Look! It's a big one!
 

[CODE]
Run 2 out of 52:
Using B1=110000000, B2=680270182898, polynomial Dickson(30), sigma=3001167417
Step 1 took 686315ms
Step 2 took 254074ms
********** Factor found in step 2: 516469933130631687266967194982169414626403685360388146231581267
Found probable prime factor of 63 digits: 516469933130631687266967194982169414626403685360388146231581267
Probable prime cofactor 3398721147883236398400986102209247253337482990178523382057924811844177557035850104125741577569050393864392779989 has 112 digits
Report your potential champion to Richard Brent <rpb@comlab.ox.ac.uk>
(see [url]ftp://ftp.comlab.ox.ac.uk/pub/Documents/techpapers/Richard.Brent/champs.txt[/url])
[/CODE]

This is a factor of 3,533+. I'll go for a beer now.

Prost,
Alex

congrats!

[QUOTE=akruppa]Look! It's a big one![/QUOTE]

Oooh er! Missus.

Most UK readers will understand that one. Everyone else can look up "Frankie Howerd" in the usual sources.


Anyway, congratulations are due. Don't get too drunk.


Paul

Congratulations! Maybe it has already been noted, but even the largest ECM factor found last year would no longer make it on to this years top 10!

:bounce:

[I sit on a bouncy ball at work]

Paul Zimmermann computed the group order:

>FindGroupOrder(p,3001167417);

[ <2, 4>, <3, 2>, <5, 1>, <29, 2>, <43, 1>, <131, 1>, <1231, 1>, <9539, 1>,
<88301, 1>, <186247, 1>, <727577, 1>, <1351957, 1>, <3520093, 1>, <10810561, 1>, <20946756331, 1> ]

The factor would even have been found with GMP-ECM 6.0.x with B1=11M due to the Brent-Suyama extension (with the default Dickson(12)).

Alex

Nice catch, congratulations!

[B]Alex[/B] congrats on such a large factor :w00t:

Is this now the second largest factor ever found by ecm?

It's the third largest: there's a p66 by Bruce Dodson and a p64 by K. Aoki & T. Shimoyama.

Alex

Congratulations, Alex! :bow: :banana:

:nuke: Respect :showoff: :showoff: :showoff: :surrender

Wow ! congrats !

Sorry to be late but a question : did you use the opteron cluster to find this one, Alex?
I mean, is this still a hope that simple mortals with one or maybe two cpu at home finds large ecm factors ? (I know the answer is yes but a mean of 7300 years to find a c60 with gmp-ecm on my desktop is a bit frustating :surrender )

The situation is the same with gnfs : with a cluster and time, you do a c180-c200. At home, a c120-c130+...

Bye.

Philippe.

Yes, this was on the cluster. Of course, having 100+ blazing fast cpus helps, but there was a very great deal of luck involved as well! Finding a p60 with one or two cpus is a perhaps not terribly likely, but a nice p5x should be managable if you keep working on it.

Alex

Gonna take

[code]
3398721147883236398400986102209247253337482990178523382057924811844177557035850104125741577569050393864392779989
[/code]

for a spin :geek: :showoff: with ggfns...

Umm... this number is prime. What do you want to use ggnfs for?

Alex

[QUOTE=akruppa]Umm... this number is prime. What do you want to use ggnfs for?

Alex[/QUOTE]

Oops ... my mistake :redface: it was 1 hour in the morning....