What are your CRUS plans? - Page 3

gd_barnes · 2011-02-13, 19:41

For a majority of the sieve files, they will need to be sieved further. Obviously Chris's (Flatlander) files are an exception and there is 1-2 exceptions for bases < 32 that I have sieved to an optimal depth. Mathew sends me additional and updated sieve files from time to time by Email. He just sieves then to a specific limit (usually P=500G or 1T) and so likely will need more sieving.

BTW, I said that n=50K-100K would take 3-4 times as long as n=25K-50K. That should be about 6 times as long. A doubleing of the n-value takes 4 times as long per test but the number of tests is twice as large so that's 8 times. Then you have to subtract the k's you've found primes for n=25K-50K. Hence ~6 times as long.

Flatlander · 2011-02-13, 20:01

It's doubtful that I could keep my i7 at 4.4Ghz during the (UK) summer unless I turn to sieving to keep the temps down. So I might do some more in a few months.

(Thinks. Or get a water cooling kit and go faster.

)

edit: Can GPUs sieve this stuff yet?

mdettweiler · 2011-02-13, 20:38

Quote:

Originally Posted by Flatlander

edit: Can GPUs sieve this stuff yet?

No, unfortunately not. The algorithm that ppsieve/tpsieve uses for sieving on GPUs only works for base 2; to sieve other bases (and more specifically, handfuls of specific k's rather than huge swaths of k- and n-ranges which are ppsieve/tpsieve's specialty) you'd need an algorithm more akin to that used by sr2sieve, which is not as easily portable to a GPU. When I talked to ppsieve/tpsieve's author about this, he said that it probably could be done, but that he wasn't planning to do it himself any time soon.

rogue · 2011-06-06, 16:59

If I had the resources I would try to get R51 or R79 to n=25000.

I would like to knock off one or two more single k conjectures (of those that I have reserved). I have a list of 15 conjectures that are on my hit list that I would like to work on.

rogue · 2011-07-14, 20:56

Quote:

Originally Posted by rogue

If I had the resources I would try to get R51 or R79 to n=25000.

I would like to knock off one or two more single k conjectures (of those that I have reserved). I have a list of 15 conjectures that are on my hit list that I would like to work on.

I did some further calculations regarding R51/R79. Way too much work. R79 would take close to 16 months with my resources and R51 would take close to two years.

My aforementioned list will be down to eight in a couple of weeks as R51 and R79 will be taken off of it.

rogue · 2013-07-24, 17:50

Quote:

Originally Posted by rogue

I did some further calculations regarding R51/R79. Way too much work. R79 would take close to 16 months with my resources and R51 would take close to two years.

My aforementioned list will be down to eight in a couple of weeks as R51 and R79 will be taken off of it.

I can't believe it has been two years since the last post in this thread.

R51 was completed to n=25K by Lennart and R79 was completed to n=25K by me.

I'm debating whether or not to complete a bunch of conjectures that are complete to n=2.5K or n=10K to n=25K or to take a bunch of single k conjectures to 200K or if I should tackle something else.

gd_barnes · 2013-07-26, 06:18

1 and 2 k'ers to n=200K would be nice.

rogue · 2013-07-26, 12:58

Quote:

Originally Posted by gd_barnes

1 and 2 k'ers to n=200K would be nice.

Yeah, right. I estimate that if I took all the 1 k conjectures to n=200K, that could be up to a year of work for me. Fortunately they all seem to be sieved, but I don't know if they are sieved deeply enough.

MyDogBuster · 2013-07-26, 14:10

There is really not that many 1kers left that are not tested to n=200K. I see about 45 in that category. Most of those sieve files were done by me and those are all sieved to 5T. Mark, on a Q6600, testing near 200K for a 1ker is around 2.5 hrs per test. I have only put a small dent in the 2kers.

I plan on doing some of the 1ker's but at a slowed down pace.

As for myself, I plan on taking the remaining bases =< 250 that are at 25K to at least 50K. Also the above mentioned 1kers. My S66 will be done shortly so I might start another large ck base to 10K. It's a challenge to keep it all straight, but the way I do it I only use 2 cores.

