Pascal's OPN roadblock files - Page 69

lavalamp · 2019-12-20, 06:38

A rather surprising result for this number in the t800 file:

(11678983384265232966613390730340859^7-1) / (p-1) = 4201 * p45 * p50 * p108

Presumably some unlucky ECM that didn't catch the p45 at least?

Edit: I believe I have now finished off all numbers of the form p^7-1 in the t800 file where p has <=35 digits.

RichD · 2020-01-09, 12:12

All the C140s are factored. I am starting with the C141s.

RichD · 2020-01-24, 01:31

Quote:

Originally Posted by lavalamp

(11678983384265232966613390730340859^7-1) / (p-1) = 4201 * p45 * p50 * p108

Presumably some unlucky ECM that didn't catch the p45 at least?

I can see where this can be argued either way. Using 0.21 * SNFS-205 needs a t43 testing level. Or using 2/9 of 205 would need a t45.5.

RichD · 2020-01-24, 01:35

A few smallish numbers have appeared in the t2100 file in the mid-C110 range. I have been working the C141 range so no conflict by me.

MisterBitcoin · 2020-01-25, 21:06

Found an P31 for this canidate.

Code:

 (8607076653121^47-1)/1409491047619715454786751986299893440

Composite Cofactor C542 survived 430 curves at B1=250K.

Number was found in the t2100 file.

RichD · 2020-06-16, 00:15

Different perspectives on the above number in this post and this post.

This is a Most Wanted number with a calculated OPN weight of 2.5Mil. Actually number 11 from the top in the MWRB file. yoyo has completed 10K curves at t65 and 9K curves at t70.

VBCurtis · 2020-06-16, 01:59

I'm game to do another hybrid CADO/15e-queue job with SNFS; we've done GNFS197 and GNFS198 somewhat profitably, and trying an iffy SNFS job seems fine.
I won't have human-time to test params nor yields with CADO until the end of June; if any of the usual suspects wishes to do some yield calcs, please be my guest.
I'd say Q-max of 800M in 15e is a reasonable target. So, what Q should be done in CADO (and should they be done with A=30 or I=16) such that doing the rest of the Q up to 800M yields enough relations?

I'm looking for an answer like "10-150 on CADO A=30, 150-800 on 15e" or "20-100 on I=16, 100-700 on ggnfs". Anyone willing to do such test sieving in a way that tries to minimize total sieve time will find a willing CADO host come 1 july.
I haven't done an SNFS job above 900 bits yet, so I'd also like the param choices to be spelled out so I set up CADO correctly: sieve which side, 3LP on which side, lims. I assume 33LP since we're stretching 15e.

henryzz · 2020-06-16, 08:16

ryanp gave in trying to factor 732541^47-1 due to it running into weird issues with lasieve(random low yield on some q if I remember right). I would suggest doing this one purely CADO would be the most sensible way. I think this one would have been done ages ago if it wasn't for this issue.

RichD · 2020-06-16, 13:23

At one point in time ryanp had an interest in working a few OPN. I suppose Real Life^TM issues came up. This was my last correspondence to him and possibly a starting point for moving forward.

Quote:

Finally got time to do some testing. Last month I had a power supply go out after a power outage. That made me behind (over-committed) on projects/assignments. Then I had to shut down for a few days while I relocated. Now I have test sieved on the Linux box with 16e and everything looks good.

I was going to wait until I had the full range tested but I believe I proved we can get enough relations (on or before Q=600M) on the -r side.

Code:

Trial sieving 2K blocks.
  Q     Yield
 20M  6299 3.15
100M  6080 3.04
200M  5612 2.81

n: 605716904027877980774625455520189647387776352555063757365644672493136637525085152114527251672682055452329862008130550673203343550128250999766605061023948523297828457779191592093682881010498969046911261346842026672855745883554109771998292748069377018429964450347583969787
# 732541^47-1, difficulty: 281.51, skewness: 9.49, alpha: 0.00
# cost: 8.45884e+19, est. time: 40280.21 GHz days (not accurate yet!)
skew: 9.494
c6: 1
c5: 0
c4: 0
c3: 0
c2: 0
c1: 0
c0: -732541
Y1: -1
Y0: 82919274927962023982932249248351337261442889121
m: 82919274927962023982932249248351337261442889121
type: snfs
rlim: 536000000
alim: 1072000000
lpbr: 33
lpba: 33
mfbr: 66
mfba: 96
rlambda: 2.7
alambda: 3.6

RichD · 2020-06-16, 16:08

Quote:

Originally Posted by henryzz

ryanp gave in trying to factor 732541^47-1 due to it running into weird issues with lasieve(random low yield on some q if I remember right). I would suggest doing this one purely CADO would be the most sensible way. I think this one would have been done ages ago if it wasn't for this issue.

