Maths for Quantum Computing

Quantum computing has a reputation for requiring heavy mathematics. That reputation is partly earned and partly exaggerated. Here is an honest breakdown of what you actually need, at what stage, and where to learn it -- without gatekeeping.

Student studying mathematics for quantum computing with laptop and textbooks

How much maths do you actually need?

The answer depends entirely on what you want to do with quantum computing.

For conceptual understanding -- grasping superposition, entanglement, and why quantum computers are interesting -- you need almost no formal mathematics. High school algebra is enough to follow intuitive explanations.

For gate-model algorithms -- implementing circuits in Qiskit or Cirq, understanding Grover's algorithm or quantum phase estimation -- you need linear algebra and complex numbers. These are genuinely necessary. A university-level linear algebra course or a dedicated self-study track of three to five weeks gets you there.

For variational algorithms and quantum machine learning -- VQE, QAOA, quantum neural networks -- you add multivariate calculus and some probability theory on top. These are used in gradient-based optimisation of circuit parameters.

For quantum error correction -- stabiliser codes, surface codes, logical qubit design -- you need group theory and abstract algebra. This is research-level territory.

For theoretical research -- proving complexity bounds, developing new algorithms -- you need a broad mathematical toolkit. This is a graduate-level pursuit and not what most practitioners need.

The takeaway: most people learning quantum computing for practical or career purposes need linear algebra and complex numbers, and that is it for a long time. Do not let unfamiliar-sounding prerequisites stop you from starting.

The essential topics

These are the mathematical building blocks that appear directly in quantum computing theory and practice.

Linear algebra

The core language of quantum computing. Quantum states are vectors. Quantum gates are matrices (specifically unitary matrices). Measurement is an eigenvalue problem. You need: vectors and vector spaces, matrix multiplication, inner products, eigenvalues and eigenvectors, and the tensor product for multi-qubit systems.

Complex numbers

Quantum amplitudes are complex-valued. Euler's formula (e^(i*theta) = cos + i*sin) is everywhere: rotation gates, the Bloch sphere, phase kickback, and the quantum Fourier transform all use it. You need: basic arithmetic with complex numbers, modulus and argument, and Euler's formula. A few hours covers it.

Probability and statistics

Quantum measurement is inherently probabilistic. The Born rule says the probability of measuring a state is the squared modulus of its amplitude. You need: basic probability, expectation values, and enough statistics to interpret shot-based results from real quantum hardware.

Calculus (optional)

Not needed for standard gate-based algorithms. Becomes important when you reach variational algorithms: VQE and QAOA use gradient descent to optimise circuit parameters, and quantum machine learning uses backpropagation. You can safely defer this until you reach variational methods.

Fourier analysis

The quantum Fourier transform (QFT) is a central subroutine in many algorithms including Shor's factoring algorithm and quantum phase estimation. Understanding the classical discrete Fourier transform first makes the quantum version much clearer. Basic familiarity is useful; deep mastery is not required.

Group theory (advanced)

Needed primarily for quantum error correction -- stabiliser codes and the Pauli group are described using group theory. Also useful for understanding symmetry-based quantum algorithms. This is advanced territory; most practitioners never need it outside research.

The topics you don't need

Quantum computing sounds intimidating in part because of its name. But many mathematical topics associated with physics are not required for computing work.

Quantum mechanics (the physics)

Quantum computing borrows the mathematical formalism of quantum mechanics but you do not need to understand the underlying physics. You do not need to know about wave-particle duality, the Schrodinger equation, or atomic structure. The circuit model is an abstraction that sits above the physics.

Differential equations

Not required for gate-based quantum computing. The Schrodinger equation is a differential equation, but quantum computing replaces continuous time evolution with discrete gate operations. You will encounter differential equations only if you go deep into Hamiltonian simulation or analog quantum computing.

Topology

Sometimes mentioned alongside quantum computing (topological quantum computing is a research approach). Not relevant for standard gate-model or annealing-based work. You can ignore this entirely unless you are specifically interested in topological qubits.

Real analysis and measure theory

These are graduate-level pure mathematics topics. They underpin rigorous probability theory but are not needed for practical quantum computing. If you have taken a real analysis course, great -- but you do not need it to start.

Learning order

If you are building your maths skills from scratch alongside learning quantum computing, this is a practical sequence that avoids spending time on topics you do not yet need.

1

Linear algebra basics

Start here. Focus on: vectors, matrices, matrix multiplication, and eigenvalues. Brilliant's Linear Algebra course or the 3Blue1Brown "Essence of Linear Algebra" series are both excellent starting points. Give this two to four weeks of part-time study.

2

Complex numbers

Short and focused. You need to understand complex arithmetic, the complex plane, modulus, and Euler's formula. This can be covered in a few dedicated hours -- it does not require weeks of study. Brilliant's Complex Numbers course is a good option.

3

Probability fundamentals

Enough to understand the Born rule and interpret measurement results. Basic probability -- events, distributions, expectation values -- is sufficient for most gate-model work. This pairs well with starting hands-on circuit work in Qiskit.

4

Calculus for variational methods (optional)

Add this only when you reach variational quantum algorithms. At that point you need partial derivatives and gradient descent. If you have covered calculus before, a quick refresher is enough. If not, defer it until you are actively working on VQE or QAOA.

Looking for a structured path? See the learning paths guide or the prerequisites overview for more detail on building up from the fundamentals.

Courses for maths and quantum computing

Courses covering the mathematical foundations of quantum computing, including linear algebra, probability, and complex numbers.

Frequently asked questions

Do you need a maths degree to learn quantum computing?
No. A maths degree is not required, especially for conceptual or introductory study. Many quantum computing courses are designed for people with only high school algebra. For practical work with real algorithms -- implementing Grover's, running Qiskit circuits, understanding quantum machine learning -- you need linear algebra and a working knowledge of complex numbers. That is a fraction of a university maths education and can be picked up in a few weeks with the right resources.
What linear algebra do you need for quantum computing?
You need: vectors and vector spaces (quantum states are column vectors), matrix multiplication (applying gates means multiplying matrices), inner products and norms (for calculating probabilities from amplitudes), eigenvalues and eigenvectors (measurement and Hamiltonian dynamics), the tensor product (combining multi-qubit systems). You do not need advanced topics like abstract algebra, measure theory, or functional analysis for most practical quantum computing work.
Are complex numbers required for quantum computing?
Yes -- complex numbers are unavoidable. Quantum amplitudes are complex numbers, not real numbers. The Born rule (squaring the amplitude to get a probability) only makes sense if you understand complex modulus. Euler's formula (e^(i*theta) = cos(theta) + i*sin(theta)) underpins rotation gates, the Bloch sphere, and the quantum Fourier transform. The good news: you only need the basics of complex numbers, not complex analysis. A few hours of study covers what you need.
How much calculus do you need?
For gate-based quantum computing: very little. Standard textbook algorithms (Grover, Shor, quantum phase estimation) require almost no calculus. Calculus becomes important if you work on variational quantum algorithms (VQE, QAOA) or quantum machine learning, where gradient descent is used to optimise circuit parameters. If you are new to quantum computing, you can safely defer calculus and come back to it when you reach variational methods.
What is the hardest maths concept in quantum computing?
For most learners, the tensor product is the first genuinely difficult concept. It explains how multi-qubit systems combine -- and why an n-qubit system lives in a 2^n-dimensional space rather than an n-dimensional one. Understanding entanglement mathematically requires the tensor product. After that, spectral decomposition (expressing operators in terms of their eigenvectors and eigenvalues) is the next conceptual hurdle, and it is central to quantum measurement and Hamiltonian simulation.