Fast Breeding (guru management) - Page 122

swellman · 2018-01-30, 17:10

QUEUED C232_146_49 is ready for SNFS on 14e.

Code:

n: 1013101528926077978679178427704027695375716974314089721050828820271712465075225788768410285923554921116684180703005038969476384074972513279143913652849762612925082241419943363578919924527774569025441804427639587057809167118875794133
# 146^49+49^146, difficulty: 250.15, anorm: 1.69e+038, rnorm: 1.23e+047
# scaled difficulty: 251.63, suggest sieving rational side
# size = 1.687e-012, alpha = 0.000, combined = 1.774e-013, rroots = 0
type: snfs
size: 250
skew: 4.3896
c6: 1
c0: 7154
Y1: -206453783524884736
Y0: 256923577521058878088611477224235621321607
rlim: 134000000
alim: 134000000
lpbr: 31
lpba: 31
mfbr: 62
mfba: 62
rlambda: 2.7
alambda: 2.7

Test sieving on the -r side with Q in blocks of 4K:

Code:

Q=20M      6076
Q=80M      4298
Q=150M     3734
Q=250M     3128
Q=300M     2974

Suggesting a sieving range for Q of 20M-270M with target # rels = 250M.

RichD · 2018-02-02, 00:00

QUEUED C197 from the OPN t1200 file.
P55.21900_5M.C197

Code:

n: 28848967990077705846988961468539295920671331242155757837897308053389516744137239692990930018313384178496879018740961164234293984375975760512498647138632711308817604574356514611333530865932937557161
# 2190066216160043027357373834534068629814437632915752857^5-1, difficulty: 217.36, skewness: 1.00, alpha: 1.45
# cost: 6.1848e+17, est. time: 294.51 GHz days (not accurate yet!)
skew: 1.000
c4: 1
c3: 1
c2: 1
c1: 1
c0: 1
Y1: -1
Y0: 2190066216160043027357373834534068629814437632915752857
type: snfs
rlim: 134000000
alim: 134000000
lpbr: 31
lpba: 31
mfbr: 92
mfba: 62
rlambda: 3.6
alambda: 2.6

Trial sieving 5K blocks.

Code:

This will be it for a while since I will be out of town next week.

swellman · 2018-02-02, 19:36

QUEUED C233_133_86 is ready for SNFS on 14e.

Code:

n: 10633354020919847828049263705668353382285616309669021200078310598696796477207436371326306802388614964256350490347978458835867676327562757887165210981714336194048764404430848960645565061266895540038717130441903406926921630132396259373
# 133^86+86^133, difficulty: 259.22, anorm: 2.47e+039, rnorm: 2.32e+048
# scaled difficulty: 260.72, suggest sieving rational side
# size = 6.137e-013, alpha = 0.000, combined = 8.413e-014, rroots = 0
type: snfs
size: 259
skew: 2.4296
c6: 86
c0: 17689
Y1: -541904769658563069794308330729
Y0: 3622145797004077275761664476965378610692096
rlim: 268000000
alim: 268000000
lpbr: 32
lpba: 32
mfbr: 64
mfba: 64
rlambda: 2.8
alambda: 2.8

Test sieving on the -r side with Q in blocks of 5K

Code:

Q=30M     9513
Q=70M     8214
Q=140M    7044
Q=230M    6493
Q=320M    6047
Q=400M    5352

suggesting a sieving range for Q of 30M-390M with a target # relations = 490M

swellman · 2018-02-07, 11:33

QUEUED XY_C230_130_77 is ready for SNFS on 15e.

Code:

n: 18315142786666101398591398195652968891950079559697267218869070977624355302066323827855137263904494058903414979203244755702312169623822646226647486135677880809134691327837076059931140659435555511440691793413686862305600824600423447
# 137^77+77^137, difficulty: 262.22, anorm: 2.05e+038, rnorm: -2.57e+049
# scaled difficulty: 264.07, suggest sieving rational side
# size = 3.233e-013, alpha = 0.000, combined = 5.186e-014, rroots = 0
type: snfs
size: 262
skew: 1.1008
c6: 77
c0: 137
Y1: -24506804088319785713436649354236658982956133
Y0: 5989180101784270567563310697
rlim: 268000000
alim: 268000000
lpbr: 32
lpba: 32
mfbr: 64
mfba: 64
rlambda: 2.8
alambda: 2.8

Test sieving on the -r side with Q in blocks of 5K

Code:

Q=30M     13535
Q=60M     11612
Q=100M    10688
Q=160M    9560
Q=220M    8185
Q=300M    7481

suggesting a sieving range for Q of 30M-260M with a target # relations = 460M

swellman · 2018-02-08, 15:26

QUEUED C234_138_130 is ready for SNFS on 14e.

