A quantum random number generator, or QRNG, produces random bits by measuring a quantum process, such as which way a single photon goes. It is a physical entropy source, not a new kind of cryptography. Most systems already get strong randomness from the operating system, so a QRNG is a niche choice, not an upgrade everyone needs.
How do computers normally get random numbers?
Cryptographic systems use a random bit generator made of two parts. NIST's SP 800-90 series describes them. SP 800-90B specifies the design principles and requirements for entropy sources, and the tests for validating them. SP 800-90A specifies deterministic random bit generators, which expand a small amount of true randomness into a long stream using hash functions or block ciphers. SP 800-90C covers how to construct complete generators from the two.
So the usual path is a noisy physical source feeding a deterministic generator. The noise might be timing jitter, thermal noise in a circuit or interrupt timing. The generator then produces as many bits as the system needs.
Where does a quantum source fit?
A QRNG replaces or supplements the noise source. Instead of electronic noise, it measures something whose outcome quantum mechanics says is unpredictable, for example photon arrival times or the state of vacuum fluctuations. Proponents describe this as randomness with a physical guarantee rather than an engineering assumption.
That is where care is needed. SP 800-90B does not ask whether a source is quantum. It asks whether the source is well characterised: a model of the noise, a stated amount of min-entropy per sample, health tests that run continuously, and validation. A quantum source with a bad detector, a bias or a failing component can still produce poor output. A good classical source with proper health tests can be fully adequate.
Why do people ask about QRNGs now?
Partly because of quantum computing news, and the two get confused. Post-quantum cryptography, the NIST standards in FIPS 203, 204 and 205, replaces algorithms that a future quantum computer could break. A QRNG is a hardware device that makes random numbers. You can use either without the other. A QRNG does not make your encryption quantum safe.
What about public randomness?
NIST runs an Interoperable Randomness Beacons project to promote trusted public randomness as a public utility, for example to make randomized processes auditable. Its page carries a plain warning: do not use beacon generated values as cryptographic secret keys. Public randomness proves fairness. It cannot be secret.
A constructed example
A lottery operator wants draws that outside auditors can verify. A hardware QRNG feeding the draw server gives them a source they can describe to regulators. For the audit trail they publish a commitment before each draw and use a public beacon value as an extra input that nobody on their side controls.
Meanwhile the operator's web servers keep using the operating system generator for session tokens and TLS keys. Nothing there needed a quantum device, and swapping it would add a part that can fail.
When is a QRNG worth considering?
Consider one when you must generate large volumes of keys in a controlled facility, when a regulator or customer specifically asks for a physical entropy source, or when you are building hardware that lacks a good noise source. In each case, require documentation of the entropy model, health tests and the validation you can show. Ask whether the module is validated under a recognised programme, and check the claim yourself.
How to check or apply this
- List where you generate keys, nonces and tokens today and which generator each uses.
- Confirm those systems use the operating system or a validated module, not a home built generator.
- For any new entropy hardware, ask for the SP 800-90B style documentation: noise model, min-entropy estimate, health tests.
- Check that the output feeds a deterministic generator rather than being used raw.
- Never use public beacon values as secrets.
- Keep post-quantum migration planning separate from entropy decisions.
Quick check
Is a QRNG more secure than a good system generator?
Not automatically. Security depends on the whole design, testing and validation. A quantum source with weak health checks can fail.
Does a QRNG protect against quantum computers?
No. It produces random bits. Protection against quantum attacks comes from post-quantum algorithms.
Sources
- NIST, Random Bit Generation project
- NIST SP 800-90B
- NIST SP 800-90A Rev. 1
- NIST, Interoperable Randomness Beacons
- NIST, Post-Quantum Cryptography project
Last reviewed: 18 October 2026.
