GFN factoring with mmff-gfn | Reservations - Page 3

Batalov · 2012-10-23, 01:49

It is basically log₂(k)+n rounded up.

firejuggler · 2012-10-23, 01:56

so thats around 1.02 % for n=25, and between 2.85% & 2.66% for 26-31?
gaaa... thats not large

Batalov · 2012-10-23, 02:03

Not quite 100%, agreed. But they (sort of) sum up, and with patience...
...
...
PROFIT!

________
*They don't really sum up; if you had two runs with p₁ and p₂, then
p = 1 - (1-p₁)(1-p₂) = p₁+p₂ - p₁p₂ (but the last part is tiny)

if you had five runs with p₁...p₅, then
p = 1 - (1-p₁)(1-p₂)(1-p₃)(1-p₄)(1-p₅)
and after many, many runs you will level at 100% ;-)

LaurV · 2012-11-05, 13:07

Ok. Got the v26, tested, pass all (5) tests, now what?
It is a bit difficult to follow with reservations, the fronts seems to move very fast, I have no idea what work should I do. Where should I go to see what is wanted, what is available?

@SB: can you assign me ("the most wanted" hehe) 1-2 (one-two) weeks worth of work on one GTX580? (about, more or less)
or, is any page so clear as doublemersenne page, where I can go and try one assignment? It is quite difficult for me to follow with all the reservations in this thread, especially because I am involved in other things, and I am not so well educted

in gfn's and there are 5 different tails/bases to follow....

edit: ok, i hope i got it right:
reserving gfn3: N=25-31 | k from 500e12
I see this is not so much claimed. I have no idea "how high" I can go, because I have no idea how fast it is. I will just fill my worktodo with lines and let it crawl.

firejuggler · 2012-11-05, 13:34

for this range, on a 560 it goes around one hour for 20e12so assuming double the speed of that 40e12 for one hour, (40*24*7)/7 you should be able to get them by 1500e12

Batalov · 2012-11-05, 18:07

Quote:

Originally Posted by LaurV

edit: ok, i hope i got it right:
reserving gfn3: N=25-31 | k from 500e12
I see this is not so much claimed. I have no idea "how high" I can go, because I have no idea how fast it is. I will just fill my worktodo with lines and let it crawl.

You got it right. For each N, there's a space from the listed k up to the sky. We don't want to have holes.
I remind that the probability of success per unit of time is ~ O(1/kN²).

rajula · 2015-01-18, 09:45

I just remembered that I am still missing that gfn6-factor. Thus, before devoting the GPU to finding Fermat factors, I'll continue a bit the search for gfn6. To start, I reserve

gfn6: N=50-69, k=260e12-300e12

Batalov · 2015-01-18, 19:40

Eggcellent! But wait - you may have it at higher n's (as a byproduct of Fermat Factor / GFN searches), no? It still counts! ;-)

rajula · 2015-01-19, 07:49

Quote:

Originally Posted by Batalov

Eggcellent! But wait - you may have it at higher n's (as a byproduct of Fermat Factor / GFN searches), no? It still counts! ;-)

Unfortunately the overall counts are only
GFN3: 2
GFN5: 1
GFN6: 0
GFN10: 1
GFN12: 2
so the search continues.

MattcAnderson · 2015-12-31, 11:18

Hi All

I just so happen to be a hungry for factors
I have three computers in my living place.
One of the is the upstairs computer, okay there is a second upstairs computer because I have a tenant. .... And for the last few weeks he owes me money.

My wife's computer, which I am right now typing on has a black and white cover. I use this computer for the "folding at home" BONIC run by Stanford University. because Parkinson's disease is a bad thing.

I don't really use her computer for Prime95 anymore.

My Black computer is cusum ordered from Blue Dragon Computer here in Salem Oregon USA. It required a second trip to the computer store, but I like that place. It is better than the Geek Squad Guys who work for Best Buy here in ... USA. I prefer to avoid chains whenever possible.

Having typed all of that.

Go Team Go

Batalov · 2016-11-03, 17:00

Awww... We can now find so many previously unknown GFN₇ and GFN₁₁'s!
I wasn't coding mmff-gfn for them because of lack of the public interest, but now... just wait and see...

I'll post the code, and perhaps the binaries can be built by the usual suspects and I will create new reservation pages for GFN₇ and GFN₁₁'s!

