[QUOTE=jrk;229531]It's fine, as long as you don't mind being blown out of the water by a nice ECM factor![/quote]I'm not that worried....it was either start this or be tied up with my next NFS jobs (two c140s just sitting there staring at me).[quote]Uh, whoa. Really? That sounds bad. 5% duplication seems high for what amounts to less than 1/10 of the needed work done... remember, duplication rate goes up as you sieve more Q. 15-25% duplication would be expected for the entire job.[/QUOTE]It's a steady 5%; I haven't seen any of the increase I expect while sieving. I had one job where I was approaching 30% as it finished; I was starting to worry that I would run out of Qs!

After sieving about 4M Qs, here's what msieve reports on a filtering run:[code]keeping 26161485 ideals with weight <= 200, target excess is 41177
commencing in-memory singleton removal
begin with 8633372 relations and 26161485 unique ideals
reduce to 135 relations and 0 ideals in 4 passes
max relations containing the same ideal: 0
filtering wants 1000000 more relations
elapsed time 00:14:51[/code]

[QUOTE=schickel;229538]It's a steady 5%; I haven't seen any of the increase I expect while sieving. [/QUOTE]
In that case I wonder if an output file might have been cat'ed twice at some point, because sieving just Q=16M to Q=20M shouldn't make 5% duplicates.

[QUOTE=schickel;229538]After sieving about 4M Qs, here's what msieve reports on a filtering run:[code]keeping 26161485 ideals with weight <= 200, target excess is 41177
commencing in-memory singleton removal
begin with 8633372 relations and 26161485 unique ideals
reduce to 135 relations and 0 ideals in 4 passes
max relations containing the same ideal: 0
filtering wants 1000000 more relations
elapsed time 00:14:51[/code][/QUOTE]
Doing the filtering so often is just slowing you down (and it'll keep getting longer and longer...). The filtering is not done incrementally, filtering runs on the whole dataset at once. And in fact if you proceed filtering this often, you might not even want to solve the first matrix it produces after successful filtering, because it will be huge. Further sieving (than what is strictly required for a matrix) will make a much nicer matrix.

[quote=jrk]Allow me to ask, why do you not allow the 15e siever?[/quote]
Well, because we haven't deployed 15e yet, because both squalyl and I are busy both at work (our respective daily jobs are unrelated to integer factoring) and outside of work :smile:

Add 106 curves at B1=11e7, B2=776278396540

My machine was quite a bit slower than the estimate - about 1 hour 8 minutes per curve.

I have another one working with a slightly different B2 and a little longer per curve. I may have to stop that one before it finishes 100.

The c170 is finished as p55*p115. Next line has some small factors and a [URL="http://factordb.com/index.php?id=1100000000230134940"]c144[/URL]. AFAIK minimal ECM has been done - I'm currently running 214@5e4 to complete 25-digit level.
Edit 1: Done with 214@5e4, (25 digits) running 430@25e4 (30 digits).
Edit 2: Done with 150@25e4, finishing the rest of that and started 300@1e6 (part of 35 digits).
Edit 3: Done P-1 at B1=100000000 (1e8 or 100M), B2=4537592002678 (default), 3 runs of P+1 at B1=50000000, (5e7 or 50M) B2=1512545248792 (default), running P-1 at B1=1e9. Current ECM status is 362@25e4 and 132@1e6.

Found probable prime factor of 40 digits: 2064317755979677441652715999336087183653
c105 cofactor is composite.
update: cracked as p36*p69. Now c167.

[QUOTE=unconnected;236125]Found probable prime factor of 40 digits: 2064317755979677441652715999336087183653
c105 cofactor is composite.[/QUOTE]

Great! Unless someone with more firepower wants to run it and finish it faster, I've started this on my computer (quad i5). (I figure it probably has enough ECM to run GNFS by now)
Edit: By the way, we lost the 7 factor that was driving the sequence higher, but it was replaced by an 11. I guess it'll go up again after this index by a bit, but hopefully we can start a downwards trend or even a downdriver soon.

[QUOTE=Mini-Geek;236127]Great! Unless someone with more firepower wants to run it and finish it faster, I've started this on my computer (quad i5). (I figure it probably has enough ECM to run GNFS by now)
Edit: By the way, we lost the 7 factor that was driving the sequence higher, but it was replaced by an 11. I guess it'll go up again after this index by a bit, but hopefully we can start a downwards trend or even a downdriver soon.[/QUOTE]

you have noticed unconnecteds edit?and stopped the gnfs?

[QUOTE=Dougal;236136]you have noticed unconnecteds edit?and stopped the gnfs?[/QUOTE]

Yes.
We are now on index 2553 with a c145. The 11 cofactor is now a 17. Still nudging the sequence up very slowly. I'm doing a little P-1 on this one (B1=100M). Edit: P-1 is done with B1=100M and B2 default.

ill do some curves if you let me know what bounds to use,i cant seem to see on the factordb,what work has been done.

I've started the c145 on 8 threads on a core i7. It will run the default ecm series on 7 threads for as long as the polyfind takes to complete. If someone with bigger guns wants to take over, or if it cracks with ecm, post here and I'll kill it. Otherwise a c145 will probably take me ~4 days.