mersenneforum.org what happens next?
 User Name Remember Me? Password
 Register FAQ Search Today's Posts Mark Forums Read

 2006-01-07, 01:34 #1 NeoGen   Dec 2005 2×33 Posts what happens next? "We have 1556088 untested values left under n=50 Million" One day when all these untested values are done, what happens? Does the project end? Or does it restart sieving and PRP'ing for some higher value like n=100 million? (for example)
 2006-01-07, 02:33 #2 Citrix     Jun 2003 1,579 Posts We haven't thought that far. It will take atleast 10 years to get there, if we had all the computers from GIMPS. This is the short answer. Though we are hoping we will find all the primes (12 left to be found) soon and the project would end by 50 million. We are open to suggestions, if you have any. Citrix edit: If the project is incomplete at that time and we have enough computing power, I personally would like to continue to a higher value. Last fiddled with by Citrix on 2006-01-07 at 02:35
 2006-01-07, 05:06 #3 NeoGen   Dec 2005 2·33 Posts 10 years?? Oh well... plenty of work for the future then. By the way... what's the relation between these untested values and the candidates in the prp range reservation thread? There are so few candidates comparing to the untested values...
2006-01-07, 05:29   #4
Citrix

Jun 2003

1,579 Posts

Quote:
 Originally Posted by NeoGen 10 years?? Oh well... plenty of work for the future then. By the way... what's the relation between these untested values and the candidates in the prp range reservation thread? There are so few candidates comparing to the untested values...
10 years is an estimate. We could be done by tommorow, if we find all the primes, but the probability of that is low.

All untested numbers need to be PRPed. The numbers in the PRP reservation thread are the numbers most likely to produce a prime and the numbers that are easiest to PRP right now.

Citrix

 2006-01-07, 12:14 #5 japelprime     "Erling B." Dec 2005 23·11 Posts I am rather new to this prime hunting but I would like to go after 10 million digit prime. It will take time..... if I am correct any PRP candidate will take 3 month to 1 year depending on today computerpower you have. Starting sieve some 10 million digit range wery deep would be fine by me. Maybe it is boring for some one not having any results for looooong time.
 2006-01-07, 15:54 #6 hhh     Jun 2005 373 Posts if you want to be the one who discovers the 10 million digit prime, the best would be to refer to GIMPS. There are some suggestions at Seventeenorbust from time to time to open a "largest Prime" queue, but it never went very far. The main problem is that these tests are quite likely to be wasted, as we might find a prime for that very k before. I prefactored some k,n-pairs fairly deep; but I am reluctant to encourage anybody to test these as they wouldn't be assigned by the server, so one wouldn't get any credit, and the residue submission would be manual, giving the admins even more work to do. If you are really interested, you will easily find the thread at sob. H.
 2006-01-07, 18:49 #7 Citrix     Jun 2003 1,579 Posts As always, you can do anything you want at PSP. If someone wants to do a 10 Million digit PRP test, they are most welcomed to do so (get in touch with ltd). But as hhh pointed out, it is not the best thing to do right now. Other than that I had an idea. If we take the k with the least number of candidates and concentrated our effort on that k, we might be able to take it to 10M+ digits and find a prime. Though the probability of finding a prime is low, and this will be like a gamble. But if we are lucky we would get the \$100K. Citrix
 2006-01-07, 18:56 #8 Citrix     Jun 2003 110001010112 Posts For k=156511 there are ~54000 numbers left. We could do some p-1, ecm , P+1 and get rid of more numbers. Then do PRP etc. Citrix
 2006-01-07, 19:12 #9 NeoGen   Dec 2005 668 Posts There is an "low n ECM factoring" thread, and has only 16 numbers there. If it helps to eliminate more of those untested values we could jump on them and get them quickly done. But alas, there isn't a tutorial on that thread. I don't have a clue on how to do it, or what program to use for the job... lol
2006-01-07, 19:13   #10
Citrix

Jun 2003

1,579 Posts

Quote:
 Originally Posted by NeoGen There is an "low n ECM factoring" thread, and has only 16 numbers there. If it helps to eliminate more of those untested values we could jump on them and get them quickly done. But alas, there isn't a tutorial on that thread. I don't have a clue on how to do it, or what program to use for the job... lol
It doesn't really help the project, it is just for fun. That is why there is no tutorial.

Citrix

 2006-01-08, 00:49 #11 japelprime     "Erling B." Dec 2005 1308 Posts I can see the problem with 10Million digit. I will jump on the train with you whatever decision will be made. If you have some wery slow PC´s (like I have) that can´t be to any use other than multisieving (NewPGen) then it could be som fun multisieving big numbers whatsoewer the maingoal is. Just an idea.

All times are UTC. The time now is 22:36.

Sat May 8 22:36:05 UTC 2021 up 30 days, 17:16, 0 users, load averages: 3.25, 3.07, 3.14

Copyright ©2000 - 2021, Jelsoft Enterprises Ltd.

This forum has received and complied with 0 (zero) government requests for information.

Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation.
A copy of the license is included in the FAQ.