Classical computers produce pseudorandom numbers. Quantum generators use truly random quantum events — with enormous implications for cryptography.
🎲 Randomness: A Fundamental Problem
Every time a computer “rolls the dice” — generates cryptographic keys, runs a Monte Carlo simulation, or decides a music shuffle order — it uses random number generators. But a classical computer is, at its core, a deterministic machine. It executes algorithms — sets of instructions that always produce the same output for the same input. How can something deterministic produce something random?
The short answer: it can't — at least not truly. What classical computers produce is called pseudorandomness. And there is only one known way to obtain genuine randomness: quantum mechanics.
⚙️ Pseudorandom Numbers: The Illusion of Randomness
A pseudorandom number generator (PRNG) starts from an initial value called a “seed” and applies a mathematical function to produce a sequence of numbers. The sequence appears random — it passes statistical tests and displays no obvious patterns — but it is entirely deterministic. If someone knows the algorithm and the seed, they can reproduce the entire sequence.
For many applications, this is not a problem. In scientific modeling, games, and even some statistical analyses, pseudorandomness suffices — and is even desirable because it allows reproducibility. In cryptography, however, the situation changes radically. If an attacker can guess the seed or the algorithm, they can break the entire encryption system. According to NIST standards (SP 800-90B), two fundamental requirements must be satisfied: forward secrecy — knowledge of past outputs must not allow prediction of future ones — and backward secrecy — knowledge of the future internal state must not reveal past data.
🔬 Why Quantum Mechanics Is Different
According to standard interpretations of quantum mechanics, microscopic phenomena are objectively random. This is not a matter of missing information or measurement limitations — randomness is inherent in nature. When a photon reaches a beamsplitter, it can either be reflected or transmitted. Quantum mechanics can only calculate the probability of each outcome — not which one will occur. According to the Born rule, the probability of each measurement result is determined by the wave function, but the actual result remains fundamentally unpredictable.
This fundamental randomness has been experimentally certified through Bell tests. Pironio et al. published in Nature (2010) a paper titled “Random Numbers Certified by Bell's Theorem,” demonstrating that quantum nonlocality can be used to certify genuine randomness in a given number sequence — something impossible with any classical method.
💻 How a Quantum Random Number Generator Works
A quantum random number generator (QRNG) uses a quantum phenomenon as an entropy source. Herrero-Collantes and Garcia-Escartin in their comprehensive review in Reviews of Modern Physics (2017) catalogued the main methods:
Single-photon beamsplitter branching. A photon from a single-photon source is sent to a beamsplitter. The photon randomly takes one of two paths and is detected by a single-photon detector. Each detection generates a random bit: 0 or 1. This method is the most intuitive — quantum mechanics in its purest form.
Vacuum fluctuations. Even the “vacuum” in quantum mechanics is not empty. The quantum vacuum state exhibits random energy fluctuations. Using homodyne detection with a laser, these generators measure variations in the vacuum state and convert them into random bits.
Laser phase noise. Phase noise at the output of a single spatial mode laser is converted to amplitude through an unbalanced Mach-Zehnder interferometer. Sampling is done by a photodetector, providing generation rates of many Gbps — suitable for high-speed applications.
Other quantum methods exist: nuclear decay (the oldest, used since the 1960s with Geiger counters), amplified spontaneous emission, Raman scattering, and optical parametric oscillation. Each method has its own trade-offs in speed, cost, and size.
🛡️ Trusted and Device-Independent Generators
Mannalath, Mishra, and Pathak in their comprehensive review (2023) in Quantum Information Processing classify QRNGs into categories based on trust. "Trusted" generators (trusted QRNGs) operate in a controlled environment where the manufacturer guarantees that the entropy source is quantum. They constitute the majority of commercial products — before 2017, 8 commercial QRNGs were already available — but they cannot mathematically prove they are not being manipulated.
At the opposite end, device-independent QRNGs use Bell inequality violations to prove that randomness is genuine, without requiring trust in the device's internal construction. This approach, while slower, offers mathematically certified security — a crucial advantage for military and governmental applications.
📊 Applications and Significance
The primary application of QRNGs is cryptography. Cryptographic keys must be unpredictable — if the seed of a PRNG is compromised, all keys are exposed. A QRNG provides entropy that depends on no initial value. In practice, it is often used as a seed source for cryptographically secure pseudorandom number generators (CSPRNG), combining true randomness with high generation rates.
Beyond cryptography, QRNGs find applications in Monte Carlo simulations, gambling (where sequence non-reproducibility is a legal requirement), quantum key distribution (QKD), and voting protocols where impartiality must be demonstrable. The market is growing rapidly, with companies like ID Quantique, QuintessenceLabs, and Toshiba producing commercial products, while online services offer quantum randomness as a service (Randomness as a Service).
🔮 The Only True Randomness
Classical phenomena — thermal noise, Zener noise, avalanche breakdown, ring oscillator circuits — are chaotic and unpredictable in practice, but not fundamentally random. If we knew the exact initial conditions, we could theoretically predict them. Quantum mechanics sets fundamental limits: Heisenberg's uncertainty principle, Bell's theorem, the Born rule — all show that there is no hidden deterministic explanation behind quantum phenomena.
From the dice of ancient Iraq 5,000 years ago to quantum chips in today's smartphones, the quest for random numbers has traveled an enormous distance. But only with quantum generators did we finally reach something that nature itself guarantees: real, provably unpredictable randomness.