My earlier whining aside, I'm not worried about putting so many curves into a number that it could have been SNFS'ed anyway. I assume a lot of contributors are running fire-and-forget, and curves are curves.

If it's fairly simple, can you skew the assignments so that numbers with higher SNFS difficulty get more curves? I assume you have thought of this; there may be some reason for not bothering to do this, as in getting the smaller gaps in the table filled first. (I have been kind of perverse in concentrating on SNFS > 200 for the past year.)

BTW, 3^479+2^479 just coughed up a p52 with B1=43e6, only 12 hours after my p53:

2908867208146159972104994923719960261660671813159759

This rash of p5x factors is positively eerie.

[quote=FactorEyes;162145]My earlier whining aside, I'm not worried about putting so many curves into a number that it could have been SNFS'ed anyway. I assume a lot of contributors are running fire-and-forget, and curves are curves.

If it's fairly simple, can you skew the assignments so that numbers with higher SNFS difficulty get more curves? I assume you have thought of this; there may be some reason for not bothering to do this, as in getting the smaller gaps in the table filled first. (I have been kind of perverse in concentrating on SNFS > 200 for the past year.)

BTW, 3^479+2^479 just coughed up a p52 with B1=43e6, only 12 hours after my p53:

2908867208146159972104994923719960261660671813159759

This rash of p5x factors is positively eerie.[/quote]

Remember to tell Paul Zimmermann about your factors - they would get on to the Top 10 list.

[QUOTE=FactorEyes;162145]My earlier whining aside, I'm not worried about putting so many curves into a number that it could have been SNFS'ed anyway. I assume a lot of contributors are running fire-and-forget, and curves are curves.

If it's fairly simple, can you skew the assignments so that numbers with higher SNFS difficulty get more curves? I assume you have thought of this; there may be some reason for not bothering to do this, as in getting the smaller gaps in the table filled first. (I have been kind of perverse in concentrating on SNFS > 200 for the past year.)

BTW, 3^479+2^479 just coughed up a p52 with B1=43e6, only 12 hours after my p53:

2908867208146159972104994923719960261660671813159759

This rash of p5x factors is positively eerie.[/QUOTE]


Very nice indeed. This and 3,2,463- saved a lot of time with NFS.

It's been over a month since the last update. Paul: How about a new one???

[QUOTE=FactorEyes;162145] ...
BTW, 3^479+2^479 just coughed up a p52 with B1=43e6, only 12 hours after my p53:

2908867208146159972104994923719960261660671813159759

This rash of p5x factors is positively eerie.[/QUOTE]

Let me second, third, ... the suggestion that you report the p53, at least.
Too many p52's for the second month of 2009. From PaulZ's top10 page:
[QUOTE]
Please send an email to ``zimmerma [at] loria (dot) fr'' for any remark about this page. [/QUOTE]
-Bruce

[QUOTE=R.D. Silverman;162201]Very nice indeed. This and 3,2,463- saved a lot of time with NFS.

It's been over a month since the last update. Paul: How about a new one???[/QUOTE]I'll see what I can do.


Paul

[QUOTE=R.D. Silverman;162201]Very nice indeed. This and 3,2,463- saved a lot of time with NFS.[/QUOTE]

Thanks. Per Bruce's suggestion, I'll send the p53 on to PaulZ, together with the p54 that just came in this afternoon. We can kiss 3^491+2^491 goodbye:

[CODE]((3^491+2^491)/5)/28938788771
Using B1=43000000, B2=240490660426, polynomial Dickson(12), sigma=1736703962
Step 1 took 331609ms
Step 2 took 113159ms
********** Factor found in step 2: 219582944554227876832019548377992760446825287299367399
Found probable prime factor of 54 digits: 219582944554227876832019548377992760446825287299367399
Probable prime cofactor (((3^491+2^491)/5)/28938788771)/219582944554227876832019548377992760446825287299367399 has 170 digits
[/CODE]

It ain't supposed to be this easy. I would be tried for witchcraft in some centuries.

Seriously: 11 Opteron cores (8 2.0 GHz Barcelona cores and 3 2.67 GHz 2218 cores) running all day, averaging around 2700 curves each day at B1=43e6. I'm hitting each of the unreserved 3,2 numbers with 3775 43e6 curves, which is halfway to the p50 level, and in under 72 hours I have pulled a p52, a p53, and a p54. And my first name is not Bruce.

Schneier?

[QUOTE=FactorEyes;162301]Thanks. Per Bruce's suggestion, I'll send the p53 on to PaulZ, together with the p54 that just came in this afternoon. We can kiss 3^491+2^491 goodbye:

[CODE]((3^491+2^491)/5)/28938788771
Using B1=43000000, B2=240490660426, polynomial Dickson(12), sigma=1736703962
Step 1 took 331609ms
Step 2 took 113159ms
********** Factor found in step 2: 219582944554227876832019548377992760446825287299367399
Found probable prime factor of 54 digits: 219582944554227876832019548377992760446825287299367399
Probable prime cofactor (((3^491+2^491)/5)/28938788771)/219582944554227876832019548377992760446825287299367399 has 170 digits
[/CODE]

It ain't supposed to be this easy. I would be tried for witchcraft in some centuries.

Seriously: 11 Opteron cores (8 2.0 GHz Barcelona cores and 3 2.67 GHz 2218 cores) running all day, averaging around 2700 curves each day at B1=43e6. I'm hitting each of the unreserved 3,2 numbers with 3775 43e6 curves, which is halfway to the p50 level, and in under 72 hours I have pulled a p52, a p53, and a p54. And my first name is not Bruce.[/QUOTE]

No, It's Michael!

G'day Bruce!

Update posted
 

An update to the tables has just been uploaded. It covers an especially productive six weeks. There are 47 new factors reported and the number of composites has now fallen from 149 to 108.

The tables will be extended when the number of composites falls below 50. At current rates that won't be long.

Thanks to all who contributed, whether or not you were able to find a factor during this period. Even unsuccessful ECM runs contribute to progress by indicating when it might be appropriate to switch to SNFS.


Paul

[QUOTE=FactorEyes;162301]Thanks. Per Bruce's suggestion, ...
in under 72 hours I have pulled a p52, a p53, and a p54. And my first name is not Bruce.[/QUOTE]

After 3-4 readings, I gather we're to take this as (an upper bound on)
a cpu count? Thanks for the ECMNET factor reports; the top10 looks
much better now. -bd

Tom must have one more cracked number hidden in his logs. He's got 107.