Code:

n: 873871377158311959084069104485448955380742878586910511066775683676005627366318727796089678603178221314609623090150332844943898936307004248150490182806133725295221071021681004836291150479521618813536089469216388339331568112155847161029
# 138^130+130^138, difficulty: 251.99, anorm: 2.76e+038, rnorm: -2.25e+048
# scaled difficulty: 253.64, suggest sieving rational side
# size = 9.276e-013, alpha = 0.000, combined = 1.125e-013, rroots = 0
type: snfs
size: 251
skew: 5.1676
c6: 1
c0: 19044
Y1: -995490680060699376274247097969055175781250
Y0: 28489183085523666542794945375463469260361
rlim: 268000000
alim: 268000000
lpbr: 32
lpba: 32
mfbr: 64
mfba: 64
rlambda: 2.8
alambda: 2.8

Test sieving on the -r side with Q in blocks of 5K

Code:

Q=20M    11629
Q=70M     9941
Q=130M    9048
Q=200M    8060
Q=280M    7355

suggesting a sieving range for Q of 20M-275M with a target # rels = 460M.

swellman · 2018-02-10, 16:37

QUEUED C180_132_95 is the composite cofactor of C235_132_95, which fell out when yoyo@Home found a p56. The C180 is close to the GNFS/SNFS threshold but test sieving shows it to be a decent SNFS candidate on 14e. GNFS would need a VERY good poly to beat SNFS, but when considering the estimated time to run a poly search and then properly test sieve it, SNFS seems to have the edge.

Code:

n: 289396665786312982141006307236099419705652823106983094814522618811166859549920672792894184549036158544227111489361203033421179354216009603309439787077402269680722529033925255839017
# 132^95+95^132, difficulty: 262.26, anorm: 4.60e+037, rnorm: -3.44e+049
# scaled difficulty: 264.24, suggest sieving rational side
# size = 8.792e-013, alpha = 0.000, combined = 1.039e-013, rroots = 0
type: snfs
size: 262
skew: 1.1282
c6: 16
c0: 33
Y1: -32353354497370936650756292250156402587890625
Y0: 4247690876126830874600666969407488
rlim: 268000000
alim: 268000000
lpbr: 32
lpba: 32
mfbr: 64
mfba: 64
rlambda: 2.8
alambda: 2.8

Test sieving on the -r side with Q in blocks of 5K

Code:

Q=20M    10037
Q=60M     9062
Q=120M    8467
Q=200M    7607
Q=280M    6754
Q=340M    6195

suggesting a sieving range for Q of 20M-310M with a target # rels = 460M

pinhodecarlos · 2018-02-10, 19:07

Quote:

Originally Posted by ATH

So apparently the large vectors makes sense even without MPI=1:

make all WIN=1 WIN64=1 ECM=1 CUDA=0 NO_ZLIB=1 VBITS=64/128/256:

msieve-svn1018-vbits64-haswell.zip
msieve-svn1018-vbits128-haswell.zip
msieve-svn1018-vbits256-haswell.zip

msieve-svn1018-vbits64-sandybridge.zip
msieve-svn1018-vbits128-sandybridge.zip
msieve-svn1018-vbits256-sandybridge.zip

For filtering stage of a GNFS130 I see a gain of 30% on those new binaries, nothing on the LA stage. Version tested were all sandybridge.

RichD · 2018-02-13, 05:19

QUEUED

also identified as phi_7(largest factor of phi_7(phi_13(7)))

C164 from the OPN t600 file.
P38.14020_5M.C164

Code:

n: 23868414476868587995325249170685588832943809103422768132521828316896025074624706715888406731264933402097870928021784797238895030842545830931640968187921163985020669
# 14020072124657538094183043839811142289^7-1
skew: 1.00
c6: 1
c5: 1
c4: 1
c3: 1
c2: 1
c1: 1
c0: 1
Y1: -1
Y0: 14020072124657538094183043839811142289
m: 14020072124657538094183043839811142289
type: snfs
rlim: 67000000
alim: 67000000
lpbr: 30
lpba: 30
mfbr: 60
mfba: 60
rlambda: 2.6
alambda: 2.6

Trial sieving 5K blocks.

