I just did it with python:

python -c 'print 2**1061-1'

[QUOTE=Xyzzy]I'm probably just too dumb to understand, but after reading the help text in that file I still have no clue what this file does or what the data in it means... If someone could explain it I would appreciate it...[/QUOTE]
Here is a typical entry, a 200 digit divisor of 3^565-1, broken up into 100 digits per line:
[CODE]
200 3, 565- 8449691865231601552902335195572082885338639752166586003836785666188853803626377370798873039841747221$
200 0577269214324064975830634040825139392880871755088050507942142548159835160332523906713144063796260421
[/CODE]
There have been 725 curves done with B1=1e6, 4800+5110 with B1=3e6, etc. t40 means that testing has been completed to the 40 digit level, i.e. any more curves should be with B1=11e6:
[CODE]
725*1e6 + 4800*3e6 + 200*11e6 Horn + 5110*3e6+386*11e6 Ecmnet t40
[/CODE]
Dividing out the following factor leaves a 160 digit number for ecm to work on:
[CODE]
found 25980807832750689658262178521983417527371 R. Horn 13.1.2000
-> c160
[/CODE]

[QUOTE=xilman]No idea, but I could create you one within a few minutes.
[/QUOTE]

Thanks for the file, but it's just raw numbers. This way it's hard to see what effort has been put into a number.

[QUOTE=smh]Thanks for the file, but it's just raw numbers. This way it's hard to see what effort has been put into a number.[/QUOTE]
Ah. I misunderstood [QUOTE=smh]Is there any cleaned version of this file?
I mean, with only the remaining composites left?[/QUOTE] so I sent you a file containing only the remaining composites. Be careful what you ask for, because you may get it.

What, exactly, do you want?

Paul.

[QUOTE=xilman]What, exactly, do you want?[/QUOTE]

An online page that keeps track off all the work done on the remaining numbers from the cunningham tables, like there is for the 2+ and 2- tables on mersenne.org

If I have time, and I'm not promising anything so do not hold me to it, I may put up such a page in the near future i.e. in a month or so after I am back from vacation.

It won't be snazzy but it will have all the required info.

How are people working on f14? I did the following benchmarks last night on a fairly fast CPU and the times and memory usage are pretty high...

[code]bits | mem | stage1 | stage2 | b1 | b2 | curves | total time
-----+-----+--------+--------+---------+-----------+--------+-------------
15 | 8 | 03s | 03s | 2000 | 119805 | 30 | 03m00s
20 | 14 | 16s | 14s | 11000 | 1359460 | 90 | 45m00s
25 | 33 | 1m14s | 1m02s | 50000 | 11757135 | 240 | 9h04m00s
30 | 97 | 6m13s | 4m58s | 250000 | 116469998 | 500 | 93h11m40s
35 | 249 | 24m58s | 19m04s | 1000000 | 839549780 | 1100 | 33d15h16m40s[/code]I had to stop at 35 bits because I don't have enough memory, but it looks like the memory usage more than doubles every time you go up 5 bits...

An unrelated question: What does this file do?

[url]http://www.loria.fr/~zimmerma/ecmnet/cunningham.in[/url]

Thanks!

[QUOTE=Xyzzy]How are people working on f14? I did the following benchmarks last night on a fairly fast CPU and the times and memory usage are pretty high...

[code]bits | mem | stage1 | stage2 | b1 | b2 | curves | total time
-----+-----+--------+--------+---------+-----------+--------+-------------
15 | 8 | 03s | 03s | 2000 | 119805 | 30 | 03m00s
20 | 14 | 16s | 14s | 11000 | 1359460 | 90 | 45m00s
25 | 33 | 1m14s | 1m02s | 50000 | 11757135 | 240 | 9h04m00s
30 | 97 | 6m13s | 4m58s | 250000 | 116469998 | 500 | 93h11m40s
35 | 249 | 24m58s | 19m04s | 1000000 | 839549780 | 1100 | 33d15h16m40s[/code]I had to stop at 35 bits because I don't have enough memory, but it looks like the memory usage more than doubles every time you go up 5 bits...

An unrelated question: What does this file do?

[url]http://www.loria.fr/~zimmerma/ecmnet/cunningham.in[/url]

Thanks![/QUOTE]
#
Check out the -k parameter to gmp-ecm. Setting it high enough reduces the memory requirements subtantially, though at the cost of cpu time.

The file noted gives you all the composite Cunningham cofactors, and nothing else.

Paul

[QUOTE=Xyzzy]I had to stop at 35 bits because I don't have enough memory, but it looks like the memory usage more than doubles every time you go up 5 bits...[/QUOTE]

I guess you mean digits instead of bits?

[QUOTE=smh]I guess you mean digits instead of bits?[/QUOTE]Oops...
:whistle:

I started some ECMing on M2018 yesterday. As P1009 as well as M1009 are still early at the 45 digit level, I hope for an efficient use of the resources I put into it. Anybody willing to join me?

(Just to be sure:
I copied the factors of P1009 and M1009 into a M2018 entry in the lowm.txt file and now it works, right?)