Now I remember. We/He tried to the run the number 1 Most Wanted Road Block (even as an octic) but ran into the stated problems you mentioned. I think we decided it needed to be run as a 34-bit job but neither one of us could find the appropriate binary at that time. Hence, it was abandon.

Edit: BTW, 6115909044841454629^17-1, SNFS-301 (octic).

chris2be8 · 2020-06-17, 16:05

Quote:

Originally Posted by RichD

Different perspectives on the above number in this post and this post.

This is a Most Wanted number with a calculated OPN weight of 2.5Mil. Actually number 11 from the top in the MWRB file. yoyo has completed 10K curves at t65 and 9K curves at t70.

I think fivemack was just saying it's too big for 15e on NFS@Home, so would need to be done with 16e. It would be doable though.

Chris

2019-12-20, 06:38	#749
lavalamp Oct 2007 Manchester, UK 2×3×223 Posts	A rather surprising result for this number in the t800 file: (11678983384265232966613390730340859^7-1) / (p-1) = 4201 * p45 * p50 * p108 Presumably some unlucky ECM that didn't catch the p45 at least? Edit: I believe I have now finished off all numbers of the form p^7-1 in the t800 file where p has <=35 digits. Last fiddled with by lavalamp on 2019-12-20 at 06:42

2020-01-09, 12:12	#750
RichD Sep 2008 Kansas 7·463 Posts	t2100.txt All the C140s are factored. I am starting with the C141s.

2020-01-25, 21:06	#753
MisterBitcoin "Nuri, the dragon :P" Jul 2016 Good old Germany 2×17×23 Posts	Found an P31 for this canidate. Code: (8607076653121^47-1)/1409491047619715454786751986299893440 Composite Cofactor C542 survived 430 curves at B1=250K. Number was found in the t2100 file.

2020-06-16, 00:15	#754
RichD Sep 2008 Kansas 6251₈ Posts	732541^47-1 Different perspectives on the above number in this post and this post. This is a Most Wanted number with a calculated OPN weight of 2.5Mil. Actually number 11 from the top in the MWRB file. yoyo has completed 10K curves at t65 and 9K curves at t70.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Passive Pascal	Xyzzy	GPU Computing	1	2017-05-17 20:22
Tesla P100 — 5.4 DP TeraFLOPS — Pascal	Mark Rose	GPU Computing	52	2016-07-02 12:11
Nvidia Pascal, a third of DP	firejuggler	GPU Computing	12	2016-02-23 06:55
Calculating perfect numbers in Pascal	Elhueno	Homework Help	5	2008-06-12 16:37
Factorization attempt to a c163 - a new Odd Perfect Number roadblock	jchein1	Factoring	30	2005-05-30 14:43

2020-01-24, 01:35	#752
RichD Sep 2008 Kansas 7×463 Posts	A few smallish numbers have appeared in the t2100 file in the mid-C110 range. I have been working the C141 range so no conflict by me.

2020-06-16, 01:59	#755
VBCurtis "Curtis" Feb 2005 Riverside, CA 7·659 Posts	I'm game to do another hybrid CADO/15e-queue job with SNFS; we've done GNFS197 and GNFS198 somewhat profitably, and trying an iffy SNFS job seems fine. I won't have human-time to test params nor yields with CADO until the end of June; if any of the usual suspects wishes to do some yield calcs, please be my guest. I'd say Q-max of 800M in 15e is a reasonable target. So, what Q should be done in CADO (and should they be done with A=30 or I=16) such that doing the rest of the Q up to 800M yields enough relations? I'm looking for an answer like "10-150 on CADO A=30, 150-800 on 15e" or "20-100 on I=16, 100-700 on ggnfs". Anyone willing to do such test sieving in a way that tries to minimize total sieve time will find a willing CADO host come 1 july. I haven't done an SNFS job above 900 bits yet, so I'd also like the param choices to be spelled out so I set up CADO correctly: sieve which side, 3LP on which side, lims. I assume 33LP since we're stretching 15e.

2020-06-16, 08:16	#756
henryzz Just call me Henry "David" Sep 2007 Cambridge (GMT/BST) 2·2,897 Posts	ryanp gave in trying to factor 732541^47-1 due to it running into weird issues with lasieve(random low yield on some q if I remember right). I would suggest doing this one purely CADO would be the most sensible way. I think this one would have been done ages ago if it wasn't for this issue.