Linear algebra reservations, progress and results - Page 128

jyb · 2016-09-19, 16:51

Reserving 8+5_269 (14e).

swellman · 2016-09-19, 21:45

Quote:

Originally Posted by swellman

Been working on C269_140_83 for awhile. Finally got it through filtering but now it's almost been a week just building the matrix. I think I'm out on the fringe with my system (16Gb memory) but trying to push through. I'll keep going until the end of the month - if it isn't built by then it will never happen. This effort has taken 5x longer than any previous 32-bit job I've attempted.

Can hardly wait to see the ETA if it makes it that far!

The matrix is finally built but it's all a lost cause. ETA is decades in length, and progress is glacial. I don't have enough memory or horsepower.

Please release C269_140_83. I hope someone else can postprocess it!

Reserving the cofactor of 150^122+122^150

unconnected · 2016-09-19, 21:56

Quote:

Originally Posted by swellman

Please release C269_140_83. I hope someone else can postprocess it!

Reserving the cofactor of 150^122+122^150

150^122+122^150 SNFS(277.63) seems even tougher than 140^83+83^140 SNFS(270.82)

swellman · 2016-09-19, 22:59

Quote:

Originally Posted by unconnected

150^122+122^150 SNFS(277.63) seems even tougher than 140^83+83^140 SNFS(270.82)

Indeed but I've done a SNFS 271 in the past with no problems. Not sure why 140_83 was so difficult. A 30M x 30M matrix had its problems, but I think my eyes were too big for my stomach. I can post the log file if it will help.

I'm willing to attempt 150_122. Should know quickly if I can post process it (once it has enough relations).

jyb · 2016-09-20, 00:25

Reserving C212_138_53 (14e).

jyb · 2016-09-20, 00:26

Quote:

Originally Posted by swellman

Indeed but I've done a SNFS 271 in the past with no problems. Not sure why 140_83 was so difficult. A 30M x 30M matrix had its problems, but I think my eyes were too big for my stomach. I can post the log file if it will help.

I'm willing to attempt 150_122. Should know quickly if I can post process it (once it has enough relations).

Do you think it would help to do more sieving on C269_140_83? Say another 50M relations?

swellman · 2016-09-20, 00:56

Quote:

Originally Posted by jyb

Do you think it would help to do more sieving on C269_140_83? Say another 50M relations?

I'm not sure but usually I can read 32-bit jobs into memory and postprocess with only a few complaints from my system. This composite drank memory like a sponge and my system just couldn't run LA after the matrix was finally built. I think it's just too big for me but I'm willing to try again if more relations are added, maybe a higher TD would help (I used TD=130)?

jyb · 2016-09-20, 04:14

Code:

p83 factor: 89924363526609478441451550341199291672323172034030431514992942534863308314735106429
p128 factor: 82166813913486588479110582880897205365471804935373013724621832979805599282412789990531853408372393288384208484049790807103334207

Log: http://pastebin.com/4YTJAwp0

jyb · 2016-09-20, 04:40

Reserving C208_148_104 (14e).

fivemack · 2016-09-20, 07:59

I have put a further 100MQ onto C269_140_83 and will attack it with the biggest hammers I have. Reserving.

If I'm going to queue up jobs which are slightly too difficult, it should at least be my hammers that I break on them.

unconnected · 2016-09-20, 08:05

C275_150_116 (23.3M matrix @ TD=136):

Code:

Tue Sep 13 12:01:07 2016  found 126419294 hash collisions in 502133477 relations
Tue Sep 13 12:01:29 2016  added 1 free relations
Tue Sep 13 12:01:29 2016  commencing duplicate removal, pass 2
Tue Sep 13 13:35:39 2016  found 126374785 duplicates and 375758693 unique relations

C260_140_81 (20M matrix @ TD=130):

Code:

Wed Jul 27 03:53:34 2016  found 118384806 hash collisions in 495023007 relations
 Wed Jul 27 03:53:53 2016  added 1 free relations
 Wed Jul 27 03:53:53 2016  commencing duplicate removal, pass 2
 Wed Jul 27 05:20:40 2016  found 113561203 duplicates and 381461805 unique relations

C269_140_83 (TD=132 fails, TD=128 in progress):

Code:

Mon Sep 19 12:18:26 2016  found 116619041 hash collisions in 452743675 relations
Mon Sep 19 12:18:49 2016  commencing duplicate removal, pass 2
Mon Sep 19 13:45:32 2016  found 121479559 duplicates and 331264116 unique relations

Slightly higher duplicate rate (26.8% vs 25.1 and 22.9) and lower total relations count (~450M vs ~500M). I think some additional sieving might help.

2016-09-20, 07:59	#1407
fivemack (loop (#_fork)) Feb 2006 Cambridge, England 1100100011000₂ Posts	I have put a further 100MQ onto C269_140_83 and will attack it with the biggest hammers I have. Reserving. If I'm going to queue up jobs which are slightly too difficult, it should at least be my hammers that I break on them. Last fiddled with by fivemack on 2016-09-20 at 08:00

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
restarting nfs linear algebra	cubaq	YAFU	2	2017-04-02 11:35
Linear algebra at 600%	CRGreathouse	Msieve	8	2009-08-05 07:25
Linear algebra crashes	10metreh	Msieve	3	2009-02-02 08:34
Linear algebra proof	Damian	Math	8	2007-02-12 22:25
Linear algebra in MPQS	R1zZ1	Factoring	2	2007-02-02 06:45

2016-09-19, 16:51	#1398
jyb Aug 2005 Seattle, WA 11011100110₂ Posts	Reserving 8+5_269 (14e).

2016-09-20, 00:25	#1402
jyb Aug 2005 Seattle, WA 2·883 Posts	Reserving C212_138_53 (14e).

2016-09-20, 04:40	#1406
jyb Aug 2005 Seattle, WA 2×883 Posts	Reserving C208_148_104 (14e).