Fast Breeding (guru management) - Page 76

RichD · 2016-12-21, 04:57

Quote:

Originally Posted by jyb

So now I'm confused. You got better results with quintics, but I've gotten much better results with sextics. What gives?.

Pull them both off until we can get this deciphered. I've been juggling several number and maybe I am the one confused.

swellman · 2016-12-21, 14:53

Here's one for 14e which has survived a full t55 plus a bit of B1=3e8. Full disclosure - it would perhaps be better to run it through 15e or as a 14e/32 job but should work as is. The 14e queue is running dry again.

I will handle the postprocessing.

C253_131_91

Code:

n: 2537063132129631115698746903522943935147375191654948413855185174155666479187812508096108428840058843829591587140039584142597315983406890888385057018512151614924726522738227137794066910268576608479110340832714298508122124050830173398972241174957675222027
# 131^91+91^131, difficulty: 258.59, anorm: 2.18e+038, rnorm: 5.74e+048
# scaled difficulty: 260.33, suggest sieving rational side
# size = 8.090e-013, alpha = 0.000, combined = 1.009e-013, rroots = 0
type: snfs
size: 258
skew: 4.7795
c6: 1
c0: 11921
Y1: -57420825906681085498936641961451
Y0: 12557715249685059365095784756336096391283081
rlim: 134000000
alim: 134000000
lpbr: 31
lpba: 31
mfbr: 62
mfba: 62
rlambda: 2.7
alambda: 2.7

RichD · 2016-12-21, 16:08

Quote:

Originally Posted by jyb

So now I'm confused. You got better results with quintics, but I've gotten much better results with sextics. What gives?

I can confirm 208097431^29-1 sieves better as a septic using 14e. I will look into 790579404481^19-1 later today. I have a sneaky feeling you are right about both.

One fallacy in my testing, I let factMsieve pick all the parameters and didn't notice it was using 15e. Another, I was only looking at the times over a small region. Much too small for any worthwhile analysis.

jyb · 2016-12-21, 16:46

Quote:

Originally Posted by RichD

I can confirm 208097431^29-1 sieves better as a septic using 14e.

Uh, sextic, right?

RichD · 2016-12-21, 16:53

Quote:

Originally Posted by jyb

Uh, sextic, right?

Damn auto-correct.

RichD · 2016-12-21, 16:58

Go with what you found with 790579404481^19-1. I am going to do some test sieving for my own edification.

swellman · 2016-12-21, 20:55

C198_148_78 is another fast sieving number that I ran ECM against for ~5K curves @B1=11e7 and another 1k curves @3e8. Not a lot of ECM by the 2/9 rule but it's another case of total time of ECM > 1/3 * time to sieve. Plus the ECM @3e8 gives me some hope there are no p51s waiting to strike...

Code:

n: 153157072086335672906087000410932457349679841465250341042913974383511710451602287883874665982231725598170508541847346158353357648318019977631504530641541678177013474805423559195823103189591132589669
# 148^78+78^148, difficulty: 239.61, anorm: 4.68e+038, rnorm: 1.67e+045
# scaled difficulty: 240.71, suggest sieving rational side
# size = 4.299e-012, alpha = 0.000, combined = 3.562e-013, rroots = 0
type: snfs
size: 239
skew: 1.4242
c6: 81
c0: 676
Y1: -487138448432162610794
Y0: 1992938609718721349935067589591558240333
rlim: 134000000
alim: 134000000
lpbr: 31
lpba: 31
mfbr: 62
mfba: 62
rlambda: 2.7
alambda: 2.7

YuL · 2016-12-22, 09:52

Quote:

Originally Posted by jyb

I'm not aware that anybody has explicitly decided to abandon it, I just think that people forgot about it.

There have been other GNFS jobs with numbers around this size with 15e, so it should be fine, though of course the 15e queue isn't hurting for candidates at the moment. Still, if you can do some test sieving to determine good parameters for the job, it can be queued.

OK, there you go, parameters are joined, sieving to be done using 15e from 50e6 to 210e6
on A side (I target around 380e6 relations). Note that as I wrote previously credit for the
polynomial goes to wombatman.
If the number gets queued you can put my name in the post-processing column.
Thanks.

