After we find a twin, our next exponent will be... - Page 2

2006-12-14, 07:31

One could go for this record first: http://primes.utm.edu/top20/page.php?id=2

smh · 2006-12-14, 08:42

Quote:

Originally Posted by KEP

Actually I hope not, that he decides to do only LLR search, since I hope to see that we will some day have the biggest sequential prime database (which we according to my knowledge already have), going higher than 1,000,000,000,000,000.

What is the point of having such a database? Calculating a prime in that range (which is way to small for RSA) is probably much faster than doing a DB lookup

Quote:

Originally Posted by MooooMoo

Finding a 100,000 digit twin takes ~8.5 times as much computing power as finding a 58,700 digit twin, so I'm not sure if n=333333 is a good idea.

So what? Finding records isn't easy. Why do people always need results in a few weeks? Projects like SoB and Rieselsieve won't finish in our lifetime.

Quote:

Originally Posted by thommy

One could go for this record first: http://primes.utm.edu/top20/page.php?id=2

Thats what we are doing alreay. The record is 51780 digits. The current search is for ~58700 digits.

2006-12-14, 10:20

Quote:

Originally Posted by smh

Thats what we are doing alreay. The record is 51780 digits. The current search is for ~58700 digits.

But you aren't sieving for Sophie Germain primes.

smh · 2006-12-14, 12:08

Quote:

Originally Posted by thommy

But you aren't sieving for Sophie Germain primes.

Ahh... Sorry, i thought this was the largest twin prime page. The record number is very similar.

Xentar · 2006-12-14, 18:26

Small question:
Is a twin prime possible for every n? Or does this n have to meet some conditions (for example, not being dividable by a specific number)?
So, for example, when you find a twin prime for 195000, couldnt you try to find one for 200000, 205000 and so on?

MooMoo2 · 2006-12-14, 19:43

Quote:

Originally Posted by Xentar

Small question:
Is a twin prime possible for every n? Or does this n have to meet some conditions (for example, not being dividable by a specific number)?

Yes, if the value of k is high enough (you may have to search a large number of k values before getting that twin).

Skligmund · 2006-12-15, 08:15

Quote:

Originally Posted by jmblazek

Looking at the top 20, briefly glancing at their computers, it appears only 2 participants are running TPS right now (17th ranked Skligmund and 19th ranked Virus-X). Congrats to Skligmund for being the TOP TPS producer! However, it's a bit odd that KEP keeps finding all those primes on his single computer.

He hacks. obviously. (just kidding!!!)

There is a lot of luck of draw there, I just so happen to get a lot of ranges without primes downloaded...

As for getting more users at Primegrid, I think it will happen a little on its own. I know I would go for a couple months not looking at the site at all (hectic summer), let alone read the forums. Now that the temperatures are well below freezing 24/7, I kick around on the computer more, that's how I learned about it. I researched a little bit, decided this would be neat to get in on, so here I am!

Other than Primegrid rank (which I've been a part of since 2005), how do you figure I am top for TPS? Just by average work?

jmblazek · 2006-12-15, 09:00

Quote:

Originally Posted by Skligmund

Other than Primegrid rank (which I've been a part of since 2005), how do you figure I am top for TPS? Just by average work?

I guess I should have said top PG TPS producer...and yes, since TPS is so new to PG, I just used the RAC. It's a snapshot estimate.

You're crunching PG TPS faster than anyone else at the moment...however, Virus-X is slowly inching up.

As far as top TPS producer, it looks like eric_v holds those honors. Since PG doesn't keep track of how many M someone finishes (I don't even know if it keeps record of how many numbers someone tests) a new method of determining overall top producer will have to be devised. Although, it seems basic that the person who tests the most numbers (not necessarily M) should be the top producer. So TPS should go back and track how many k were tested by each person (yah, that's going to go over well)...since number of k's varies with each M depending on sieving.

(sieving...well, that's a whole other criteria to mix in with top producer)

Skligmund · 2006-12-15, 09:31

Yeah, I was thinking of that, and since I don't know how all the data is organized, I was hesitant to suggest anything. This could be a can of worms (if it already isn't?). Anyways, if anything should be changed, the sooner, the easier. I guess that is why it is beta eh?

MooMoo2 · 2006-12-18, 20:41

It looks like n=333,333 has got my vote.

Here's why:

In the BOINC thread, biwema estimated that the average amount of work per twin is 73 Pentium 4 CPU years. There are about 50-60 participants now (people who have contibuted enough work to find one prime), and I'm assuming that the average participant has one P4 CPU working on the project. I also expect participation to increase slightly, to around 70-80 members, which means we should find a twin in about a year.

pacionet · 2006-12-19, 09:34

I vote too for 333,333.

It would be better if we decide the NEXT N (5, 10 or 50,.. ?) exponents to test; this way, soon, we can start parallel sieving for these exponents.

View Poll Results: The next exponent should be...
under n=250,000	4	25.00%
between 250,000-300,000	2	12.50%
between 300,000-350,000	10	62.50%
between 350,000-400,000	0	0%
above n=400,000	0	0%
Voters: 16. You may not vote on this poll

2006-12-19, 09:34	#22
pacionet Oct 2005 Italy 3·113 Posts	I vote too for 333,333. It would be better if we decide the NEXT N (5, 10 or 50,.. ?) exponents to test; this way, soon, we can start parallel sieving for these exponents. Last fiddled with by pacionet on 2006-12-19 at 09:35

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
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
Where can I find a Reverse and Add program? I can't find any!	Stargate38	Programming	18	2015-07-10 06:08
What if we don't find twin prime n=333333?	cipher	Twin Prime Search	5	2009-04-16 21:53
If you find a twin prime...	MooMoo2	Twin Prime Search	2	2006-05-11 23:38
TWIN MOS RAM	ET_	Hardware	6	2004-10-21 09:41

2006-12-14, 07:31	#12
thommy 2²·43·53 Posts	One could go for this record first: http://primes.utm.edu/top20/page.php?id=2

2006-12-14, 18:26	#16
Xentar Sep 2006 Germany 2×5×19 Posts	Small question: Is a twin prime possible for every n? Or does this n have to meet some conditions (for example, not being dividable by a specific number)? So, for example, when you find a twin prime for 195000, couldnt you try to find one for 200000, 205000 and so on?

2006-12-15, 09:31	#20
Skligmund Dec 2006 Anchorage, Alaska 2·3·13 Posts	Yeah, I was thinking of that, and since I don't know how all the data is organized, I was hesitant to suggest anything. This could be a can of worms (if it already isn't?). Anyways, if anything should be changed, the sooner, the easier. I guess that is why it is beta eh?

2006-12-18, 20:41	#21
MooMoo2 "Michael Kwok" Mar 2006 1,181 Posts	It looks like n=333,333 has got my vote. Here's why: In the BOINC thread, biwema estimated that the average amount of work per twin is 73 Pentium 4 CPU years. There are about 50-60 participants now (people who have contibuted enough work to find one prime), and I'm assuming that the average participant has one P4 CPU working on the project. I also expect participation to increase slightly, to around 70-80 members, which means we should find a twin in about a year.