The composite cofactor of 3,637+ was the product of 63-digit, 66-digit, and 109-digit prime numbers:

63-digit prime factor:
146653942939676501492528967729962226235222591603243036129895041

66-digit prime factor:
281492062846509383614609316625232146351554744294280968767884161701

109-digit prime factor:
6224453161048991898758218648741426180592061927925707078089776778181066640120779850430389525264538633126800183

Thanks go to M. Vang for performing the postprocessing. The factors have been reported to the Cunningham project and have been recorded on Page 125.

How much ecm was done on this integer?

How much ecm was done on this integer?

Same thing I was wondering on seeing p63*p66. Looks
like 26,987 curves with

Using B1=400000000, B2=5821851770290, polynomial Dickson(30)

with GMP-ECM 6.3 (Dec 2011) and 6.4 (May, 2012). This
was supposed to be a test sufficient to find a p60 to
probability of 62%; the limits are non-trivially above
p60-optimal (B1=260M), but well below p65-optimal (B1=850M).

This was a 15e number? If I recall, I went to 2t60 on
the most recent 16e pretest, but even that wouldn't have
given very substantial odds at finding one of p63, p66.
Maybe 50-50, if I'd doubled my curve count.

Good you've asked though; how much do we have on 11p301
and 11m301? I've just finished .6t60 on the c210 and
.4t60 on the c261 as initial passes for this recent
extension to the Cunningham lists. Not sure whether Sam
got his "usual" t60 before adding the extension; as these
two most recent ones were added early (to catch the p79
record!). Perhaps it's worth adding some more?

-bdodson (not sure that my p70 will stay in the top5 for 2012!)

bdodson ,

I really though it was an ecm miss when I saw those p63*p66, that's why I was wondering, like you, how much ecm was done. You answered my question.
BTW, aren't you going to go after first place here at NFS@Home? It would also be a way to quickly finish 2,1037- sieve until the end of the year.

Carlos

bdodson ,

...
BTW, aren't you going to go after first place here at NFS@Home? It would also be a way to quickly finish 2,1037- sieve until the end of the year.

Carlos

I'm happy to see Greg in first place. The University discontinued
the semi-public linux labs that I was using for most of the nfs@Home
computing I was doing. Those were commodity quadcores. I ran some
on the replacement research machines, but nfs sieving was causing them
to need occasional rebooting, which meant asking sysadmin assistance
instead of heading over to a lab where I could hit reboot. I'm still
doing gnfs sieving on our research machines, just not on Boinc - they're
run under a PBS scheduler, and don't allow interactive logins.

-Bruce (11p301 tested to t60, working on 11m301)