Riemann Hypothesis

The Riemann hypothesis is a conjecture in number theory stating that every nontrivial zero of the Riemann zeta function has real part $1/2$. Proposed by Bernhard Riemann in 1859, it connects the behavior of a function of a complex variable with the distribution of prime numbers. It remains an unsolved problem and is one of the Millennium Prize Problems designated by the Clay Mathematics Institute. (dlmf.nist.gov)

Mathematical statement

For a complex number $s=\sigma+it$ with $\sigma>1$, the zeta function is defined by the absolutely convergent series

Through analytic continuation, it extends to the entire complex plane except for a simple pole at $s=1$. The values relevant to the hypothesis generally lie outside the region where this defining series converges, so the continued function—not a direct summation of the original series—is essential. (dlmf.nist.gov)

A zero is a value $s$ for which $\zeta(s)=0$. The zeros fall into two classes:

• Trivial zeros: the negative even integers, $s=-2,-4,-6,\ldots$.
• Nontrivial zeros: infinitely many zeros lying in the critical strip, $0<\operatorname{Re}(s)<1$.

The hypothesis asserts that all zeros in the second class lie on the critical line:

The trivial zeros are excluded from the statement, not exceptions to it. (dlmf.nist.gov)

The nontrivial zeros are symmetric about both the real axis and the critical line. Consequently, a zero $\beta+i\gamma$ off the critical line would be accompanied by reflected zeros, including $1-\beta+i\gamma$. These symmetries explain why the central line is distinguished, but do not establish that every zero lies there: a symmetric collection of points can contain points away from its axis of symmetry. (dlmf.nist.gov)

Connection with prime numbers

The zeta function connects directly to primes through its Euler product:

This identity follows from the unique factorization of positive integers into primes. It expresses the same function both as a sum over positive integers and as a product over primes. The product representation in this form applies only in its convergence region; it cannot simply be substituted at a nontrivial zero. (dlmf.nist.gov)

The prime number theorem gives the leading asymptotic distribution of primes. If $\pi(x)$ denotes the number of primes not exceeding $x$, then

A more accurate smooth approximation is the logarithmic integral, $\operatorname{li}(x)$. The Riemann hypothesis controls the size of the discrepancy between the actual count and this approximation. In particular, it is equivalent to the estimate

Here big-O notation means that the absolute error is eventually bounded by a constant times the displayed expression. (dlmf.nist.gov)

The connection becomes especially clear through the Chebyshev function

where the sum includes all positive powers of primes. Explicit formulas relate $\psi(x)$ to the nontrivial zeros, with contributions involving $x^\rho/\rho$. Since

the imaginary part of a zero determines an oscillation in $\log x$, while its real part determines the scale of that contribution. Locating all zeros at $\beta=1/2$ therefore places the oscillatory terms on a square-root scale, although estimating their combined effect also requires controlling the sum over zeros. (dlmf.nist.gov)

Historical development

Riemann introduced the conjecture in his 1859 memoir on the number of primes below a given magnitude. His approach used complex analysis to relate prime counting to the zeros of the zeta function, establishing the framework in which the problem is still studied. (claymath.org)

In 1896, Jacques Hadamard and Charles-Jean de la Vallée Poussin independently proved the prime number theorem by establishing that the zeta function has no zeros on the line $\operatorname{Re}(s)=1$. This was an important distinction: proving the average distribution of primes required excluding zeros on the boundary of the critical strip, not proving that all interior zeros lie on its middle line. (dlmf.nist.gov)

The Clay Mathematics Institute included the conjecture among its seven Millennium Prize Problems in 2000, allocating a US$1 million prize to its resolution under the institute’s rules. Its official classification remains unsolved. (claymath.org)

Equivalent formulations

An equivalent formulation is a statement whose proof would establish the hypothesis and which, conversely, follows from it. Such reformulations translate the zero-location problem into questions about arithmetic functions, approximation errors, or inequalities.

