What if we don't find twin prime n=333333?

cipher · 2009-04-14, 11:45

What is the probability that we will find twin prime for n=333333 before 100G is up?

What if we don't find a twin prime before 100G? do we start sieving 100G to 200G? or do we abandon n=333333 for a higher n=Value?

Thanks

Svenie25 · 2009-04-14, 14:20

Over at primegrid Lennart said, that they will stop LLR until the sieve is at 12P. I´ve seen for myself, that sieving needs atm ~45 seconds for one k and a LLRtest takes 2:40 minutes.
I think that nobody expected, that we will need so long to find this little twin. One of the problems ist, that primes you find here (no twin) will not get into the top5000. And so not many people are interested. But on the other hand, in the last few days were more than 2 times more LLRtest were crunched at Primegrid.
And for myself I would like to see the end of this project. There must be a twin! ;)
Hopefully it is not to far in the future...

MooooMoo · 2009-04-15, 15:28

Quote:

Originally Posted by cipher

What is the probability that we will find twin prime for n=333333 before 100G is up?

What if we don't find a twin prime before 100G? do we start sieving 100G to 200G? or do we abandon n=333333 for a higher n=Value?

Thanks

The probability that we'll find a twin between now (60G) and 100G is about 60-65%.

There are no plans for sieving 100G-200G, so it looks like we'll abandon n=333333 for a higher n-value.

Svenie25 · 2009-04-15, 15:35

Oh, doesn´t sound good.

But we will see. I´m switching now back to sieving until at least sunday. should be the next n? n=666666?

geoff · 2009-04-16, 02:48

It should be possible to make a program that sieves x different (carefully chosen) exponents at once at only a small extra cost in time (but requiring x times more memory).

Minimising the number of 1 bits in the binary expansion of the exponent should also reduce sieve time (i.e. an exponent which is a power of two would be fastest).

E.g. sieving n=2^19+1=524289 and n=2^19+2^18+1=786433 together should be almost as fast as sieving n=524289 alone.

Joshua2 · 2009-04-16, 21:53

A 64 bit sieve would be good too. But it would be good to have a n that finds top-5000 primes again.

2009-04-14, 11:45	#1
cipher Feb 2007 211 Posts	What if we don't find twin prime n=333333? What is the probability that we will find twin prime for n=333333 before 100G is up? What if we don't find a twin prime before 100G? do we start sieving 100G to 200G? or do we abandon n=333333 for a higher n=Value? Thanks

2009-04-15, 15:35	#4
Svenie25 Aug 2008 Good old Germany 3×47 Posts	Oh, doesn´t sound good. But we will see. I´m switching now back to sieving until at least sunday. should be the next n? n=666666? Last fiddled with by Svenie25 on 2009-04-15 at 15:43

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Highest Prime is also a twin prime... NOT	hydeer	Lone Mersenne Hunters	9	2018-04-03 22:54
Twin Prime Days, Prime Day Clusters	cuBerBruce	Puzzles	3	2014-12-01 18:15
n=333333 Twin Found!	Kosmaj	Twin Prime Search	12	2009-08-14 13:10
After we find a twin, our next exponent will be...	MooooMoo	Twin Prime Search	49	2006-12-31 21:52
If you find a twin prime...	MooooMoo	Twin Prime Search	2	2006-05-11 23:38

2009-04-14, 14:20	#2
Svenie25 Aug 2008 Good old Germany 3×47 Posts	Over at primegrid Lennart said, that they will stop LLR until the sieve is at 12P. I´ve seen for myself, that sieving needs atm ~45 seconds for one k and a LLRtest takes 2:40 minutes. I think that nobody expected, that we will need so long to find this little twin. One of the problems ist, that primes you find here (no twin) will not get into the top5000. And so not many people are interested. But on the other hand, in the last few days were more than 2 times more LLRtest were crunched at Primegrid. And for myself I would like to see the end of this project. There must be a twin! ;) Hopefully it is not to far in the future...

2009-04-16, 02:48	#5
geoff Mar 2003 New Zealand 1157₁₀ Posts	It should be possible to make a program that sieves x different (carefully chosen) exponents at once at only a small extra cost in time (but requiring x times more memory). Minimising the number of 1 bits in the binary expansion of the exponent should also reduce sieve time (i.e. an exponent which is a power of two would be fastest). E.g. sieving n=2^19+1=524289 and n=2^19+2^18+1=786433 together should be almost as fast as sieving n=524289 alone.

2009-04-16, 21:53	#6
Joshua2 Sep 2004 13×41 Posts	A 64 bit sieve would be good too. But it would be good to have a n that finds top-5000 primes again.