# Work down on the Upper bound of the Twin Primes [duplicate]

This question already has an answer here:

It can be shown using the Selberg Sieve method, that the maximum number of Twin primes less than $N$ is $$\frac{CN}{\ln^2(N)}$$ does anyone know if there has been any work done on finding an upper bound for the constant $C$?

## marked as duplicate by Gerry Myerson, Andrés E. Caicedo, Ramiro de la Vega, Daniel Moskovich, Andrey RekaloJul 9 '13 at 5:33

This question has been asked before and already has an answer. If those answers do not fully address your question, please ask a new question.

## 1 Answer

It is conjectured that the number of twin primes less than $N$ is $(\mathfrak{S}+o(1))N/(\log N)^2$, where $$\mathfrak{S}=2\prod_{p>2}(1-(p-1)^{-2})$$ is the so-called twin-prime constant. Using the large sieve it is easy to show that the number of twin primes less than $N$ is at most $(8\mathfrak{S}+o(1))N/(\log N)^2$. According to page 76 of Tenenbaum's Introduction to analytic and probabilistic number theory, the best result in this direction is by Wu (1990) which replaces 8 by 3.418.

EDIT: According to MathSciNet, Wu (2004) improved 3.418 to 3.3996.

EDIT: The constant 8 also follows from the Selberg sieve, see page 65 in Greaves' Sieves in number theory.

• And of course the result (for some $C$) is due to V. Brun. – Denis Chaperon de Lauzières Mar 15 '11 at 15:47
• I think the Brun sieve misses a power of $\log\log N$. – GH from MO Mar 15 '11 at 15:54
• Only the first version of Brun's sieve has this $\log \log N$; he improved it later to get the right order of magnitude. – Denis Chaperon de Lauzières Mar 15 '11 at 17:35
• Great, so Brun was the first to get the right order of magnitude. Do you know a reference? – GH from MO Mar 15 '11 at 19:13
• gallica.bnf.fr/ark:/12148/bpt6k3121p/… – Denis Chaperon de Lauzières Mar 15 '11 at 20:26