Short answer: A quantum random number generator (QRNG) uses the intrinsically probabilistic behaviour of quantum mechanics as its source of randomness, instead of a deterministic classical process. Reviews describe designs that differ in how much you must trust the device, from standard devices to device-independent and semi-device-independent schemes. A 2020 testing study found issues with current certification schemes, so a quantum origin alone is not proof of good output.
Why randomness matters
An arXiv review of QRNGs says random numbers are central to cryptography and various other tasks. A 2020 paper in an ACM journal adds that generating truly random numbers for cryptography can be a slow and expensive process, and that quantum physics offers sources of entropy that may be genuinely unpredictable.
Classical and quantum sources
A review in the International Journal of Theoretical Physics says classical sources are deterministic at the microscopic level and rely on computational pseudo randomness. By contrast, it says quantum theory provides an intrinsically probabilistic framework in which genuine randomness arises independent of any ignorance or incomplete knowledge about the system. The arXiv review says this intrinsic probabilistic nature has allowed researchers to construct a large number of QRNGs that differ from the traditional true number generators.
Device types and certification
The arXiv review focuses on features that are not achievable in the classical world, naming device independence and semi-device independence. It also discusses the origin of randomness from a hierarchy of axioms for quantum mechanics, including incompatibility and uncertainty, contextuality, entanglement, nonlocality and indistinguishability of identical particles.
The Springer review covers device-independent (DI) and semi-device-independent (SDI) randomness certification schemes, with particular focus on dimension witness approaches, quantum steering and protocols for sequential randomness certification. It traces the field from early theoretical proposals to recent experimental milestones. The idea behind certification is that the randomness is checked from the statistics of the experiment, with fewer assumptions about the internal workings of the device.
What testing has found
The ACM study presents what it calls the first published analysis of one family of QRNGs, along with a second hardware device and two online services. Each was examined with five batteries of statistical tests: Dieharder, NIST SP800-22, Ent, Tuftests and TestU01. The authors say their analysis highlights issues with current certification schemes. This article does not repeat product-level findings, which readers should take from the paper directly and which describe the devices as tested in that study.
What this does not tell you
This article summarises the abstracts of three papers. It does not recommend or rank any device, and a QRNG does not by itself make a system secure. Randomness is one input to cryptography, and the surrounding implementation matters too.
FAQ
What is a quantum random number generator?
A device that uses the intrinsic probabilistic nature of quantum mechanics as its source of randomness, according to an arXiv review.
How is quantum randomness different from classical randomness?
A Springer review says classical sources are deterministic at the microscopic level and rely on pseudo randomness, while quantum theory provides genuine randomness independent of any ignorance about the system.
What does device-independent mean?
Device-independent and semi-device-independent schemes certify randomness with fewer assumptions about the device. The reviews cover them as features not achievable in the classical world.
Are QRNG outputs always unbiased?
Not necessarily. A 2020 ACM study tested several generators with five statistical test batteries and says its analysis highlights issues with current certification schemes.
Does a QRNG make a system secure?
No. Randomness is one ingredient of cryptography, and none of these papers claims otherwise.
All field notes