Quantum computing glossary

Short, plain definitions. When a term needs more room, we link to the article that explains it properly.

CBOM (cryptographic bill of materials): An inventory of the cryptography in your software and systems: which algorithms, key lengths, certificates and libraries are used, and where. The CycloneDX standard supports CBOMs. Building one is usually the first step of a quantum readiness assessment.
CNSA 2.0: The NSA's Commercial National Security Algorithm Suite 2.0, announced in September 2022, which sets the quantum-resistant algorithms for US national security systems (including ML-KEM-1024 and ML-DSA-87). It sets deadlines by system type, with 2035 as the target for the full transition. Suppliers to government feel it first.
Coherence time: How long a qubit keeps its quantum state before decoherence corrupts it. It varies widely by technology, from microseconds for superconducting qubits to seconds or longer for trapped ions. What matters in practice is how many gates fit inside that window.
CRQC (cryptographically relevant quantum computer): A quantum computer large and reliable enough to break RSA or elliptic-curve cryptography with Shor's algorithm. None exists today. Estimates of when one could appear vary widely, which is why governments set migration deadlines instead of waiting for a date.
Crypto-agility: The ability to swap cryptographic algorithms, key sizes or libraries without rewriting applications. Systems that hard-code RSA or elliptic curves are the slowest to migrate, so crypto-agility is a design goal for any PQC program and for whatever comes after it.
CUDA-Q: NVIDIA's open-source platform for programming hybrid quantum-classical applications in Python or C++. It runs circuits on GPU-accelerated simulators and on QPUs from several hardware vendors.
Decoherence: The loss of a qubit's quantum state through unwanted interaction with its environment: heat, vibration, stray electromagnetic fields. Decoherence is the main reason current quantum computers make errors and can only run short circuits.
Entanglement: A correlation between qubits that has no classical equivalent: the state of the group cannot be described one qubit at a time. Measuring one entangled qubit tells you something about the others. Entanglement cannot be used to send information faster than light.
Fault tolerance: The ability of a quantum computer to run arbitrarily long computations because error correction removes errors faster than they build up. No machine is fully fault tolerant today; it is the stated goal of every major hardware roadmap.
Fidelity: A measure of how close an operation or state is to the ideal one, usually given as a percentage. A two-qubit gate fidelity of 99.9% means roughly one error per thousand operations. Small differences in fidelity decide how deep a circuit you can run.
FN-DSA (FIPS 206): A post-quantum signature algorithm based on Falcon that NIST is standardizing as FIPS 206. It has smaller signatures than ML-DSA, but its signing uses floating-point sampling that is hard to implement safely. Plan for it and test it, but it is not yet a final standard.
Grover's algorithm: A quantum search algorithm published by Lov Grover in 1996 that finds a marked item among N possibilities in about the square root of N steps. Against symmetric ciphers it roughly halves the effective key length in theory, which is why AES-256 is the usual recommendation. It does not break AES or SHA-2.
Harvest now, decrypt later: An attack where someone records encrypted traffic today and stores it until a quantum computer can break the key exchange. It makes the quantum threat a present problem for data that must stay secret for many years, such as financial, health or government records.
HQC (Hamming Quasi-Cyclic): A key-encapsulation algorithm based on error-correcting codes that NIST selected in March 2025 as a backup to ML-KEM. Because it relies on different mathematics than lattices, it gives a fallback if a weakness is found in ML-KEM.
Hybrid key exchange: A key exchange that combines a classical algorithm and a post-quantum one, so the connection stays secure as long as either holds. The most common example is X25519MLKEM768 in TLS 1.3, already enabled by default in Chrome and supported by Cloudflare.
Hybrid quantum-classical computing: A workflow where classical computers do most of the work and call a quantum processor for specific steps, often in a loop. Nearly every practical quantum application today is hybrid. A hybrid proof of concept is the usual way to test one on your own data.
Interference: The way amplitudes of quantum states add up or cancel out, like waves. Quantum algorithms are designed so that wrong answers cancel and the right answer is reinforced before measurement.
Lattice-based cryptography: Cryptography built on hard problems involving high-dimensional grids of points (lattices), such as Learning With Errors. No known quantum algorithm solves these problems efficiently. ML-KEM, ML-DSA and FN-DSA are all lattice based.
Logical qubit: A qubit encoded across many physical qubits using quantum error correction, so it has a much lower error rate than any single physical qubit. Useful fault-tolerant algorithms will need hundreds to thousands of logical qubits.
