Factoring database

[MisterBitcoin]

Originally Posted by vebis

I add factors for 10666 c109s today.

WHOA.
Did you "only" do some small ECM or NFS to find those factors. I noticed an small amount of c78´s, which speeaks for ECM.
Anyway, good job!


Will you expend your work on C109? I can also do some ECM/NFS, so that the amount will go even lower.

[vebis]

Originally Posted by MisterBitcoin

WHOA.
Did you "only" do some small ECM or NFS to find those factors. I noticed an small amount of c78´s, which speeaks for ECM.
Anyway, good job!


Will you expend your work on C109? I can also do some ECM/NFS, so that the amount will go even lower.

It was ecm to pretesting depth of 33.54 or 40% of c109

[MisterBitcoin]

Starting YAFU for C73. We have over 700 from them.


Edit: Should be done in 3-4 hrs. Slowly, but steady process. (The Perl script from yoyo is nice. )

[chris2be8]

I was working on 73 digits, but I've switched my scripts to skip that range (I'm doing 70-72 digits on one system and 74-78 on another now).

Thanks for the help (someone has added a load of smallish composites to factordb and I was struggling to keep up with them).

Chris

[MisterBitcoin]

Originally Posted by chris2be8

I was working on 73 digits, but I've switched my scripts to skip that range (I'm doing 70-72 digits on one system and 74-78 on another now).

Thanks for the help (someone has added a load of smallish composites to factordb and I was struggling to keep up with them).

Chris

When I´m done I´ll switch to C79 up to C94 for the next couple of days.
Edit: Sofar 0 collisions; the rdm select function worked fine.

[MisterBitcoin]

Originally Posted by MisterBitcoin

When I´m done I´ll switch to C79 up to C94 for the next couple of days.
Edit: Sofar 0 collisions; the rdm select function worked fine.

C73 is now done, you can have it "back".
Now working on the range C79 C89-C94, will take some time.

[EdH]

Originally Posted by MisterBitcoin

C73 is now done, you can have it "back".
Now working on the range C79 C89-C94, will take some time.

We might be colliding a bit. I'm not sure how easy it is for you to set where you work, but I'd have to change a lot of machines. My setup is working from 79 up with a 100 element random window. So, if I could ask you to work a number higher than whatever has a lot, we might be less likely to duplicate. Of, course, I'm not sure if we really are colliding (I'm not seeing many). As I'm writing this, there are 130 entries for 92 dd and about 30 from 79-91 dd. That means my machines are hitting from 79 to 92 and nothing higher ATM. If you could work above the bulky low boundary by a couple hundred, it might be better. But, if you're not seeing any "already knowns" then there really isn't much duplication at this level.

[chris2be8]

Originally Posted by MisterBitcoin

C73 is now done, you can have it "back".

Thanks, I'll tackle 70-78 digits from now on.

Chris

[MisterBitcoin]

Originally Posted by EdH

We might be colliding a bit. I'm not sure how easy it is for you to set where you work, but I'd have to change a lot of machines. My setup is working from 79 up with a 100 element random window. So, if I could ask you to work a number higher than whatever has a lot, we might be less likely to duplicate. Of, course, I'm not sure if we really are colliding (I'm not seeing many). As I'm writing this, there are 130 entries for 92 dd and about 30 from 79-91 dd. That means my machines are hitting from 79 to 92 and nothing higher ATM. If you could work above the bulky low boundary by a couple hundred, it might be better. But, if you're not seeing any "already knowns" then there really isn't much duplication at this level.

I´m only running process on my tower. So only 3-6 hrs/day.
I dont see any need for numbers <C95. I´ll do some ECM on C100´s; using random generator range of 500 to prevent collisions.

[VBCurtis]

Nearly all of the massive backlog of C100's split as P50*P50, so ECM is wasted time. I'd skip ECM and take my chances on specifically C100s.

[EdH]

Originally Posted by VBCurtis

Nearly all of the massive backlog of C100's split as P50*P50, so ECM is wasted time. I'd skip ECM and take my chances on specifically C100s.

I modified my scripts to skip ECM whenever the composite is 100 dd based on richs' post elsewhere, but it will be a little while before I start touching any c100s again.