Code:

 Q  Yield
20M 16361
50M 12366
80M 10196

VBCurtis · 2018-02-13, 06:29

QUEUED 13*2^847-1 for 14e queue:

Code:

n: 42246880894524383035892167318063256701851446750403840464286176017669351238138572605165092732275908592011242081859470348782066064144827879104751823280198945957929888255355032133967798016490782003
m: 2787593149816327892691964784081045188247552
type: snfs
size: 257
skew: 0.60
c6: 26
c0: -1
rlim: 134000000
alim: 200000000
lpbr: 33
lpba: 33
mfbr: 65
mfba: 92
rlambda: 2.7
alambda: 3.7

Trial sieving blocks of Q=2000:
Q=40M 0.152 sec/rel, 9863 rels
Q=100M 0.198 sec/rel, 8329 rels
Q=160M 0.204 sec/rel, 7925 rels
Suggest Q 15M-165M, for 620M rels.

I tested mfbr of 65, 91, 92, 93, 94, 95; 92 was the fastest of the 3LP options, with 20% better yield than 65 at a penalty of 5% speed (at the same Q). Since 92 will find more relations at lower/faster Q, choosing 92 means avoiding the slowest Q and is overall no slower. Also, it's fun to try something new in parameter selection with an eye toward extending 14e use via improved yields.
I'll do the LA.

Dubslow · 2018-02-14, 05:43

Quote:

Originally Posted by RichD

QUEUED

also identified as phi_7(largest factor of phi_7(phi_13(7)))

C164 from the OPN t600 file.
P38.14020_5M.C164

Under my aforementioned scheme, this would be named OPN_C164_7_13_1_7_6_7_7.

(Perhaps the OPN subscheme should include an extra letter to help discern what each field is, like OPN_C164_7_p13_n1_p7_n6_p7_n7 ? Where p means "phi" or "power" and "n" means "nth factor of previously-calculated number", or maybe something else.)

RichD · 2018-02-15, 04:57

QUEUED AS C168_AS3408_1679 C166 from Aliquot Sequence 3408.
AS 3408:i1679

Code:

n: 1849959698458141896282042340312824214549892860101677080063475429346433861792907412970467950297875110048406275074129147348999744724448993266489199137326482342635948767
skew: 2852369.7
lss: 0
c0: 79263284085565649297873532999343489290
c1: 79084889934473764927625577490723
c2: 44061537669541890460319774
c3: -87634034888970002123
c4: -7574786715264
c5: 265680
Y0: -106938861903569094031722097992052
Y1: 15868117662209966512381
type: gnfs
rlim: 134000000
alim: 134000000
lpbr: 31
lpba: 31
mfbr: 62
mfba: 62
rlambda: 2.6
alambda: 2.6

Trial sieving 5K blocks.

Code:

2018-02-13, 06:29	#1340
VBCurtis "Curtis" Feb 2005 Riverside, CA 2×2,437 Posts	QUEUED 132^847-1 for 14e queue: Code: n: 42246880894524383035892167318063256701851446750403840464286176017669351238138572605165092732275908592011242081859470348782066064144827879104751823280198945957929888255355032133967798016490782003 m: 2787593149816327892691964784081045188247552 type: snfs size: 257 skew: 0.60 c6: 26 c0: -1 rlim: 134000000 alim: 200000000 lpbr: 33 lpba: 33 mfbr: 65 mfba: 92 rlambda: 2.7 alambda: 3.7 Trial sieving blocks of Q=2000: Q=40M 0.152 sec/rel, 9863 rels Q=100M 0.198 sec/rel, 8329 rels Q=160M 0.204 sec/rel, 7925 rels Suggest Q 15M-165M, for 620M rels. I tested mfbr of 65, 91, 92, 93, 94, 95; 92 was the fastest of the 3LP options, with 20% better yield than 65 at a penalty of 5% speed (at the same Q). Since 92 will find more relations at lower/faster Q, choosing 92 means avoiding the slowest Q and is overall no slower. Also, it's fun to try something new in parameter selection with an eye toward extending 14e use via improved yields. I'll do the LA. Last fiddled with by fivemack on 2018-02-13 at 13:57*

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