Join PrimeGrid
Returning Participants
Community
Leader Boards
Results
Other
drummers-lowrise
|
Message boards :
General discussion :
Intersection of primes
| Author |
Message |
|---|
|
|
|
Hi all,
I returned after near two years pause - and with few "philosophical" questions. I'll post them in different threads...
1. One of them already found in the thread "some basic questions" - it is difficult to track found prime to workunit...
2. Next - it is a public project - so it would allow to download all found primes. Is that possible?
3. It could happen that a prime in one project is also a prime from another project... Are these some of such ones?
Tnx for your opinions....
-js- | |
|
 Michael Goetz Volunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13520
ID: 53948
Credit: 241,958,166
RAC: 322,156
                          
|
Hi all,
I returned after near two years pause - and with few "philosophical" questions. I'll post them in different threads...
1. One of them already found in the thread "some basic questions" - it is difficult to track found prime to workunit...
2. Next - it is a public project - so it would allow to download all found primes. Is that possible?
3. It could happen that a prime in one project is also a prime from another project... Are these some of such ones?
Tnx for your opinions....
-js-
1. Yes, it's difficult, in part because the workunits are ephemeral and disappear in time. If you want to know which WU found the prime, the easiest thing to do is... ask. :)
2. There's a new primes page (see the menu on the left). Currently, it lists only the first 1000 primes of each type, but I intend to remove that restriction eventually. I don't see any reason why it shouldn't be public.
3. Absolutely, positively, there ARE duplicates. Coincidentally a few days ago I checked and there's about 2000 primes in the database that are in there twice.
Primes found at PrimeGrid have been previously found elsewhere, and primes found elsewhere have been previously found here. Primes found by PrimeGrid on BOINC have also been found by PrimeGrid on PRPNet. Primes found by PrimeGrid on one PRPNet port have been found by PrimeGrid on other PRPNet ports.
There's been a lot of different prime searches, and there's obviously been a fair amount of overlap.
____________
My lucky number is 75898524288+1 | |
|
|
|
|
Tnx, Michael - good explanation...
I had thought that it would be good idea to have some API to access these data directly. As all data are stored in DB, that could not be very difficult thing to prepare such interface...
Then interested people could download these primes directly and perform some extra analysis of them...
There was misunderstanding about p.3 (sorry, maybe in hurry I expressed not very clearly)...
I had in my mind -- the prime of one form (for example, Cullen) could be prime found in another search (for example, Proth or Sophie)...
But going to your answer ("here's been a lot of different prime searches") - maybe somewhere is accumulated (common/integrated) DB of them?
Regards, -js-
| |
|
|
|
3. It could happen that a prime in one project is also a prime from another project... Are these some of such ones?
Right now I see on u-g-f.de: pending tests for project GFN524288 the candidate 910116^524288+1 being assigned to 288larsson.
The problem with 910116 is that it is a perfect square. Therefore:
910116^524288+1 = 954^1048576+1
so that prime really belongs to GFN1048576, not GFN524288.
In general we should exclude from GFN searches all b values that are perfect powers. For if they are squares, they belong to another GFN series and have most certainly already been checked there. And if they are non-square perfect powers, say b=a^{2j-1}, it is easy to show that they have an "algebraic" factorization, hence are composite.
/JeppeSN
| |
|
|
|
Right now I see on u-g-f.de: pending tests for project GFN524288 the candidate 910116^524288+1 being assigned to 288larsson.
Same work unit is now assigned to zzcmeyer.
/JeppeSN | |
|
|
|
|
Now the work unit 913936^524288+1 is assigned to zzcmeyer on PRPNet port 11001. 913936 is the next even square after 910116 (see above), and:
GF(19, 913936) = 913936^524288+1 = 956^1048576+1 = GF(20, 956)
This number is known to be composite.
/JeppeSN | |
|
JimBHonorary cruncher
Send message
Joined: 4 Aug 11
Posts: 916
ID: 107307
Credit: 974,494,172
RAC: 911
        
|
Now the work unit 913936^524288+1 is assigned to zzcmeyer on PRPNet port 11001. 913936 is the next even square after 910116 (see above), and:
GF(19, 913936) = 913936^524288+1 = 956^1048576+1 = GF(20, 956)
This number is known to be composite.
/JeppeSN
I have no intention of pulling it. Doing that with PRPNet is just asking for trouble. I'd rather that test ran and the server didn't get screwed up as opposed to the other way around. | |
|
|
|
|
Now 917764^524288+1 is assigned to zzcmeyer, and as you guessed 917764=958^2. I conclude that PRPNet does not skip b that are square numbers. I shall not report any more instances of it.
Do you know if Generalized Fermat Prime Search on BOINC ("short" and "WR") also includes square numbers? In particular, can anyone see if GF(22, 10000) = 10000^{2^22}+1 and GF(22, 20736) = 20736^{2^22}+1 have been tested in the World Record subproject (see other thread for motivation)?
/JeppeSN | |
|
Message boards :
General discussion :
Intersection of primes |
