mersenneforum.org (https://www.mersenneforum.org/index.php)
-   15k Search (https://www.mersenneforum.org/forumdisplay.php?f=16)
-   -   K=13236795 Status Post (https://www.mersenneforum.org/showthread.php?t=3705)

 ValerieVonck 2005-01-21 05:31

I will sieve from 164000 to 220000. Is this ok?

 Kosmaj 2005-01-22 05:32

Better go at least to 340k, that's one of the "milestones" where LLR exe times (and FFT lengths used) become longer. If you want to find a 100k-digit prime then sieve to 425k. Also note that the bottom limit of reportable primes is currently at about 180k.

 ValerieVonck 2005-01-22 13:04

[QUOTE=Kosmaj]Better go at least to 340k, that's one of the "milestones" where LLR exe times (and FFT lengths used) become longer. If you want to find a 100k-digit prime then sieve to 425k. Also note that the bottom limit of reportable primes is currently at about 180k.[/QUOTE]

Kosmaj,

I am currently testing the following ranges:

164K to 220K => sieved + llring
220K to 500K => currently sieving

Are there "better" (faster) sieving programs then NewPgen?

 ValerieVonck 2005-01-29 07:07

164K -> 220K 45% complete - 1 prime found (50.000 digits)

 Jwb52z 2005-01-29 17:22

Cedric, if you could list the prime itself, it would be nice.

 ValerieVonck 2005-02-03 08:29

[QUOTE=Jwb52z]Cedric, if you could list the prime itself, it would be nice.[/QUOTE]

13236795*2^167188-1 50000+ digits
13236795*2^199020-1 59919 digits -> on the primepages
13236795*2^199516-1 60068 digits -> on the primepages

I will show this time the primes I have found.
From now on, I only list the primes I have found on the "Post your primes you have found" pages

Currently resieving the range 220000 through 500000 with Proth 0.42
Therefore I created my own Riesel.dat file

[code]
1
220000
500000
k=13236795
<factors of input file of NewPGen>
...
[/code]

I then stripped the CR of this file

Regards

 Jwb52z 2005-02-03 09:29

Thank you Cedric, but you're right. They should be posted there. I forgot about cross posting.

 ValerieVonck 2005-02-03 20:23

No problem!

 ValerieVonck 2005-02-08 18:00

13236795 done to 220k. 4 primes found
Now currently sieving to 500k.
I will take this K to 1 million. :coffee:

 ValerieVonck 2005-02-13 17:47

K=13236795 Status Post

For not soiling the "I'd like to help out" thread,
I open a seperate thread wich will contain info about my progress of this K.

Moderators, if it is possible please move all the reply's in the "I'd like to help out" thread to here.

Update
[code]
: stopped sieving of 13236795 of the range 220k => 500k
: started llr'ing this range => 18.000 candidates left (of the 280K potentials)
: continuing sieving of the 500K => 1M range
[/code]

Primes found so far:

[code]
13236795*2^167188-1 is prime! 50000+ digits
13236795*2^199020-1 is prime! 59919 digits
13236795*2^199516-1 is prime! 60068 digits
13236795*2^215582-1 is prime! 64904 digits
[/code]

 ValerieVonck 2005-02-15 17:49

Currently resieving the range 231k => 1 Million.
This range spans approximately 101.000 tests.
Did LLR from 220k => 231 (744 tests).
So llr'ing _all_ these test would take a long time.
So lets say 1000 test each 2-3 days * 100 = 200 ~ 300 days of crunch time (12 of the 24 hours) :w00t: :w00t: :w00t:

So hopefully with sieving this range to 1 T or more?? I'll get this range to about 50.000 (or less) potential candidates.

Regards,

 lsoule 2005-02-15 20:36

For comparison, I've sieved 515106735 for k=460k-1M
up to 2.5T and still have 78,000 candidates.

-Larry

 ValerieVonck 2005-02-16 07:11

[QUOTE=lsoule]For comparison, I've sieved 515106735 for k=460k-1M
up to 2.5T and still have 78,000 candidates.

-Larry[/QUOTE]

Ouch! :devil: :devil: :alien:

 Kosmaj 2005-02-17 07:56

[QUOTE=CedricVonck]Currently resieving the range 231k => 1 Million.
[/QUOTE]
What do you mean by "resieving" ? Sieving again against the same possible factors? Why?? I'm not aware that any publicly availabe siever (NewPGen, ksieve) is missing composites. If you found otherwise please provide examples.
[QUOTE]
This range spans approximately 101.000 tests.
...
So hopefully with sieving this range to 1 T or more?? I'll get this range to about 50.000 (or less) potential candidates.
[/QUOTE]
I don't know how deep have you sieved those 101,000 candidates but even if you sieved only to 1 bn you will soon realize that a reduction to 50,000 is impossible...