Measurement: Reading a qubit, which returns a classical 0 or 1 and destroys the superposition. Because each run gives one sample, a quantum program is usually executed thousands of times (shots) to estimate the distribution of results.
ML-DSA (FIPS 204): The Module-Lattice-Based Digital Signature Algorithm, NIST's primary post-quantum signature standard. It is based on CRYSTALS-Dilithium and was published as FIPS 204 on August 13, 2024. Its signatures and keys are larger than RSA or ECDSA ones, which affects certificates and protocols.
ML-KEM (FIPS 203): The Module-Lattice-Based Key-Encapsulation Mechanism, NIST's main standard for establishing shared keys resistant to quantum attacks. It is based on CRYSTALS-Kyber and was published as FIPS 203 on August 13, 2024, with three parameter sets: ML-KEM-512, ML-KEM-768 and ML-KEM-1024.
Mosca's theorem: A rule of thumb proposed by Michele Mosca: if x is how long your data must stay secret, y is how long your migration will take, and z is the time until a CRQC exists, you have a problem when x + y is greater than z. It turns an uncertain threat into a planning decision.
Neutral-atom qubits: Qubits stored in uncharged atoms held in arrays by focused laser beams called optical tweezers. QuEra, Pasqal and Atom Computing work on this approach. It scales to large arrays of atoms, and the atoms can be moved to change which qubits interact.
NISQ (Noisy Intermediate-Scale Quantum): A term coined by John Preskill in 2018 for today's quantum computers: tens to a few thousand physical qubits with no full error correction. NISQ machines are useful for research and experiments, but no NISQ algorithm has yet beaten classical methods on a practical business problem.
NIST IR 8547: A NIST report, released as a draft in November 2024, that proposes a transition timeline to post-quantum cryptography. It would deprecate quantum-vulnerable algorithms such as RSA and elliptic-curve cryptography at 112-bit security after 2030 and disallow them after 2035.
Photonic quantum computing: Quantum computing that uses particles of light (photons) as qubits. Photons work at or near room temperature for most components and travel well over fiber, but making them interact reliably is hard. PsiQuantum and Xanadu follow this approach.
Physical qubit: An actual hardware qubit: a superconducting circuit, a trapped ion, a neutral atom or a photon. Physical qubits are noisy, so a qubit count on a vendor's roadmap says little on its own without error rates.
Post-quantum cryptography (PQC): Public-key cryptography designed to resist attacks from both classical and quantum computers, while running on today's hardware. NIST published its first three PQC standards (FIPS 203, 204 and 205) on August 13, 2024. Moving to them is a multi-year PQC migration for most organizations.
Q-Day: Informal name for the day a CRQC exists and current public-key encryption can be broken. Nobody knows the date, and it may not be announced publicly when it happens. Data harvested before Q-Day becomes readable after it.
QAOA (Quantum Approximate Optimization Algorithm): A hybrid algorithm introduced by Farhi, Goldstone and Gutmann in 2014 for combinatorial optimization problems such as routing and scheduling. A classical optimizer tunes the parameters of a short quantum circuit. It has not yet shown a clear advantage over classical solvers on real logistics problems.
Qiskit: IBM's open-source software development kit for quantum computing, written mainly in Python. You use it to build circuits, run them on simulators or IBM quantum hardware, and work with algorithms and error mitigation tools.
QKD (quantum key distribution): A method for two parties to share an encryption key using quantum states of light, where any eavesdropping disturbs the states and can be detected. The best-known protocol is BB84 (1984); QKD needs dedicated hardware and has distance limits, and the NSA does not recommend it for national security systems, preferring post-quantum cryptography. It is most discussed in telecommunications.
QPU (quantum processing unit): The chip or device that holds the qubits and runs quantum operations. In practice a QPU works as an accelerator next to CPUs and GPUs, which prepare the circuits and process the results.
QRNG (quantum random number generator): A device that produces random numbers from a quantum process, such as photon behavior, instead of from an algorithm. It is already sold commercially and used in some security products and phones.
Quantum advantage: When a quantum computer solves a useful problem faster, cheaper or more accurately than the best classical method. Claims are often challenged by better classical algorithms, so treat any announcement as provisional until others reproduce it. Finding where advantage might appear first is the goal of quantum use case discovery.
Quantum annealing: A form of quantum computing built for optimization: the machine starts in a simple state and slowly evolves toward the lowest-energy state of a problem you encode. D-Wave builds commercial annealers. Annealers are not general-purpose and cannot run Shor's algorithm.
Quantum circuit: A sequence of quantum gates applied to a set of qubits, followed by measurements. It is the standard way to write a program for a gate-based quantum computer. Circuit depth (how many gate layers run one after another) is limited by how long the qubits stay coherent.
