Factoring humongous Cunningham numbers

[LaurV]

Originally Posted by xilman

15245685683654194070528451784367735927134564102584982532265254072253746820313011591722538299058918175786099377638450142276578178842074686064451121444334808941
67913146583098242945234710034040860323734180924846788218077857823909839832664083294529021164010342768692899249006235945630154935616014528457767660874893572764737

Ok, I will do the first one, and in case that nobody works for the second one, I will do that one too.

[jyb]

Originally Posted by xilman

The smallest remaining HCN is this C161 from 8,3+,248

Originally Posted by LaurV

Ok, I will do the first one, and in case that nobody works for the second one, I will do that one too.

Be aware that 8^248+3^248 is probably easier by SNFS (difficulty 224).

[R.D. Silverman]

Originally Posted by jyb

Be aware that 8^248+3^248 is probably easier by SNFS (difficulty 224).

Yes.

[xilman]

Originally Posted by jyb

Be aware that 8^248+3^248 is probably easier by SNFS (difficulty 224).

Indeed. That's true of all remaining HCN candidates AFAIK because the SNFS cut-off limit is 255 digits and the smallest composites are all greater than 160 digits (~= 240 digits by SNFS). The raw data is easily available to anyone who wishes to check.

Those two were given because GNFS was specifically requested.

[xilman]

Originally Posted by LaurV

Ok, I will do the first one, and in case that nobody works for the second one, I will do that one too.

As has been pointed out, it will be easier by SNFS. Are you comfortable with finding the polynomial yourself? I'm not trying to be patronizing; I genuinely don't know your level of expertise in this field. Most anyone contributing to this thread can give assistance if required.

Paul

[jyb]

Originally Posted by debrouxl

I have just queued "7293_minus_5293" to NFS@Home's 14e.

I've reserved this number to NFS@Home on the reservation page. Not knowing what else to use, I did it with my own email address. PM me if you need the reservation code or address.

[LaurV]

Originally Posted by jyb

Be aware that 8^248+3^248 is probably easier by SNFS (difficulty 224).

Originally Posted by R.D. Silverman

Yes.

The million dollar question: Is yafu so clever to see that? (because certainly I am not, I was counting on an "easy" job, copy/paste the two numbers, launch "yafu @batchfile", job done. I know BB did a wonderful job in teaching yafu about SNFS, but I am still at that stage. The first number is sieving already after it went through 0.3334 ECM last night (that is, 33% according with yafu encoding). Now you two started confusing me with inconspicuous mathematical terms...

[debrouxl]

If yafu can't produce polynomials for the homogeneous Cunningham numbers, it's the duty of the "phi" tool, which I found around on MersenneForum a while ago.

[jyb]

Originally Posted by debrouxl

If yafu can't produce polynomials for the homogeneous Cunningham numbers, it's the duty of the "phi" tool, which I found around on MersenneForum a while ago.

Specifically, here. Make sure you grab the latest version.

[LaurV]

Thanks a lot.
On the other hand, I said that the first one is sieving, which was wrong. In the morning when I left there were only few curves left from the ECM, so I assumed that at the time of posting (from job, lunch break) it already moved to sieving. But I totally missed the poly part. I just arrived home this evening and it still has 50 hours or so of poly... Sorry for misinforming you.
I will get over it, one step at a time...

[bsquared]

Originally Posted by debrouxl

If yafu can't produce polynomials for the homogeneous Cunningham numbers, it's the duty of the "phi" tool, which I found around on MersenneForum a while ago.

yafu can produce polynomials for these. And it chooses snfs over gnfs. When I tried it just now it started test sieving between the degree 5 and degree 6 polynomials.

[EDIT}
It picked the degree 5:

c5: 144
c0: 1
ETA: 70 days 3h 11m 11s (with 1 threads)

c6: 64
c0: 9
ETA: 78 days 2h 1m 54s (with 1 threads)

c6: 81
c0: 1
ETA: 83 days 8h 56m 10s (with 1 threads)