Another possibility is to stop tests early. Indeed, there is nothing special about this k. No need to marry the first love :smile: :shock: :wink:

 ValerieVonck 2005-02-17 17:06

[QUOTE=Kosmaj]What do you mean by "resieving" ? Sieving again against the same possible factors? Why?? I'm not aware that any publicly availabe siever (NewPGen, ksieve) is missing composites. If you found otherwise please provide examples.

I don't know how deep have you sieved those 101,000 candidates but even if you sieved only to 1 bn you will soon realize that a reduction to 50,000 is impossible...

Another possibility is to stop tests early. Indeed, there is nothing special about this k. No need to marry the first love :smile: :shock: :wink:[/QUOTE]

LOL :rolleyes: :rolleyes:
With resieve I mean: I have put 2 ranges together, to be absolute sure.
I almost restart the sieve from scratch.

Other point, per day I am eliminating approx 1900 factors.
I'll keep this up until the find a factor rate will drop :geek: :geek:
I am currently dividing the workload between 2 machines.

The last point: I know that there are other K's to be searched but I really want to find some big primes.

Ps. down to 98000+ tests.
3000 factors eliminated.

 lsoule 2005-02-17 19:04

Sieving seems to be much more effective on my k=231. Sieving the
range n=400k-1M only up to 137B leaves just 49,000 tests.

 ValerieVonck 2005-02-17 20:46

Isoule,

Mine is P4 3 Ghz: 648 k
P4 2.4Ghz: 484 k

 lsoule 2005-02-18 00:48

For k=231, it looks like 629k/s on a P4 2.2GHz

 ValerieVonck 2005-02-18 07:18

OK thank you.

 ValerieVonck 2005-02-20 11:30

Currently testing the range 231K => 240k : 576 candidates on my P4 3G.
Continuing sieving the range 500G => 1T on my P4 2.4G.

 ValerieVonck 2005-02-21 17:05

Range 230K => 240K done. No primes found
Range 240K => 250K currently testing - 674 candidates
Continuing sieving with lower bound 250k.
95886 candidates remaining :geek: :geek: :coffee: :coffee:

 ValerieVonck 2005-02-23 16:52

Range 240K => 250K done. No primes found
Range 250K => 280K currently testing - 1997 candidates
Continuing sieving with lower n bound 280k.

 ValerieVonck 2005-03-02 18:43

Crunched to 279021 => still no prime found.
Added 1295 tests to the input file. To 300K :showoff:
After my lmh range is done, I will install a server on my main pc to service my local lan.

 ValerieVonck 2005-03-08 19:30

Small update:
Approximately 140 tests to 300.000 to go.
Total tests done: 14.261
From n = 164.000 to n = 297.759
Number of tests to 1M = 92.059 :showoff: :showoff: (I will do a deeper sieve of this..)
Number of Primes found: 4

 ValerieVonck 2005-03-11 07:22

[QUOTE=CedricVonck]Small update:
Approximately 140 tests to 300.000 to go.
Total tests done: 14.261
From n = 164.000 to n = 297.759
Number of tests to 1M = 92.059 :showoff: :showoff: (I will do a deeper sieve of this..)
Number of Primes found: 4[/QUOTE]

Project is currently on hold because my pc is in the shop to swap the dvd rom drive.

I am also testing the range 1 -> 164K.

 ValerieVonck 2005-03-13 09:57

Added 650 tests to the queue.

91080 test remaining.

 ValerieVonck 2005-03-16 17:18

Added another 450 test to the queue.
90606 factors remaining

 ValerieVonck 2005-03-18 15:33

Added 786 tests to the queue
I currently at 322029

Total tests to do: 89.819

 ValerieVonck 2005-03-23 05:34

88.198 tests remaining

 ValerieVonck 2005-09-14 15:44

No primes to 346638.
Starting resieve to 1T

 ValerieVonck 2005-09-23 07:39

No primes between 340K & 350 (will check this evening)

 ValerieVonck 2005-10-06 17:04

Small update

Small update

k=13236795 no prime to 354708 (from 340k) (going to test to 360k)
k=4075360575 no prime to 208965 (from 200k)
k=673394865 no primes to 153723 (from 130k) (130208 was prime)

 ValerieVonck 2005-10-17 08:10

I will continue work as soon as my work for LMH is done.

Cedric

 ValerieVonck 2005-10-31 22:54

I am back

 All times are UTC. The time now is 22:57.