[QUOTE=wblipp;158876]That all said, I don't know if Paul has actually configured this server to spread the load around.[/QUOTE]The allocation is a mix of smallest number of curves yet run and random choice.


Paul

[QUOTE=xilman;159246]The allocation is a mix of smallest number of curves yet run and random choice.[/QUOTE]

That seems reasonable. I think the selection probability within each strategy is more important that the strategy choice when trying to avoid the overwork of any one number. I usually leave the time-adjustments near zero, but I like to make the base probability 10-30%. That spreads the choices among the top numbers in a geometric distribution. This especially helps at the transitions where it is possible for several people to get a P-1 assignment before any of them are returned.

William

[QUOTE=xilman;159246]The allocation is a mix of smallest number of curves yet run and random choice.

Paul[/QUOTE]
Given the ECM server setup -- and, in general, I would rather crank curves as part of a coordinated effort than go solo -- it is probably most efficient at this time to sieve some of the moderate-sized (SNFS 180-195) numbers to divert ECM resources to those above SNFS 200, right?

I am trying to reform my "OOOH! Sucha big number -- I must factor it!" attitude, and get more with a program of doing the most with the fewest resources.

BTW, I just ran through 750 curves at 43e6 for all of the remaining composites above 198 SNFS, and pulled a p51, a p46, and four p43's, which suggests that 11e6 is still the sweet spot.

I'll set the server refresh parameter at a few hours, instead of a few days, so there will be less likelihood of running 20000 curves at 11e6 on an SNFS 180.:redface:

[QUOTE=FactorEyes;159400]Given the ECM server setup -- and, in general, I would rather crank curves as part of a coordinated effort than go solo -- it is probably most efficient at this time to sieve some of the moderate-sized (SNFS 180-195) numbers to divert ECM resources to those above SNFS 200, right?

I am trying to reform my "OOOH! Sucha big number -- I must factor it!" attitude, and get more with a program of doing the most with the fewest resources.

BTW, I just ran through 750 curves at 43e6 for all of the remaining composites above 198 SNFS, and pulled a p51, a p46, and four p43's, which suggests that 11e6 is still the sweet spot.

I'll set the server refresh parameter at a few hours, instead of a few days, so there will be less likelihood of running 20000 curves at 11e6 on an SNFS 180.:redface:[/QUOTE]

It might be a good idea to concentrate on the 3,2+ and 3,2- tables,

[QUOTE=R.D. Silverman;159404]It might be a good idea to concentrate on the 3,2+ and 3,2- tables,[/QUOTE]
For aesthetic reasons, or for some technical reason?

Paul's tables push highest on 3,2. These seem to be more desirable than the others, for some reason.

11+3,199 = p98 . p109
Oh wait *. Wrong thread. Should go to Gratutitous Factors, I guess.

3+-2 are also fun, but obviously harder by design (500 lg3 is quite a big number). There are still two reducible-poly numbers there. OK. Let's do 'em in.

______
[SIZE=1][COLOR=navy]* "Coping with senility class?"[/COLOR][/SIZE]
[SIZE=1][COLOR=navy]"No! I am here for Microwave cookery! ...No, wait. Coping with senility."[/COLOR][/SIZE]

[quote=Batalov;156743]The name of the game is to get the number of composites under [B][SIZE=3]50[/SIZE][/B]. THEN the tables will be extended.

..this sounds easy 'nuff... :whistle:[/quote]
99 numbers remain unreserved (with 20 reserved)... :whistle:

One less 3,2
 

3^437-2^437 finished with an ECM hit. 3600 curves at B1=43e6, and out pops a p52:

3725509136887757555860365095946747063636252355137721

This eases my pain at finding three p51 factors in SNFS 20x-digit composites. The ECM gods can be kind after all.

[quote=FactorEyes;160094]3^437-2^437 finished with an ECM hit. 3600 curves at B1=43e6, and out pops a p52:

3725509136887757555860365095946747063636252355137721

This eases my pain at finding three p51 factors in SNFS 20x-digit composites. The ECM gods can be kind after all.[/quote]

That would come 7th on the top 10 ECM for 2009 - tell PZ about it!

Here are the factors for the C130 cofactor of 11^173-10^173

[code]PRP69 = 101386629190386639955758278863203184725117397301200142062026064955497
PRP62 = 14571403852010447907042557137975252494659540165402852287113331[/code]

This otherwise unremarkable result is perhaps more significant because I did it with QS, specifically, the development version of YAFU. Don't try jobs this big with version 1.06, it has a bug which will seriously impede progress. It is also generally wasteful to use QS at all... ggnfs would be much faster.

It took about 5 1/2 days on 50 cpus to gather ~1.4 million relations, which was about twice as many as were required. It probably could have been done in 4 days.

[code]
matrix is 730000 x 1451350 (578.4 MB) with weight 140007570 (96.47/col)
sparse part has weight 140007570 (96.47/col)
filtering completed in 4 passes
matrix is 668469 x 668530 (181.5 MB) with weight 42235578 (63.18/col)
sparse part has weight 42235578 (63.18/col)
saving the first 48 matrix rows for later
matrix is 668421 x 668530 (129.1 MB) with weight 34966551 (52.30/col)
sparse part has weight 27154059 (40.62/col)
matrix includes 64 packed rows
using block size 65536 for processor cache size 4096 kB
commencing Lanczos iteration
memory use: 114.0 MB
lanczos halted after 10570 iterations (dim = 668400)
recovered 7 nontrivial dependencies
Lanczos elapsed time = 2714.7400 seconds.
Sqrt elapsed time = 25.5900 seconds.
Total elapsed time = 2762.2900 seconds.

***factors found***
PRP69 = 101386629190386639955758278863203184725117397301200142062026064955497
PRP62 = 14571403852010447907042557137975252494659540165402852287113331
[/code]

[QUOTE=bsquared;161872]Here are the factors for the C130 cofactor of 11^173-10^173

[code]PRP69 = 101386629190386639955758278863203184725117397301200142062026064955497
PRP62 = 14571403852010447907042557137975252494659540165402852287113331[/code]

This otherwise unremarkable result is perhaps more significant because I did it with QS, specifically, the development version of YAFU. [/QUOTE]

WOOOHOOOOOO!! :w00t::bow wave::showoff: :banana:

This might be the biggest number factored with QS ever!!