Tensor Products (Multi-Qubit States)
Track: Foundations · Difficulty: Beginner · Est: 15 min
Tensor Products (Multi-Qubit States)
Overview
A single qubit has states like . But real quantum computation becomes interesting only when we combine qubits into joint systems.
This page introduces the mathematical rule for combining quantum systems: the tensor product. You will learn why it is required (not optional), how it acts on basis states, and how the compact notation , , , arises.
Intuition
For two classical bits, the joint state is a pair like . There are four possibilities: , , , .
For two qubits, the joint system also has four computational-basis outcomes: , , , . But the state is not just “a pair of qubit states.”
Why not?
- A qubit state is a vector (a rule for probabilities across outcomes).
- A two-qubit state must be a single object that predicts probabilities for joint outcomes.
- Those probabilities must allow superposition and interference across the four joint outcomes.
So the state space of a joint system must be a space where you can:
- represent the four joint basis outcomes, and
- take linear combinations of them with complex amplitudes.
That is exactly what the tensor product construction provides.
A useful way to think about it: “two qubits” means the number of basis outcomes multiplies (2 outcomes per qubit → outcomes total), and therefore the number of amplitude coordinates multiplies too.
Formal Description
Single-qubit basis
Recall the computational basis for one qubit:
A general single-qubit state is
Tensor product of basis states
For a two-qubit system, the computational basis is built from tensor products of the one-qubit basis states:
We use the shorthand:
Here is what the notation means in words:
- means: “first qubit is in the basis state and second qubit is in the basis state .”
- The left digit refers to the first subsystem; the right digit refers to the second.
Tensor product distributes over addition
The tensor product is linear in each input. The most important expansion rule is:
In shorthand:
This is how “a state of qubit A and a state of qubit B” becomes a single joint state.
General two-qubit state
A general pure state of two qubits can be written as a linear combination of the four basis states:
Each coefficient is a complex amplitude.
If you measure both qubits in the computational basis, then
and normalization requires
Cartesian product vs tensor product
A Cartesian product of sets lists pairs, like . It describes classical joint possibilities.
A tensor product of vector spaces creates a space where you can form superpositions of joint basis states. That is why quantum joint states use tensor products rather than simple pairs.
Worked Example
Let qubit A be in
and qubit B be in .
The joint state is the tensor product:
Expand using linearity:
Now measure both qubits in the computational basis.
- Outcome occurs with probability .
- Outcome occurs with probability .
- Outcomes and have probability 0.
This matches the intuition: the second qubit is definitely 0, while the first is probabilistic.
Turtle Tip
If you feel lost, anchor on one fact: two qubits have four joint basis outcomes. A two-qubit pure state is “a list of four complex amplitudes” attached to .
Common Pitfalls
Don’t read as a number “one” or as a product . It is shorthand for a tensor product of basis states.
Also, be careful about ordering: and are different states because they refer to different subsystems.
Quick Check
- How many computational-basis states does a two-qubit system have, and why?
- Expand in the basis.
What’s Next
Tensor products tell us how to build joint state spaces. Next we’ll classify two-qubit states into two fundamentally different types: product states (which factor into single-qubit states) and entangled states (which do not).
