Running GGNFS - Page 37

fivemack · 2010-02-26, 12:07

Quote:

Also, x^5-843714919 is the correct one.

Ah, good point, I was confusing sigma(x^18) and x^18-1 ... that would make the algebraic side around 10^44 for the quintic, so that's the way to go. Probably not worth restarting the job.

fivemack · 2010-02-26, 14:32

Not quite sure what happened there, I think I must have edited apocalypse's post rather than replying to it.

siew · 2010-03-11, 16:53

Hi!

Received following messages today:

Code:

FBsize 1149879+0 (deg 5), 1151366+0 (deg 1)
total yield: 572566, q=43000019 (0.85104 sec/rel)
113431 Special q, 367095 reduction iterations
reports: 316699993->3967808970->3499071217->617340989->189940885->79581099
Number of relations with k rational and l algebraic primes for (k,l)=:

Total yield: 572566
milliseconds total: Sieve 294360313 Sched 0 medsched 21601276
TD 51687587 (Init 801049, MPQS 80555) Sieve-Change 85867914

TD side 0: init/small/medium/large/search: 823253 3152762 1637884 10636302 12641
999

sieve: init/small/medium/large/search: 3001799 39113818 2193568 95135990 2269319
8

TD side 1: init/small/medium/large/search: 774681 862414 1638581 10464387 820494
3

sieve: init/small/medium/large/search: 2204269 28042425 2253189 95940634 3781423

What happens, please advice

Batalov · 2010-03-12, 00:21

This is the usual message when the sieiving finishes its requested job.
~~(looks like you were sieiving for the c171 for aliq.4788?) (probably not)~~

Did you sieve in one single run -f 41000000 -c 2000000 (judging by the number of special q's)?
Not that there's anything wrong with it, but it is more convenient to sieve in chunks; easier to parallelize etc (you will also get slightly more relations, because FB_lim will be rising together with the -c value and not be frozen at 41000000).

Your yield is low (572566 from a 2M q-range?); parameters that you are using are a bit suspect, as some people may say.
What is the number?

siew · 2010-03-12, 09:08

Hi, tanx, number is C155

i running many jobs for each core at each PC.

yes, the yield is low, but i searched for a good poly about a week and didnt find better.
Poly is:

Code:

m: 
skew: 404308.59
# norm 1.71e+021
c5: 69977040                            
c4: -40919910306512                     
c3: -38951048564673876494               
c2: 6659089867695790241111157           
c1: 2469483236376026733832761212030     
c0: -84022074362447707231091322472985569
# alpha -6.13
Y1: 1396353296355407797
Y0: -163635006307491521022993089052
type: gnfs
rlim: 18000000
alim: 18000000
lpbr: 28
lpba: 28
mfbr: 52
mfba: 52
rlambda: 2.6
alambda: 2.6

average speed is 0.54 sec/rel at Clock 3.2 Ghz of Core2Quad.

siew · 2010-03-16, 20:57

Collectet 23000 unique relations on 9-95M range.
No success,

'filtering wants more 1 000 000 relations...'

any idea to improve yield ?

maybe i have to chose different poly ?

tanx

Batalov · 2010-03-16, 21:05

Is it 23 million or 23 thousand?
If 23 million, then are they really unique (if they were unique, it should have converged)?
What is the filtering output, in detail?

P.S. If you think that your N is so secret, you should have obscured some digits in your polynomials.
Because your number is their common root, it is easily computed. It is http://factordb.com/search.php?id=16325400 and it is a C154 (but it does have 512 bits).

FactorEyes · 2010-03-16, 21:21

Quote:

Originally Posted by Batalov

P.S. If you think that your N is so secret, you should have obscured some digits in your polynomials.

Shhhhh!

