Fast Breeding (guru management) - Page 92

swellman · 2017-04-14, 19:13

Another composite ready for 14e. C209_125_122 has survived a full t55 by yoyo@Home and a few thousand more curves @B1=3e8 by me, with no factors found.

Code:

n: 27365418530407784409149457734678145942421565154918349069416779248441859303354412278473612336441552673499736719637468141998274557145007503297364086365784811478821747785405259009052821315703944637780030717006553
# 125^122+122^125, difficulty: 262.90, anorm: 2.21e+037, rnorm: 3.71e+049
# scaled difficulty: 264.93, suggest sieving rational side
# size = 6.983e-013, alpha = 0.000, combined = 8.808e-014, rroots = 0
type: snfs
size: 262
skew: 2.2270
c6: 1
c0: 122
Y1: -4336808689942017736029811203479766845703125
Y0: 65096320940612888928992369305913995581980672
rlim: 200000000
alim: 200000000
lpbr: 32
lpba: 32
mfbr: 64
mfba: 64
rlambda: 2.8
alambda: 2.8

Some test sieving in blocks of 5000 Q:

Code:

Q=20M total yield: 9032, q=20005003 (0.46331 sec/rel)
Q=100M total yield: 7807, q=100005001 (0.59778 sec/rel)
Q=200M total yield: 6545, q=200005007 (0.82368 sec/rel)

VBCurtis · 2017-04-14, 22:44

Quote:

Originally Posted by swellman

Some test sieving in blocks of 5000 Q:

Code:

Q=20M total yield: 9032, q=20005003 (0.46331 sec/rel)
Q=100M total yield: 7807, q=100005001 (0.59778 sec/rel)
Q=200M total yield: 6545, q=200005007 (0.82368 sec/rel)

I tested your settings (though I goofed and set lambdas to 2.7 rather than 2.8, so not quite precisely the same; then I tested mfbr = 93 and rlambda = 3.7 to compare 3 large primes:

Code:

2R2A: Q=20M total yield: 9024, q=20005003 (0.14645 sec/rel)
3R2A: Q=20M total yield: 14978, q=20005003 (0.11966 sec/rel)

50% better yield, nice improvement in sec/rel. Sorry, didn't have time to test higher Q values to make sure the sec/rel improvement is across the board, but the large yield improvement means fewer of the big Q to sieve anyway.

swellman · 2017-04-14, 22:57

@VBCurtis thank you so very much. I didn't think to check for 3LP with what seemed (to me at least) a somewhat generic poly. Was there something that made you investigate 3LPs or was it just good practice?

Either way you've made this composite much easier to sieve.

Not to be a pest, but can one of the gatekeepers feed 14e? Queue is almost empty.

VBCurtis · 2017-04-15, 06:29

Quote:

Originally Posted by swellman

@VBCurtis thank you so very much. I didn't think to check for 3LP with what seemed (to me at least) a somewhat generic poly. Was there something that made you investigate 3LPs or was it just good practice?

Two things: SNFS-260 seemed way too big for 14e, and yield even with 32LP was below 2. I was worried this was going to be much easier on 15e, but we need 14e food; so I thought I'd try 3LP or 33LP to see if either one made the stretch into the 260s palatable for 14e. It has never before occurred to me to try 3LP on a 14e task, so I can't claim "good practice".
I did not try 33LP, but I would not be surprised to find a case in this size range where 14e/33 was likely to be quicker than 14e/32 (though I imagine such a case would be yet easier with 15e/31 or 15e/32). There's good fun in "how far can we stretch the 14e siever?".

fivemack · 2017-04-15, 12:40

C209_125_122, C209_127_91, C185_86353_47 fed to the queue.

Thanks everyone for the more usable format of presenting trial sieving results; I have done the integrations in Excel and picked parameters I-hope-appropriately.

I am a much happier gatekeeper once I realise I should use Firefox rather than Safari for fettling the queue page from my Mac at home - Safari really slows down on a page with ten thousand editable text boxes.

swellman · 2017-04-15, 12:50

There are two more composites available to feed 14e - see posts 1000-1001 of this thread. Sorry - I see you've found them.

Good to know about Safari vs Firefox performance. Of course, if you had a Windows box it wouldn't be an issue [ducks, runs away in a serpentine manner...];-)

RichD · 2017-04-15, 20:04

Another OPN from the t600 file.

