Help test 210885 - Find a new top 5000 prime! - Page 10

Kosmaj · 2007-03-26, 22:44

Nuggetprime

If you want to process them all there are no problems, I can send you the whole file, but as Curtis said can you begin with 560-570 just to check exe times on your machines and have a better idea what kind of numbers you are going to deal with.

FYI, in the 560-1M range there are 51154 remaining candidates, or about 116.25 per 1000 n. Here is the list of FFT jumps in LLR. The next one is cominng almost at once at n=560828.

Code:

   fftlen       nmax
--------------------
    49152     560828
    57344     650886
    65536     745444
    81920     923559
    98304    1096675

Kosmaj · 2007-03-27, 01:55

I just added up files to 600k, in 5k chunks.

nuggetprime · 2007-03-27, 15:05

I will start testing when I found a prime in the 5th Drive or it goes over 700k.

gd_barnes · 2007-05-05, 10:02

I'm taking 560 to 562K as my first hack at this. Please let me know if that range has already been done.

G

lsoule · 2007-05-07, 04:14

560-562 is yours. Welcome to the project!

gd_barnes · 2007-05-13, 08:26

I have completed testing on the range of 560K to 562K for k=210885. There were no primes.

Gary

Cruelty · 2007-06-25, 19:31

I'd like to take this k to n=750000. Can someone mail me an input file?

<my nick>[at]poczta[dot]onet[dot]pl

OK, I've changed my mind

I would like to sieve it first from existing 5.5T till 10T and then start testing - in such case I would go with entire 562000<n<1M range for sieving purposes and then LLR it till n=750000.

Kosmaj · 2007-06-25, 20:13

YGM!

There are 50949 candidates remaining (562-1000k) .

Cruelty · 2007-06-25, 20:50

package delivered

for some unknown reason KIS has marked it as a SPAM

Cruelty · 2007-07-24, 09:52

I have tested this k till n=600000 and will not test it further. Additionally I have mailed Kosmaj input file for 600000<n<1M sieved till p=10T. This sieve depth should be sufficient till n~750000 when FFT lenght will change from 64k to 80k.

gd_barnes · 2007-07-24, 15:11

Quote:

Originally Posted by Cruelty

I have tested this k till n=600000 and will not test it further. Additionally I have mailed Kosmaj input file for 600000<n<1M sieved till p=10T. This sieve depth should be sufficient till n~750000 when FFT lenght will change from 64k to 80k.

Cruelty,

Is there a 'quick and easy' way to tell or calculate at what value of n that the FFT length will change? Or do you just have to manually test it? Or is there a table somewhere that shows approximately when it will change for various sizes of k's?

Also, can you or someone tell me how the FFT length is calculated by the primality checking programs in the first place? I'm assuming it is based on the true length of the number being tested.

Thanks for any help that you can give.

Gary

2007-03-26, 22:44	#100
Kosmaj Nov 2003 111000100110₂ Posts	Nuggetprime If you want to process them all there are no problems, I can send you the whole file, but as Curtis said can you begin with 560-570 just to check exe times on your machines and have a better idea what kind of numbers you are going to deal with. FYI, in the 560-1M range there are 51154 remaining candidates, or about 116.25 per 1000 n. Here is the list of FFT jumps in LLR. The next one is cominng almost at once at n=560828. Code: fftlen nmax -------------------- 49152 560828 57344 650886 65536 745444 81920 923559 98304 1096675

2007-05-05, 10:02	#103
gd_barnes May 2007 Kansas; USA 2²·19·137 Posts	Taking 560 to 562K I'm taking 560 to 562K as my first hack at this. Please let me know if that range has already been done. G

2007-05-13, 08:26	#105
gd_barnes May 2007 Kansas; USA 2²×19×137 Posts	560 to 562 completed I have completed testing on the range of 560K to 562K for k=210885. There were no primes. Gary

2007-06-25, 19:31	#106
Cruelty May 2005 658₁₆ Posts	I'd like to take this k to n=750000. Can someone mail me an input file? <my nick>[at]poczta[dot]onet[dot]pl OK, I've changed my mind I would like to sieve it first from existing 5.5T till 10T and then start testing - in such case I would go with entire 562000<n<1M range for sieving purposes and then LLR it till n=750000. Last fiddled with by Cruelty on 2007-06-25 at 19:54

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Choose your own K and work on finding a top-5000 prime!	lsoule	Riesel Prime Search	421	2021-07-27 23:51
Test a Specific Exponent in Prime95. When it is 100% done, does it mean I find a new prime number?	king	Information & Answers	5	2018-02-21 18:15
Help test 2995125705 - Find a new top 5000 prime!	SlashDude	Riesel Prime Search	538	2007-05-08 01:42
The fastest way to a top-5000 prime?	lsoule	15k Search	13	2005-09-19 20:24
Beta user stats for Phase 2 - Test of 210885	SlashDude	15k Search	1	2003-12-16 23:09

2007-03-27, 01:55	#101
Kosmaj Nov 2003 2·1,811 Posts	I just added up files to 600k, in 5k chunks.

2007-03-27, 15:05	#102
nuggetprime Mar 2007 Austria 100101110₂ Posts	I will start testing when I found a prime in the 5th Drive or it goes over 700k.

2007-05-07, 04:14	#104
lsoule Nov 2004 California 2³·3·71 Posts	560-562 is yours. Welcome to the project!

2007-06-25, 20:13	#107
Kosmaj Nov 2003 E26₁₆ Posts	YGM! There are 50949 candidates remaining (562-1000k) .

2007-06-25, 20:50	#108
Cruelty May 2005 11001011000₂ Posts	package delivered for some unknown reason KIS has marked it as a SPAM

2007-07-24, 09:52	#109
Cruelty May 2005 2³·7·29 Posts	I have tested this k till n=600000 and will not test it further. Additionally I have mailed Kosmaj input file for 600000<n<1M sieved till p=10T. This sieve depth should be sufficient till n~750000 when FFT lenght will change from 64k to 80k.