Quantum error correction: Techniques that spread one qubit of information across several physical qubits and detect errors without measuring the data directly. It is the only known path to running long algorithms like Shor's. In December 2024, Google's Willow chip showed error rates falling as the error-correcting code grew larger.
Quantum gate: An operation that changes the state of one or more qubits, the quantum equivalent of a logic gate. Common examples are the Hadamard gate, which creates superposition, and the CNOT gate, which can create entanglement between two qubits.
Quantum machine learning: Research that uses quantum circuits as models or subroutines in machine learning. Results so far are mostly small experiments, and loading large classical datasets into a quantum computer remains a serious bottleneck. Treat claims of practical speedups with caution.
Quantum sensing: Using quantum systems to measure magnetic fields, gravity, time or acceleration with very high precision. It is closer to commercial use than quantum computing, with applications in navigation, medical imaging and mineral exploration.
Quantum simulation: Using a quantum computer to model molecules, materials or other quantum systems that are expensive to simulate classically. It is widely considered the most likely early application, with interest in drug discovery, batteries and catalysts. See how it applies to healthcare and pharma.
Quantum supremacy: A term proposed by John Preskill in 2012 for a quantum computer doing any task, useful or not, that classical computers cannot do in a reasonable time. Google claimed it in 2019 with its 53-qubit Sycamore processor on a random sampling task. Many researchers now prefer the term "quantum advantage".
Quantum volume: A single-number benchmark introduced by IBM researchers that combines qubit count, gate fidelity and connectivity. It measures the largest square circuit (same width and depth) a machine can run reliably. It is useful for comparing gate-based systems but says little about performance on a specific application.
Quantum-inspired algorithms: Classical algorithms that borrow ideas from quantum computing, such as tensor networks or annealing-style search, and run on ordinary hardware. Sometimes they solve a problem well enough that a quantum computer is not needed, which is a legitimate result of a quantum project.
Qubit: The basic unit of quantum information. A classical bit is either 0 or 1; a qubit can be in a combination of both until it is measured, and measuring it always gives 0 or 1.
Shor's algorithm: A quantum algorithm published by Peter Shor in 1994 that factors large numbers and computes discrete logarithms exponentially faster than any known classical method. On a large fault-tolerant quantum computer, it would break RSA, Diffie-Hellman and elliptic-curve cryptography. No existing machine is anywhere near large enough to run it against real keys.
SLH-DSA (FIPS 205): The Stateless Hash-Based Digital Signature Algorithm, published as FIPS 205 on August 13, 2024 and based on SPHINCS+. Its security rests only on hash functions, a conservative assumption, but its signatures are large and slower to create. It is a backup in case lattice-based schemes are weakened.
Superconducting qubits: Qubits made from tiny superconducting circuits cooled to near absolute zero (about 10 to 20 millikelvin) in a dilution refrigerator. IBM, Google and Rigetti use them. Their gates are fast, but coherence times are short and scaling the wiring and cooling is hard.
Superposition: A qubit state that combines 0 and 1 with specific amplitudes. Those amplitudes set the probability of each result when you measure. Superposition by itself does not make a computer fast; algorithms need interference to turn it into a useful answer.
Surface code: A quantum error correction code that arranges physical qubits on a 2D grid, with each qubit only talking to its neighbors. It tolerates relatively high physical error rates, which is why Google and others use it. Its cost is overhead: one logical qubit can need hundreds of physical qubits or more.
Topological qubit: A proposed qubit that stores information in the global properties of an exotic state of matter, which should make it naturally resistant to local noise. Microsoft announced its Majorana 1 chip in February 2025, but the evidence that it contains working topological qubits is still debated by physicists.
Trapped-ion qubits: Qubits stored in individual charged atoms held in place by electromagnetic fields and controlled with lasers. Quantinuum and IonQ build them. They have long coherence times and high gate fidelities, with slower gate speeds than superconducting qubits.
VQE (Variational Quantum Eigensolver): A hybrid algorithm, first demonstrated in 2014, that estimates the lowest energy of a molecule or material. The quantum computer prepares a trial state, the classical computer adjusts it, and the loop repeats. It is one of the most studied algorithms for chemistry on NISQ hardware.

Get in before the queue forms

We are taking a short list of companies for our first readiness assessments and post-quantum migrations. Tell us what you are working on and we will get back to you within two business days.

Write to us