Msieve / lattice siever with degree 7/8 poly

[henryzz]

Originally Posted by 10metreh

They are algebraic factors: the p20 is from 5^29+4^29, the 23 * 256147 are from 5^11+4^11 and the 3^2 is from 5^1+4^1.

thanks
i presume that why they arent in the tables

[Andi47]

Originally Posted by henryzz

BTW Andi47 read posts 17,18 and 20 of this thread

Sorry - I thought "just call me Henry" below your avatar really meant that...

edit: when i came to calculate the number there were lots of addtional factors that werent mentioned here including a p20 that i had to find myself

When you click the reservation link, you will see the remaining cofactor - so you will not have to split small factors off...

BTW: Which kind of computer do you use for this factorization? For a Core 2 Duo (2 GHz, using both threads) I would estimate ~2 weeks. Parameters (rlim, alim, lpbr, and such...) can be found over there in the alq4788-thread. For the lambdas it seems to be near the crossover between 2.5 and 2.6 - maybe you want to do some test sieving.

[henryzz]

Originally Posted by Andi47

Sorry - I thought "just call me Henry" below your avatar really meant that...

surely u have been around the forum long enough to know that the only way to get comments below ur name is for xyzzy to add them as a joke
its ok i expect this will happen a lot
i am just hoping if i spread the word people will read that before posting

When you click the reservation link, you will see the remaining cofactor - so you will not have to split small factors off...

BTW: Which kind of computer do you use for this factorization? For a Core 2 Duo (2 GHz, using both threads) I would estimate ~2 weeks. Parameters (rlim, alim, lpbr, and such...) can be found over there in the alq4788-thread. For the lambdas it seems to be near the crossover between 2.5 and 2.6 - maybe you want to do some test sieving.

a Q6600 which is currently at stock
my pc doesnt run 24/7 which limits my size quite a bit so does a limit of 2Gb of memory
also i am impatient
thats exactly what i was thinking
pinch the parameters from the aliquot forum
so far i only have the one poly from msieve a 2.387e-011 which is over 2/3 of the one for the latest 4788 factorization

[Andi47]

Originally Posted by henryzz

surely u have been around the forum long enough to know that the only way to get comments below ur name is for xyzzy to add them as a joke
its ok i expect this will happen a lot
i am just hoping if i spread the word people will read that before posting



a Q6600 which is currently at stock
my pc doesnt run 24/7 which limits my size quite a bit so does a limit of 2Gb of memory
also i am impatient
thats exactly what i was thinking
pinch the parameters from the aliquot forum
so far i only have the one poly from msieve a 2.387e-011 which is over 2/3 of the one for the latest 4788 factorization

Q6600 (quad-core) @2.4 GHz? If you run it ~10 hours/day it should be equally fast as a C2D (=dual core) @2.0 GHz running 24/7, so ~2 weeks should still apply.

BTW: With 2 GB RAM it should be possible to GNFS anything up to ~c155 (maybe even c260) when using a linux environment. With Windows I won't do anything above c150.

[10metreh]

Originally Posted by henryzz

surely u have been around the forum long enough to know that the only way to get comments below ur name is for xyzzy to add them as a joke
its ok i expect this will happen a lot
i am just hoping if i spread the word people will read that before posting

Xyzzy can be nasty... look at what happened to flouran. It might have been a joke, but it was still nasty. Somehow I have escaped without a subdescription - and I would like to keep it that way.

[henryzz]

Originally Posted by Andi47

Q6600 (quad-core) @2.4 GHz? If you run it ~10 hours/day it should be equally fast as a C2D (=dual core) @2.0 GHz running 24/7, so ~2 weeks should still apply.

BTW: With 2 GB RAM it should be possible to GNFS anything up to ~c155 (maybe even c260) when using a linux environment. With Windows I won't do anything above c150.

i supose i can run it 12 hours a day for the next week and a half so yes i should hopefully finish in 2 weeks and after that i will only be down to 4-5 if not
i am not hugely patient with the long jobs though so i will probably do some other jobs in the middle of the factorization but i should still finish in 2 weeks as i might recruit another pc to do a few hours work
something that will significantly affect my patience is having to do it in linux away from the things i usually do

edit: mods would a threadsplit be worthwhile? split the posts about my factorization into thier own thread and maybe all the posts about the "Just call me Henry" thing into another(maybe including the posts i quoted to start that off here)

[xilman]

Originally Posted by 10metreh

Xyzzy can be nasty... look at what happened to flouran. It might have been a joke, but it was still nasty.

Believe me, it could get a lot worse.

Paul

[henryzz]

i have reached the linear algebra stage with 5^319+4^319
it should take about 6 and a half hours for the 2M^2 which means i should get the results tomorrow evening

[henryzz]

Originally Posted by henryzz

i have reached the linear algebra stage with 5^319+4^319
it should take about 6 and a half hours for the 2M^2 which means i should get the results tomorrow evening

prp55 factor: 6852013642934374356846035958993519237208519773511625423
prp83 factor: 12968222269216341341220108397357041906689524861847658294907716830211054521792631607

[Batalov]

I suggest splitting posts from #40 into a separate thread.
There's no degree 7-8 data in them.

[henryzz]

Originally Posted by Batalov

I suggest splitting posts from #40 into a separate thread.
There's no degree 7-8 data in them.

i suggested this at the bottom of #50 as well