-----BEGIN PGP MESSAGE-----
Version: GnuPG v1.4.9 (MingW32)

hQQOA0FupPU0EeiyEA/+OQXrC+qRZrzqEtN88S/eJYDB6ujx/38PYGW9DALP1xVK
LUemjpC9aSslVFDzWwzag0NYLgSiZsPlbPDAhVBKTf0VhO+6+F1YnxvXtsLkp3aE
sf0BMHogdpBF5z0Y1WzNlNCeya7DHPlM3muPe1xXPs63rq9prrkH8uKrALgJjUuK
jnWlqkVboukg7ciuvEnXhal/D4PQypiziAr53kmjHQJeVDkSnwJgMhcSqHCvDdmA
Klro4gtzDFc6ALy+Bfns2LBpBvcY7roiFjX7E18rHQZiONrXNbRl3TAP5oBf/3Nl
vhjtwe/a6jW4p+9uRvmNOfpgJhjdweHVGBigKBXH
=g/BY
-----END PGP MESSAGE-----

miklin · 2010-03-16, 23:15

Quote:

Originally Posted by Batalov

P.S. If you think that your N is so secret, you should have obscured some digits in your polynomials.
Because your number is their common root, it is easily computed. It is http://factordb.com/search.php?id=16325400 and it is a C154 (but it does have 512 bits).

Code:

Sun Jul 27 16:46:12 2008  Msieve v. 1.36
Sun Jul 27 16:46:12 2008  random seeds: 4f9eb0f9 c69b5c7a
Sun Jul 27 16:46:12 2008  factoring 8209836131236623540106248032442393614982637631141339598467747244677911494877848116936777629774248859738700738214196985237506543807286694994580847683477459 (154 digits)
Sun Jul 27 16:46:13 2008  searching for 15-digit factors
Sun Jul 27 16:46:14 2008  commencing number field sieve (154-digit input)
Sun Jul 27 16:46:14 2008  R0: -366501892512856641124573959989
Sun Jul 27 16:46:14 2008  R1:  153803794949520818707
Sun Jul 27 16:46:14 2008  A0: -609153808778499403500623798053669920
Sun Jul 27 16:46:14 2008  A1:  988756559882381766374145859260
Sun Jul 27 16:46:14 2008  A2:  6199402379998072308296273
Sun Jul 27 16:46:14 2008  A3: -3390691833060500638
Sun Jul 27 16:46:14 2008  A4: -9008320667086
Sun Jul 27 16:46:14 2008  A5:  1245300
Sun Jul 27 16:46:14 2008  size score = 7.632233e-16, Murphy alpha = -5.440536, combined = 4.680043e-15
Sun Jul 27 16:46:14 2008  
Sun Jul 27 16:46:14 2008  commencing square root phase
Sun Jul 27 16:46:14 2008  reading relations for dependency 1
Sun Jul 27 16:46:15 2008  read 2839528 cycles
Sun Jul 27 16:46:30 2008  cycles contain 9501724 unique relations
Sun Jul 27 16:48:43 2008  read 9501724 relations
Sun Jul 27 16:50:12 2008  multiplying 14368592 relations
Sun Jul 27 18:14:24 2008  multiply complete, coefficients have about 817.17 million bits
Sun Jul 27 18:15:03 2008  initial square root is modulo 21440369
Sun Jul 27 20:34:15 2008  reading relations for dependency 2
Sun Jul 27 20:34:17 2008  read 2841381 cycles
Sun Jul 27 20:34:28 2008  cycles contain 9504728 unique relations
Sun Jul 27 20:36:30 2008  read 9504728 relations
Sun Jul 27 20:37:59 2008  multiplying 14375930 relations
Sun Jul 27 22:03:01 2008  multiply complete, coefficients have about 817.59 million bits
Sun Jul 27 22:03:40 2008  initial square root is modulo 21628877
Mon Jul 28 00:25:30 2008  reading relations for dependency 3
Mon Jul 28 00:25:39 2008  read 2841027 cycles
Mon Jul 28 00:25:50 2008  cycles contain 9501574 unique relations
Mon Jul 28 00:28:08 2008  read 9501574 relations
Mon Jul 28 00:29:36 2008  multiplying 14367656 relations
Mon Jul 28 01:53:46 2008  multiply complete, coefficients have about 817.13 million bits
Mon Jul 28 01:54:25 2008  initial square root is modulo 21422053
Mon Jul 28 04:13:39 2008  prp77 factor: 7525920029946****************************************************************
Mon Jul 28 04:13:39 2008  prp78 factor: 1090874750006*****************************************************************
Mon Jul 28 04:13:39 2008  elapsed time 11:27:27

miklin · 2010-03-17, 00:00

Quote:

Originally Posted by siew

Hi, tanx, number is C155

i running many jobs for each core at each PC.

yes, the yield is low, but i searched for a good poly about a week and didnt find better.