RichD · 2016-12-23, 15:39

I started with degree 5 and sent it to factMsieve which in turn picked 15e and lpb=30/60 among other parameters. So I changed to 14e and performed some trial sieving. After the first run I noticed lpb=30/60 so I changed it to 31/62 and continued. Below are my yield numbers for 5K blocks.

Code:

   special-Q =>    30M     80M
deg-5 -r 30/60    1394    1052
deg-5 -a 30/60     608     579

deg-6 -r 31/62     531     275
deg-6 -a 31/62     503     408

deg-5 -r 31/62    2788    2014
deg-5 -a 31/62    1540    1118

Now I see why I picked a quintic.

jyb · 2016-12-23, 17:41

Quote:

Originally Posted by RichD

I started with degree 5 and sent it to factMsieve which in turn picked 15e and lpb=30/60 among other parameters. So I changed to 14e and performed some trial sieving. After the first run I noticed lpb=30/60 so I changed it to 31/62 and continued. Below are my yield numbers for 5K blocks.

Code:

   special-Q =>    30M     80M
deg-5 -r 30/60    1394    1052
deg-5 -a 30/60     608     579

deg-6 -r 31/62     531     275
deg-6 -a 31/62     503     408

deg-5 -r 31/62    2788    2014
deg-5 -a 31/62    1540    1118

Now I see why I picked a quintic.

With those numbers I see why too. But the thing is, I got numbers which were very much at odds with yours. I only used 1K blocks, but the difference between my numbers and yours is still striking:

Code:

   special-Q =>    20M     100M
deg-6 -r 31/62     1508     763
deg-6 -a 31/62     1766    1257

deg-5 -r 31/62     682     492
deg-5 -a 31/62     649     258

BTW, these are all with an alim and rlim of 67000000.

As you can see, the degree-5 numbers are similar to yours (after accounting for the smaller block of special-q and what I'm guessing is a larger fb size), and the degree-6 numbers are substantially better. So why are your degree-6 yields so terrible? Tell me, what is your degree-6 polynomial?

RichD · 2016-12-23, 17:57

Quote:

Originally Posted by jyb

So why are your degree-6 yields so terrible? Tell me, what is your degree-6 polynomial?

These are the default parameters out of factMsieve and then changing to 14e and lpb=31/62. Most likely not optimal.

Code:

n: 620595445212886197477191763266087920164778283551144144831883282223415660307863206244244816417687295355308926397314396118416074156134623096759384176319664870366271916273988877636289052746091391
m: 494124614835927587916699085265816641
c6: 790579404481
c5: 0
c4: 0
c3: 0
c2: 0
c1: 0
c0: -1
Y1: -1
Y0: 494124614835927587916699085265816641
skew: 96.16
rlim: 25450000
alim: 49700000
lpbr: 31
lpba: 31
mfbr: 62
mfba: 62
rlambda: 2.6
alambda: 2.6

P.S Off the top of my head, before trial sieving, I would have guessed deg-6 would have been better because you are not elevating the difficulty by the large base.

2016-12-23, 15:39	#834
RichD Sep 2008 Kansas 3×1,129 Posts	My test sieving for 790579404481^19-1 I started with degree 5 and sent it to factMsieve which in turn picked 15e and lpb=30/60 among other parameters. So I changed to 14e and performed some trial sieving. After the first run I noticed lpb=30/60 so I changed it to 31/62 and continued. Below are my yield numbers for 5K blocks. Code: special-Q => 30M 80M deg-5 -r 30/60 1394 1052 deg-5 -a 30/60 608 579 deg-6 -r 31/62 531 275 deg-6 -a 31/62 503 408 deg-5 -r 31/62 2788 2014 deg-5 -a 31/62 1540 1118 Now I see why I picked a quintic.

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

2016-12-21, 16:58	#831
RichD Sep 2008 Kansas 3×1,129 Posts	Go with what you found with 790579404481^19-1. I am going to do some test sieving for my own edification.