The Riemann Zeta Function: Zeros, the Critical Strip and the Primes

A sum that shouldn't converge, and does anyway

The Riemann zeta function starts as an infinite sum, ζ(s) = 1 + 1/2ˢ + 1/3ˢ + 1/4ˢ + ..., which converges only for complex numbers s with real part greater than 1. Bernhard Riemann's 1859 insight was to analytically continue this sum to almost the entire complex plane — extend it uniquely, preserving differentiability, to every s except s = 1, where the function has a simple pole (the sum literally diverges to infinity, matching the divergence of the harmonic series at s = 1). Everywhere else, including for negative and complex s where the original sum makes no numerical sense at all, ζ(s) is a well-defined, smooth complex number.

Euler's product: zeta is really about primes

Leonhard Euler had already shown, a century before Riemann, that this sum equals an infinite product over every prime number:

zeta(s) = sum(1/n^s, n=1..inf) = product(1/(1 - p^-s), p prime)
// the sum ranges over all positive integers; the product ranges only over primes —
// they're equal because every integer factors uniquely into primes (unique factorisation)

This identity is the whole reason number theorists care about zeta at all: it's a bridge between the additive world of sums over all integers and the multiplicative world of primes, and it exists precisely because of unique factorisation. Any question about how the primes are distributed can, at least in principle, be translated into a question about the analytic behaviour of ζ(s) — and that translation turns out to be extraordinarily productive.

Trivial zeros and the critical strip

ζ(s) = 0 at every negative even integer — s = −2, −4, −6, and so on — a direct consequence of the functional equation Riemann derived, which relates ζ(s) to ζ(1 − s) via a factor that vanishes at those points. These are called the trivial zeros because their location is fully understood and uninteresting for number theory. Every other zero of ζ(s) is proven to lie in the critical strip, 0 < Re(s) < 1 — a fact that follows from the functional equation plus the original sum's non-vanishing for Re(s) > 1.

The Riemann Hypothesis

Riemann conjectured — and it remains unproved 165+ years later — that every non-trivial zero has real part exactly 1/2, lying precisely on the critical line Re(s) = 1/2 that bisects the critical strip. Over 10 trillion zeros have been computed numerically, and every single one sits on that line, which is strong empirical evidence but not a proof — number theory has been burned before by conjectures that held for astronomically large but still finite ranges before failing (Skewes' number and the Mertens conjecture are the classic cautionary tales). The Riemann Hypothesis is one of the seven Clay Millennium Prize Problems, worth a $1,000,000 prize for a proof or disproof, and remains the single most important open question in pure mathematics.

functional equation:  zeta(s) = 2^s * pi^(s-1) * sin(pi*s/2) * Gamma(1-s) * zeta(1-s)
trivial zeros:        s = -2, -4, -6, ...      (from the sin factor vanishing)
critical strip:       0 < Re(s) < 1           (where all non-trivial zeros must lie)
Riemann Hypothesis:   every non-trivial zero has Re(s) = 1/2 exactly (unproven)

Why this would pin down the primes

The connection to primes runs through the prime-counting function π(x) (the number of primes ≤ x) and its smooth approximation, the logarithmic integral Li(x). The Prime Number Theorem (proved in 1896, independently by Hadamard and de la Vallée Poussin, using exactly the fact that ζ(s) has no zeros on the line Re(s) = 1) already guarantees π(x) ~ Li(x) as x → ∞. But the error between π(x) and Li(x) — how tightly the actual prime distribution hugs the smooth curve — is controlled by exactly how close the non-trivial zeros are to Re(s) = 1/2. If the Riemann Hypothesis is true, that error is the smallest it could possibly be, O(√x log x); the further any zero strays from the critical line, the larger and less predictable the error is allowed to get. Proving it wouldn't change what the primes are — it would prove they're distributed as regularly as the evidence already suggests they are.

Frequently asked questions

Try it live

Everything above runs in your browser — open Riemann Zeta and change the parameters while it is running. Nothing is installed, nothing is uploaded, the whole model lives in one tab.