Code:

n: 352118671977091369297589999783927719255273432509208039367661933290917550037863164371471854669554332817359372845763627978778660919700801877630593402798278084179981920891755570883480907
# 705670360649^19-1, difficulty 225.12
# sieve on algebraic side
lss: 0
skew: 0.0106
c6: 705670360649
c0: -1
Y1: -1
Y0: 351403133751847046579328315380439449
rlim: 134000000
alim: 134000000
lpbr: 31
lpba: 31
mfbr: 62
mfba: 62
rlambda: 2.6
alambda: 2.6

Test sieving 5K blocks.

Code:

  Q  Yield
 20M 10120
 60M  7578
100M  7988

RichD · 2017-04-15, 21:53

Add the following test sieve results to the above.

Code:

140M  7874

richs · 2017-04-19, 15:23

Does anyone have any advice on how to resume a download of the relations datafile? I was about a day and a half into downloading C237_12161_59 (10.5 Gb out of 12.4 Gb) when the download stopped due to a network error. I am using Chrome on Windows 10.

richs · 2017-04-19, 21:03

Quote:

Originally Posted by richs

Does anyone have any advice on how to resume a download of the relations datafile? I was about a day and a half into downloading C237_12161_59 (10.5 Gb out of 12.4 Gb) when the download stopped due to a network error. I am using Chrome on Windows 10.

I figured it out. I used wget64 with the following command line:

Code:

wget64.exe --continue --no-check-certificate --user ******** --password ******** http://escatter11.fullerton.edu/********/C237_12161_59/C237_12161_59.dat.gz

unconnected · 2017-04-20, 09:53

127^121+121^127 is already factored, please remove it from 15e queue

2017-04-14, 22:57	#1004
swellman Jun 2012 11×281 Posts	@VBCurtis thank you so very much. I didn't think to check for 3LP with what seemed (to me at least) a somewhat generic poly. Was there something that made you investigate 3LPs or was it just good practice? Either way you've made this composite much easier to sieve. Not to be a pest, but can one of the gatekeepers feed 14e? Queue is almost empty. Last fiddled with by swellman on 2017-04-14 at 22:58

2017-04-15, 12:40	#1006
fivemack (loop (#_fork)) Feb 2006 Cambridge, England 2³×11×73 Posts	C209_125_122, C209_127_91, C185_86353_47 fed to the queue. Thanks everyone for the more usable format of presenting trial sieving results; I have done the integrations in Excel and picked parameters I-hope-appropriately. I am a much happier gatekeeper once I realise I should use Firefox rather than Safari for fettling the queue page from my Mac at home - Safari really slows down on a page with ten thousand editable text boxes. Last fiddled with by fivemack on 2017-04-15 at 12:45

2017-04-15, 12:50	#1007
swellman Jun 2012 11·281 Posts	There are two more composites available to feed 14e - see posts 1000-1001 of this thread. Sorry - I see you've found them. Good to know about Safari vs Firefox performance. Of course, if you had a Windows box it wouldn't be an issue [ducks, runs away in a serpentine manner...];-) Last fiddled with by swellman on 2017-04-15 at 12:51 Reason: Oops

2017-04-15, 21:53	#1009
RichD Sep 2008 Kansas 2×3×5×113 Posts	705670360649^19-1 Add the following test sieve results to the above. Code: 140M 7874

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
System management notes	kriesel	kriesel	7	2020-10-21 18:52
Improving the queue management.	debrouxl	NFS@Home	10	2018-05-06 21:05
Script-based Primenet assignment management	ewmayer	Software	3	2017-05-25 04:02
Do normal adults give themselves an allowance? (...to fast or not to fast - there is no question!)	jasong	jasong	35	2016-12-11 00:57
Power Management settings	PrimeCroat	Hardware	3	2004-02-17 19:11

2017-04-19, 15:23	#1010
richs "Rich" Aug 2002 Benicia, California 2³×3×5×11 Posts	Does anyone have any advice on how to resume a download of the relations datafile? I was about a day and a half into downloading C237_12161_59 (10.5 Gb out of 12.4 Gb) when the download stopped due to a network error. I am using Chrome on Windows 10.

2017-04-20, 09:53	#1012
unconnected May 2009 Russia, Moscow 2,593 Posts	127^121+121^127 is already factored, please remove it from 15e queue