Future direction of NPLB - Page 2

henryzz · 2008-09-07, 13:54

we could maybe start 1003-1400 earlier so it is at a higher n range

mdettweiler · 2008-09-07, 14:20

Quote:

Originally Posted by IronBits

With my limited knowledge...
Put any range that is made up of many smaller ranges on a server so everyone can just put a core or three on it until it has been exhausted.

Unfortunately with the specific range Gary is talking about above, the k/n pairs take an extremely short time to test, and hence server load will be extremely high. However, yes, you're right, some of the higher parts of the proposed drive would be quite viable to run through LLRnet.

mdettweiler · 2008-09-07, 14:22

Quote:

Originally Posted by Flatlander

I'm wondering if k=1003-1400 would get neglected if it was seperated from k=1400-2000? (Because primes are more likely to have already been found.)

Actually, as henryzz suggested, from what Gary and I had discussed, our plan is to do k=1003-1400 first, then worry about 1400-2000--precisely because 1003-1400 is so fragmented and it desperately needs cleaning up.

IronBits · 2008-09-07, 14:47

How long would 2 cores @ 3Ghz take to clean up 1003-1400 ?

Brucifer · 2008-09-07, 18:53

Quote:

Originally Posted by Anonymous

chop snip 1003-1400 is so fragmented and it desperately needs cleaning up.

Ignorant old fart here. I have been sieving for a long time. One thing I noticed is it is much easier to get people running llrnet than sieving because all the glory seems to be in finding the primes, which sieving doesn't get squat for glory. Another thing I've noticed is that it is really easy to fragment things up and then you have real problems cause people just head off in different directions and quit. The title is "No Prime Left Behind." That doesn't have connotations about top 5,000 primes or anything. It does suggest an effort to not leave primes left unfound. I don't think we should be heading off in too many directions at the same time. Maybe one issue might be to reanalyze the purpose of the rallys. For instance, if 1003-1400 is so fragmented, then do some controlled sieving (not general public), and then use the rallys to llrnet the results of the controlled sieving to take bites out the the fragmented space. It would get that issue taken care of, and it would not detract from everyones main thrust or desire to pursue other stuff. Keep a rally at being something "special" for a specific purpose.

OR.............. to open Pandora's BOX........... devil's advocate here..... much as I hate boinc... Set up a boinc system to work away at the hard stuff that the manual folks don't care to finish off. Like the stuff that takes a long time to run results/tests on. The only thing is that you have to keep a constant supply of work in the boinc pipeline or you turn off the boinc lovers out there in the world. Or just set up the boinc thing for a public sieving effort, running a wrapper around the sr2sieve.

What am I saying mainly? That if you are going to go big then you will have to do it boinc. If you stay manual, then realize that you can't cover all the ranges that you think need to be done, cause you don't have the assets to handle it all at once. And by fragmenting into too many directions here you risk trashing the project. Just an old man's thoughts..............

IronBits · 2008-09-07, 19:00

Do you know where/how to setup a Boinc Server?
I would be interested in doing that...

em99010pepe · 2008-09-07, 19:01

Quote:

Originally Posted by IronBits

Do you know where/how to setup a Boinc Server?
I would be interested in doing that...

http://boinc.berkeley.edu/trac/wiki/ServerIntro

EDIT: We can always contact PrimeGrid for a few hints and help.

Quote:

Originally Posted by IronBits

How long would 2 cores @ 3Ghz take to clean up 1003-1400 ?

A few years. We need 1000 cores...lol...--->>>BOINC. ( I also hate BOINC but it is what we need in here). The only thing I can give is a free space to host the possible BOINC server with free electric energy and permanent Internet connection.

IronBits · 2008-09-07, 19:57

Downloading the vmplayer and debian boinc vm server
Where do we want to go from here? ;)

em99010pepe · 2008-09-07, 20:05

IB, could you wait?
Rytis from PrimeGrid will post here soon for some hints and proposals.

Rytis · 2008-09-07, 20:05

I've been pointed to this thread by em99010pepe, so I'm posting a few thoughts.

* PrimeGrid is running on an Athlon 64 X2 4800+, 4GB RAM, and it barely holds, that's why we're upgrading to a two-server configuration in the nearest future.
* BOINC LLR isn't really suitable for n<200K because of server overhead.

What I can offer is not actual help setting up server (because it simply takes too much time, which I don't have; hundreds if not thousands of hours have been invested into PrimeGrid), but integrating NPLB as a subproject in PrimeGrid (it takes only about an hour for me, because all infrastructure is here). You would provide me LLR work in batches, I'd feed it into BOINC, and then give you results.

IronBits · 2008-09-07, 22:47

That sounds better

2008-09-07, 20:05	#20
em99010pepe Sep 2004 2·5·283 Posts	IB, could you wait? Rytis from PrimeGrid will post here soon for some hints and proposals. Last fiddled with by em99010pepe on 2008-09-07 at 20:05

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Move 19 game direction	MooMoo2	Other Chess Games	8	2016-02-01 17:50
NPLB future direction	gd_barnes	No Prime Left Behind	16	2009-05-13 16:45
Thoughts on future NPLB searches	em99010pepe	No Prime Left Behind	38	2008-12-23 10:02
Future project direction and server needs synopsis	gd_barnes	No Prime Left Behind	6	2008-02-29 01:09
Poll on direction of conjectures effort	gd_barnes	Conjectures 'R Us	2	2007-12-19 18:15

2008-09-07, 13:54	#12
henryzz Just call me Henry "David" Sep 2007 Cambridge (GMT/BST) 2³×3×5×7² Posts	we could maybe start 1003-1400 earlier so it is at a higher n range

2008-09-07, 14:47	#15
IronBits I ♥ BOINC! Oct 2002 Glendale, AZ. (USA) 3·7·53 Posts	How long would 2 cores @ 3Ghz take to clean up 1003-1400 ?

2008-09-07, 19:00	#17
IronBits I ♥ BOINC! Oct 2002 Glendale, AZ. (USA) 1113₁₀ Posts	Do you know where/how to setup a Boinc Server? I would be interested in doing that...

2008-09-07, 19:57	#19
IronBits I ♥ BOINC! Oct 2002 Glendale, AZ. (USA) 2131₈ Posts	Downloading the vmplayer and debian boinc vm server Where do we want to go from here? ;)

2008-09-07, 20:05	#21
Rytis Nov 2006 83₁₀ Posts	I've been pointed to this thread by em99010pepe, so I'm posting a few thoughts. * PrimeGrid is running on an Athlon 64 X2 4800+, 4GB RAM, and it barely holds, that's why we're upgrading to a two-server configuration in the nearest future. * BOINC LLR isn't really suitable for n<200K because of server overhead. What I can offer is not actual help setting up server (because it simply takes too much time, which I don't have; hundreds if not thousands of hours have been invested into PrimeGrid), but integrating NPLB as a subproject in PrimeGrid (it takes only about an hour for me, because all infrastructure is here). You would provide me LLR work in batches, I'd feed it into BOINC, and then give you results.

2008-09-07, 22:47	#22
IronBits I ♥ BOINC! Oct 2002 Glendale, AZ. (USA) 3×7×53 Posts	That sounds better