[QUOTE=Andi47;176340]... it sieved another 60kQ range (getting 614k additional relations) and ended up quite oversieved.)[/QUOTE]Lower the qintsize in the parameter file for such small numbers.

[QUOTE=smh;176358]Lower the qintsize in the parameter file for such small numbers.[/QUOTE]

That was my first thought too - I just had to wait for another (c100) GNFS to finish.

A second c93 GNFS took ~1:43 hours (still with the original quintsize.)

[QUOTE=unconnected;176149]48 composites remaining, it's time to extend tables?[/QUOTE]

:bump:

Extending these tables?
 

If we continue to limit ourselves to bases under 12, the remaining composites are SNFS 192 or larger. So there will be ECM and large-scale sieving activity from this point: no more small SNFS targets, with a few post-ECM GNFS cofactors of moderate size.

Does anyone have plans to extend the bases past 12?

I really don't know why I keep doing these - there's plenty of Cunningham stuff to do, and the Brent tables contain oodles of juicy medium-sized jobs. I think I suffer from because-it-is-there syndrome. Also, I have seen too damn many projective equations, so that factoring (11/4)^181 - 1 seems as valid as factoring 11^181-1.

I'm not in a hurry to have the tables re-filled, because the remaining 45 composites contain many GHz-years of work.

What direction do we now head with these tables?

What do you mean by "large-scale sieving"? SNFS220 is quite accessible - say two months on one quad-core - and SNFS192 is a lot easier than that.

[QUOTE=fivemack;179044]What do you mean by "large-scale sieving"?.[/QUOTE]Medium (or small?) by current standards, but large by the standards of most of the people working on these particular numbers.

How many people have sieved an SNFS 192 or larger on these? Counting on my fingers, and adding 2 for good measure, I come up with 8. I believe that only 4 of us have run more than two sieving jobs above SNFS 200.

[QUOTE]SNFS220 is quite accessible - say two months on one quad-core - and SNFS192 is a lot easier than that[/QUOTE]

[CODE]5^298+2^298 162.5 208.3 Reserved by Tom Womack at 20080906003858; if that's you, click here to unreserve it
5^298+4^298 174.2 208.3 Reserved by Tom Womack at 20080906003909; if that's you, click here to unreserve it[/CODE]

:wink:

It doesn't matter how quickly these numbers get cracked, but it appears that there are few spare CPU cycles for SNFS jobs on these. They are mostly for fun - for the old PC in the den.

[QUOTE=FactorEyes;179043]If we continue to limit ourselves to bases under 12, the remaining composites are SNFS 192 or larger. So there will be ECM and large-scale sieving activity from this point: no more small SNFS targets, with a few post-ECM GNFS cofactors of moderate size.

Does anyone have plans to extend the bases past 12?

I really don't know why I keep doing these - there's plenty of Cunningham stuff to do, and the Brent tables contain oodles of juicy medium-sized jobs. I think I suffer from because-it-is-there syndrome. Also, I have seen too damn many projective equations, so that factoring (11/4)^181 - 1 seems as valid as factoring 11^181-1.

I'm not in a hurry to have the tables re-filled, because the remaining 45 composites contain many GHz-years of work.

What direction do we now head with these tables?[/QUOTE]I still intend to extend the tables Real Soon Now.

Unfortunately, Real Life (tm) and ill health has intervened. Despite a large degree of automation, it still takes time and effort to keep things running smoothly and I just haven't had the spare bandwidth recently.


Paul

[QUOTE=xilman;179111]I still intend to extend the tables Real Soon Now.

Unfortunately, Real Life (tm) and ill health has intervened. Despite a large degree of automation, it still takes time and effort to keep things running smoothly and I just haven't had the spare bandwidth recently.[/QUOTE]At last, I've had a bit of spare time.

The final version of the present tables is uploading right now. There are 44 remaining composites.

The extended tables are in preparation. There will be over 1400 composites, enough to keep people busy for a long time to come. It will take me a little time to update the ECM server to include the new composites at an appropriate level of B1 allocations. It will probably take Tom Womack some time for him to extend his allocation server likewise. Once the preparatory work is done I'll announce it here.


Paul

Dmitry, aka unconnected, reminded me that I'd missed one of his factors. That is now incorporated in the new tables.

I still don't have the factorization of 11+5,142 c113 which must have been done a long time ago. If anyone can find the historical record, please mail it to me. Otherwise, if anyone wants an easy SNFS to complete ...


Paul

I'll do it in few hours.

[QUOTE=xilman;180601]The extended tables are in preparation. There will be over 1400 composites, enough to keep people busy for a long time to come.[/QUOTE]The extended tables are now loaded into Phil Carmody's ECMNET server located at 83.143.57.194:8194. They are also available on my web page at [url]http://www.leyland.vispa.com/numth/factorization/anbn/main.htm[/url] but Tom Womack has not yet had chance to change his reservation page at [url]http://www.chiark.greenend.org.uk/ucgi/~twomack/homcun.pl[/url] (entirely because I've not yet told him that he needs to!)

There are 1452 composites in the new list, enough to keep everyone busy for some substantial time.


Paul

P.S. Changed the thread title in celebration.