For example, the hypothesis is equivalent to

for every $\varepsilon>0$. The implied constant may depend on $\varepsilon$. The quantifier matters: establishing the estimate for just one fixed, sufficiently large exponent is not equivalent to the hypothesis. (dlmf.nist.gov)

A formulation due to Jeffrey Lagarias uses only positive integers, their divisors, and elementary functions. Define

Then the Riemann hypothesis is equivalent to

Here $H_n$ is the harmonic number and $\sigma(n)$ is the sum of the positive divisors of $n$. Although the inequality is elementary to state, proving it uniformly for all positive integers retains the full difficulty of the original conjecture. (websites.umich.edu)

Partial results and computational evidence

It is known unconditionally that infinitely many zeros lie on the critical line. Stronger results establish positive proportions: for example, work by Kyle Pratt, Nicolas Robles, Alexandru Zaharescu, and Dirk Zeindler established that more than five-twelfths of the nontrivial zeros lie there, in the asymptotic counting sense. Such a proportion theorem is not a proof that every zero lies on the line. (dlmf.nist.gov)

Computations can also establish rigorous statements over finite regions. Dave Platt and Tim Trudgian reported a verification using interval arithmetic that every nontrivial zero with

has real part $1/2$. This is a bound on the zeros’ imaginary coordinates, not on the sizes of primes and not the number of zeros checked. (arxiv.org)

Finite verification, however extensive, cannot settle a statement about zeros at arbitrarily large heights without an additional argument covering the remaining region. Conversely, one rigorously established nontrivial zero off the critical line would disprove the hypothesis. The difference between evidence and a mathematical proof is therefore a difference in logical coverage, not merely computational precision. (arxiv.org)

Spectral and statistical approaches

The Hilbert–Pólya approach seeks a spectral interpretation of the zeros. Roughly, if their imaginary coordinates could be identified with the spectrum of an appropriately constructed self-adjoint operator, the reality of that spectrum could enforce the required zero locations. The essential difficulty is constructing the operator and proving the exact correspondence, rather than observing a resemblance to a known spectrum. (dlmf.nist.gov)

A related direction concerns random matrix theory. Montgomery’s pair-correlation conjecture predicts that suitably normalized spacings between zeta zeros have statistics resembling those of eigenvalues of large random Hermitian matrices. These connections concern the fine-scale distribution of zeros; they are not, by themselves, proofs of their location on the critical line. (pmc.ncbi.nlm.nih.gov)

Generalizations and analogues

The generalized Riemann hypothesis, commonly abbreviated GRH, extends the same critical-line assertion to other functions, notably Dirichlet $L$-functions. These encode arithmetic information beyond the ordinary prime count, including primes in residue classes. Assuming GRH is therefore a stronger assumption than assuming the classical Riemann hypothesis alone. (claymath.org)

There are also analogues for zeta functions associated with algebraic varieties over finite fields. The corresponding Riemann-hypothesis statements were proved in the development of the Weil conjectures, with André Weil’s work on curves and Pierre Deligne’s work in higher dimensions playing central roles. These results provide a profound geometric analogue, but their proofs do not resolve the classical conjecture for $\zeta(s)$. (arxiv.org)

References

1. Riemann Hypothesis — Clay Mathematics Instituteclaymath.org
2. The Millennium Prize Problems — Clay Mathematics Instituteclaymath.org
3. Unsolved Archives — Clay Mathematics Instituteclaymath.org
4. The Riemann Hypothesis — Enrico Bombiericlaymath.org
5. DLMF §25.2: Definition and Expansionsdlmf.nist.gov
6. DLMF §25.10: Zerosdlmf.nist.gov
7. DLMF §25.16: Mathematical Applicationsdlmf.nist.gov
8. DLMF §25.17: Physical Applicationsdlmf.nist.gov
9. DLMF §27.4: Euler Products and Dirichlet Seriesdlmf.nist.gov
10. DLMF §27.12: Asymptotic Formulas: Primesdlmf.nist.gov
11. An Elementary Problem Equivalent to the Riemann Hypothesiswebsites.umich.edu
12. More than five-twelfths of the zeros of ζ are on the critical linearxiv.org