BEGIN POLY #skewness 450875.80 norm 2.35e+021 alpha -6.10 Murphy_E 2.61e-012
BEGIN POLY #skewness 458270.11 norm 3.92e+021 alpha -6.43 Murphy_E 2.54e-012
BEGIN POLY #skewness 508663.06 norm 4.05e+021 alpha -6.48 Murphy_E 2.58e-012
BEGIN POLY #skewness 508947.15 norm 5.71e+021 alpha -6.39 Murphy_E 2.57e-012
BEGIN POLY #skewness 522987.56 norm 8.65e+021 alpha -6.84 Murphy_E 2.57e-012
BEGIN POLY #skewness 557204.51 norm 2.17e+021 alpha -6.00 Murphy_E 2.67e-012
BEGIN POLY #skewness 594103.33 norm 4.21e+021 alpha -6.49 Murphy_E 2.55e-012
BEGIN POLY #skewness 594322.68 norm 7.63e+020 alpha -4.93 Murphy_E 2.54e-012
BEGIN POLY #skewness 605334.77 norm 2.04e+022 alpha -7.85 Murphy_E 2.55e-012
BEGIN POLY #skewness 690937.77 norm 7.44e+020 alpha -5.00 Murphy_E 2.52e-012

P.S The Post ( http://mersenneforum.org/showpost.ph...&postcount=405 ) above, too for you.

FactorEyes · 2010-03-17, 00:21

[QUOTE=siew;208121]
Poly is:

Code:

rlim: 18000000
alim: 18000000
lpbr: 28
lpba: 28
mfbr: 52
mfba: 52

You should play around with all six of these. Hint: larger.

You will probably be able to pool your existing relations with any new ones you find, so don't throw out what you have so far.

2010-03-12, 00:21	#400
Batalov "Serge" Mar 2008 Phi(4,2^7658614+1)/2 10010100010000₂ Posts	This is the usual message when the sieiving finishes its requested job. ~~(looks like you were sieiving for the c171 for aliq.4788?) (probably not)~~ Did you sieve in one single run -f 41000000 -c 2000000 (judging by the number of special q's)? Not that there's anything wrong with it, but it is more convenient to sieve in chunks; easier to parallelize etc (you will also get slightly more relations, because FB_lim will be rising together with the -c value and not be frozen at 41000000). Your yield is low (572566 from a 2M q-range?); parameters that you are using are a bit suspect, as some people may say. What is the number? Last fiddled with by Batalov on 2010-03-12 at 00:28

2010-03-16, 21:05	#403
Batalov "Serge" Mar 2008 Phi(4,2^7658614+1)/2 2510₁₆ Posts	Is it 23 million or 23 thousand? If 23 million, then are they really unique (if they were unique, it should have converged)? What is the filtering output, in detail? P.S. If you think that your N is so secret, you should have obscured some digits in your polynomials. Because your number is their common root, it is easily computed. It is http://factordb.com/search.php?id=16325400 and it is a C154 (but it does have 512 bits). Last fiddled with by Batalov on 2010-03-16 at 21:19

2010-03-17, 00:21	#407
FactorEyes Oct 2006 vomit_frame_pointer 2³·3²·5 Posts	[QUOTE=siew;208121] Poly is: Code: rlim: 18000000 alim: 18000000 lpbr: 28 lpba: 28 mfbr: 52 mfba: 52 You should play around with all six of these. Hint: larger. You will probably be able to pool your existing relations with any new ones you find, so don't throw out what you have so far.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Installation of GGNFS	LegionMammal978	Msieve	17	2017-01-20 19:49
Running other programs while running Prime95.	Neimanator	PrimeNet	14	2013-08-10 20:15
Error running GGNFS+msieve+factmsieve.py	D. B. Staple	Factoring	6	2011-06-12 22:23
GGNFS or something better?	Zeta-Flux	Factoring	1	2007-08-07 22:40
ggnfs	ATH	Factoring	3	2006-08-12 22:50

2010-02-26, 14:32	#398
fivemack (loop (#_fork)) Feb 2006 Cambridge, England 2·13²·19 Posts	Not quite sure what happened there, I think I must have edited apocalypse's post rather than replying to it.

2010-03-16, 20:57	#402
siew Oct 2009 3²×7 Posts	Collectet 23000 unique relations on 9-95M range. No success, 'filtering wants more 1 000 000 relations...' any idea to improve yield ? maybe i have to chose different poly ? tanx