2012-11-05, 13:07	#26
LaurV Romulan Interpreter Jun 2011 Thailand 21651₈ Posts	Ok. Got the v26, tested, pass all (5) tests, now what? It is a bit difficult to follow with reservations, the fronts seems to move very fast, I have no idea what work should I do. Where should I go to see what is wanted, what is available? @SB: can you assign me ("the most wanted" hehe) 1-2 (one-two) weeks worth of work on one GTX580? (about, more or less) or, is any page so clear as doublemersenne page, where I can go and try one assignment? It is quite difficult for me to follow with all the reservations in this thread, especially because I am involved in other things, and I am not so well educted in gfn's and there are 5 different tails/bases to follow.... edit: ok, i hope i got it right: reserving gfn3: N=25-31 \| k from 500e12 I see this is not so much claimed. I have no idea "how high" I can go, because I have no idea how fast it is. I will just fill my worktodo with lines and let it crawl. Last fiddled with by LaurV on 2012-11-05 at 13:22

2012-11-05, 13:34	#27
firejuggler Apr 2010 Over the rainbow 5×503 Posts	for this range, on a 560 it goes around one hour for 20e12so assuming double the speed of that 40e12 for one hour, (40247)/7 you should be able to get them by 1500e12 Last fiddled with by firejuggler on 2012-11-05 at 13:53

2015-12-31, 11:18	#32
MattcAnderson "Matthew Anderson" Dec 2010 Oregon, USA 27A₁₆ Posts	Hi All I just so happen to be a hungry for factors I have three computers in my living place. One of the is the upstairs computer, okay there is a second upstairs computer because I have a tenant. .... And for the last few weeks he owes me money. My wife's computer, which I am right now typing on has a black and white cover. I use this computer for the "folding at home" BONIC run by Stanford University. because Parkinson's disease is a bad thing. I don't really use her computer for Prime95 anymore. My Black computer is cusum ordered from Blue Dragon Computer here in Salem Oregon USA. It required a second trip to the computer store, but I like that place. It is better than the Geek Squad Guys who work for Best Buy here in ... USA. I prefer to avoid chains whenever possible. Having typed all of that. Go Team Go Last fiddled with by MattcAnderson on 2015-12-31 at 11:18 Reason: should add quotes . ..

2016-11-03, 17:00	#33
Batalov "Serge" Mar 2008 Phi(4,2^7658614+1)/2 22051₈ Posts	Awww... We can now find so many previously unknown GFN₇ and GFN₁₁'s! I wasn't coding mmff-gfn for them because of lack of the public interest, but now... just wait and see... I'll post the code, and perhaps the binaries can be built by the usual suspects and I will create new reservation pages for GFN₇ and GFN₁₁'s! Attached Thumbnails

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
GFN factoring with mmff-gfn \| Results	Batalov	Operazione Doppi Mersennes	126	2016-11-12 12:22
mmff checkpoints	houding	Factoring	1	2015-09-02 19:32
mmff parameters	MattcAnderson	Operazione Doppi Mersennes	2	2015-07-08 15:28
Output bug in mmff	ATH	Operazione Doppi Mersennes	8	2012-09-18 15:57
P minus 1 factoring - information and reservations	philmoore	Five or Bust - The Dual Sierpinski Problem	1	2010-04-07 22:25

2012-10-23, 01:49	#23
Batalov "Serge" Mar 2008 Phi(4,2^7658614+1)/2 9,257 Posts	It is basically log₂(k)+n rounded up.

2012-10-23, 01:56	#24
firejuggler Apr 2010 Over the rainbow 100111010011₂ Posts	so thats around 1.02 % for n=25, and between 2.85% & 2.66% for 26-31? gaaa... thats not large

2012-10-23, 02:03	#25
Batalov "Serge" Mar 2008 Phi(4,2^7658614+1)/2 22051₈ Posts	Not quite 100%, agreed. But they (sort of) sum up, and with patience... ... ... PROFIT! ________ *They don't really sum up; if you had two runs with p₁ and p₂, then p = 1 - (1-p₁)(1-p₂) = p₁+p₂ - p₁p₂ (but the last part is tiny) if you had five runs with p₁...p₅, then p = 1 - (1-p₁)(1-p₂)(1-p₃)(1-p₄)(1-p₅) and after many, many runs you will level at 100% ;-)

2015-01-18, 09:45	#29
rajula "Tapio Rajala" Feb 2010 Finland 100111011₂ Posts	I just remembered that I am still missing that gfn6-factor. Thus, before devoting the GPU to finding Fermat factors, I'll continue a bit the search for gfn6. To start, I reserve gfn6: N=50-69, k=260e12-300e12

2015-01-18, 19:40	#30
Batalov "Serge" Mar 2008 Phi(4,2^7658614+1)/2 9,257 Posts	Eggcellent! But wait - you may have it at higher n's (as a byproduct of Fermat Factor / GFN searches), no? It still counts! ;-)