to the woefull and wonderful wacky world of sieve post-processing. We've all had our share of problems....
But at least most of you knew what you were doing.

Although it was the first number that gave me problems with this release of GGNFS i'm hesitating to try another number from the cunningham list. It seems 140 -150 digits is really pushing the limit.

Are there any numbers left with a SNFS difficulty of around 200?

2006-11-10, 13:33   #10
R.D. Silverman

Nov 2003

22·5·373 Posts

Quote:
 Originally Posted by smh But at least most of you knew what you were doing. Although it was the first number that gave me problems with this release of GGNFS i'm hesitating to try another number from the cunningham list. It seems 140 -150 digits is really pushing the limit. Are there any numbers left with a SNFS difficulty of around 200?
Sure. There are plenty of them. All have composite exponents divisible
by 3,5,7, or 11.

e.g. 2,1630M, 11, 279+, 11,291+ (just over 200), 5, 411+, 5,429+,
6,369+, 2,1582L (just over 200) etc. etc.

2006-11-11, 11:12   #11
smh

"Sander"
Oct 2002
52.345322,5.52471

100101001012 Posts

Quote:
 Originally Posted by R.D. Silverman Sure. There are plenty of them. All have composite exponents divisible by 3,5,7, or 11.
Thanks.

I checked out a few of the with Alex's phi.exe (Is there a version that can handle the +1 side?).

This is new teritory to me so i've got a few questions:

Is deg. 4 poly suitable for numbers this size (SNFS dif. ~200)? How much harder is a deg.4 compared to a deg. 5 poly?

What would be a suitable factor base?

 Similar Threads Thread Thread Starter Forum Replies Last Post garo Cunningham Tables 100 2021-01-04 22:36 garo Cunningham Tables 86 2021-01-04 22:35 garo Cunningham Tables 80 2021-01-04 22:33 garo Cunningham Tables 82 2020-03-15 21:47 garo Cunningham Tables 41 2016-08-04 04:24

All times are UTC. The time now is 08:31.

Wed Apr 21 08:31:44 UTC 2021 up 13 days, 3:12, 0 users, load averages: 2.89, 2.92, 2.82