Math Foundations
Tensor Product
The mathematical operation that combines the states of separate quantum systems into the state of the joint system.
Definition
The tensor product is how quantum mechanics describes more than one thing at a time. If you have one qubit and add another, the joint system is described by the tensor product of the two. It is the reason a register of n qubits has 2^n amplitudes instead of 2n, and it is also where entanglement comes from: some joint states cannot be written as a tensor product of states of the parts.
In depth
A single qubit state needs two amplitudes. Two qubits in a product state look like a tensor product of two such pairs, and multiplying them out gives four amplitudes for the basis states 00, 01, 10, and 11. A general two-qubit state has four independent amplitudes that need not factor. When they do not factor, the qubits are entangled, and neither has a state of its own.
Operations obey the same rule. A gate that acts on only the first qubit of two is written as the gate tensor the identity on the second qubit. A circuit with many gates on different qubits corresponds to tensor products of small matrices, combined by ordinary multiplication for sequential steps. This structure is why quantum circuit simulators store states as large vectors and apply gates by reshaping them, and why tensor network methods can compress states that have limited entanglement.
Because dimensions multiply, adding one qubit doubles the size of the space. This exponential growth is the root of both quantum computing's promise and the difficulty of classical simulation. It also explains why a quantum computer's usefulness depends on arranging that the final state has structure the measurement can reveal, since the exponentially large space is not directly readable.
Technical Definition
For vector spaces V and W with bases, the tensor product V tensor W has dimension equal to the product of the dimensions, and basis vectors that are pairs of basis vectors. For states, the product state of a and b is written a tensor b. For operators A and B acting on separate systems, A tensor B acts on the joint system. A joint state that cannot be written as a single tensor product is entangled.
Visual Explanation: An Analogy
Imagine describing two dice. If the dice are independent, you can describe each separately and multiply the odds. But if the dice are secretly linked, you need a full table of 36 outcomes. The tensor product is the full table, and a product state is the special case where the table factors into two independent lists.
Key facts
| Item | Detail |
|---|---|
| Dimension rule | Dimensions multiply: 2 by 2 by 2 gives 8 for three qubits |
| Notation | Written with the symbol for tensor product, or simply as adjacent kets such as 01 |
| Operators | A gate on one qubit becomes A tensor identity on the rest |
| Entanglement test | A pure two-qubit state is entangled exactly when it cannot be written as a tensor product |
| Software | Libraries use the Kronecker product, the matrix form of the tensor product |
Where the idea comes from
Tensor products were developed in mathematics in the 19th and early 20th centuries, and they entered physics as the right way to combine systems in quantum mechanics. Einstein, Podolsky, and Rosen's 1935 paper and Schrödinger's reply drew attention to joint states that do not factor, which Schrödinger called entangled.
Real-World Use Cases
- Building multi-qubit states and operators in simulators and textbooks.
- Identifying entanglement by checking whether a joint state factors.
- Compressing quantum states with tensor networks, which exploit limited entanglement.
- Counting the cost of simulating larger systems, since dimension grows multiplicatively.
How it differs from related ideas
- The tensor product combines independent systems multiplicatively, while a direct sum of spaces, used for alternatives, combines them additively.
- A product state has no correlations beyond classical ones, while an entangled state has correlations that cannot be reproduced by any classical shared randomness.
- The Kronecker product of two matrices is the concrete numerical version of the tensor product of two operators.
Common Misconceptions
- The tensor product is not ordinary multiplication of numbers. It builds a bigger object from two smaller ones.
- Not every state of two qubits is a tensor product. Most are not, and those are the entangled ones.
- Gates on different qubits are not applied one after another by necessity. If they act on separate qubits, they commute and can run in parallel.
Frequently asked questions
Why does a quantum register have 2^n amplitudes?
Because the Hilbert space of n qubits is the tensor product of n two-dimensional spaces, and dimensions multiply.
How do I tell whether a state is entangled?
For a pure two-qubit state, check whether it can be written as a product of one-qubit states. If it cannot, it is entangled.
What is the Kronecker product?
The operation that turns two matrices into a larger block matrix, and it is how the tensor product of operators is written in code.
Related Terms
Keep exploring
- Algorithm: Deutsch-Jozsa Algorithm
- Algorithm: Quantum Fourier Transform (QFT)
- Hardware: IBM Heron (r1, r2, and r3)