Another candidate for the next Port 1400 project might be R25 and S25 (the k's not being tested by the Prime Grid SR5 project) to n=200K.

I know its time to stop typing when I get a headache.

Siemelink · 2013-07-26, 20:10

Quote:

Originally Posted by Siemelink

I am taking Prime Riesel Conjectures to n = 200,000. That's going to take me through 2011.

Willem.

I have reached the 200,000. I have upped my stakes to 500,00, that will take me through 2015.

Willem.

rebirther · 2013-11-03, 15:41

@Lennart

It would be good if some of the conjectures can go into primegrid`s PRPnet.

2011-02-13, 19:41	#23
gd_barnes May 2007 Kansas; USA 101·103 Posts	For a majority of the sieve files, they will need to be sieved further. Obviously Chris's (Flatlander) files are an exception and there is 1-2 exceptions for bases < 32 that I have sieved to an optimal depth. Mathew sends me additional and updated sieve files from time to time by Email. He just sieves then to a specific limit (usually P=500G or 1T) and so likely will need more sieving. BTW, I said that n=50K-100K would take 3-4 times as long as n=25K-50K. That should be about 6 times as long. A doubleing of the n-value takes 4 times as long per test but the number of tests is twice as large so that's 8 times. Then you have to subtract the k's you've found primes for n=25K-50K. Hence ~6 times as long. Last fiddled with by gd_barnes on 2011-02-13 at 19:48

2011-02-13, 20:01	#24
Flatlander I quite division it "Chris" Feb 2005 England 31×67 Posts	It's doubtful that I could keep my i7 at 4.4Ghz during the (UK) summer unless I turn to sieving to keep the temps down. So I might do some more in a few months. (Thinks. Or get a water cooling kit and go faster. ) edit: Can GPUs sieve this stuff yet? Last fiddled with by Flatlander on 2011-02-13 at 20:02

2013-07-26, 14:10	#31
MyDogBuster May 2008 Wilmington, DE B24₁₆ Posts	There is really not that many 1kers left that are not tested to n=200K. I see about 45 in that category. Most of those sieve files were done by me and those are all sieved to 5T. Mark, on a Q6600, testing near 200K for a 1ker is around 2.5 hrs per test. I have only put a small dent in the 2kers. I plan on doing some of the 1ker's but at a slowed down pace. As for myself, I plan on taking the remaining bases =< 250 that are at 25K to at least 50K. Also the above mentioned 1kers. My S66 will be done shortly so I might start another large ck base to 10K. It's a challenge to keep it all straight, but the way I do it I only use 2 cores. Another candidate for the next Port 1400 project might be R25 and S25 (the k's not being tested by the Prime Grid SR5 project) to n=200K. I know its time to stop typing when I get a headache. Last fiddled with by MyDogBuster on 2013-07-26 at 14:14 Reason: fix typos

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Plans for the end of the world	Oddball	Lounge	4	2011-04-18 04:06
how high will CRUS go	Mini-Geek	Conjectures 'R Us	1	2010-11-08 20:50
Further Plans	Kosmaj	Riesel Prime Search	6	2009-05-20 01:27
Further Plans	Kosmaj	Riesel Prime Search	6	2006-09-29 22:32
64 bit plans	pyrodave	Software	17	2004-06-05 12:27

2011-06-06, 16:59	#26
rogue "Mark" Apr 2003 Between here and the 2⁴·397 Posts	If I had the resources I would try to get R51 or R79 to n=25000. I would like to knock off one or two more single k conjectures (of those that I have reserved). I have a list of 15 conjectures that are on my hit list that I would like to work on.

2013-07-26, 06:18	#29
gd_barnes May 2007 Kansas; USA 101×103 Posts	1 and 2 k'ers to n=200K would be nice.

2013-11-03, 15:41	#33
rebirther Sep 2011 Germany 2·1,427 Posts	@Lennart It would be good if some of the conjectures can go into primegrid`s PRPnet.