An operation on two matrices A (m×n) and B (p×q) producing the block matrix A⊗B of size (mp)×(nq) whose (i,j) block equals a_{ij} B; represents the matrix of the tensor product of linear maps in chosen bases.

Principle

The Kronecker product implements the tensor product at the level of coordinate matrices: (A⊗B)(C⊗D)=(AC)⊗(BD) when dimensions are compatible, and vec(AXB)=(B^T⊗A) vec(X) provides a useful identity for linear algebra manipulations.

Demonstration

If A=[[a_{11},a_{12}],[a_{21},a_{22}]] and B is 2×2, then A⊗B=[[a_{11}B,a_{12}B],[a_{21}B,a_{22}B]] gives a 4×4 block matrix; used to form large structured matrices from small factors.

Misapplication

Confusing the Kronecker product with the Hadamard (elementwise) product yields incorrect algebraic relations; likewise assuming A⊗B is commutative (A⊗B ≠ B⊗A in general) leads to errors.

Consequence

Enables separation of variables, compact representation of tensorized linear operators, and efficient exploitation of structure in numerical methods (e.g., solving Sylvester equations, representing multi‑index operators).

Reversal

Direct sum (A⊕B) combines matrices along the diagonal to form block‑diagonal operators representing independent action, in contrast to Kronecker product which models coupled tensor‑product action across factors.

Boundary

Defined for finite‑dimensional matrices over a field or ring; extension to infinite‑dimensional operators requires topological tensor product constructions and completion; algebraic identities assume conformable dimensions.

Semantic Tension

Kronecker product versus abstract tensor product: Kronecker is a concrete matrix representation dependent on chosen bases, while the abstract tensor product is basis‑free and requires canonical isomorphisms to identify with Kronecker matrices.

Synthesis

The Kronecker product is the blockwise matrix realization of the tensor product of linear maps in fixed bases: it constructs large structured matrices from smaller factors and preserves algebraic identities that facilitate vectorization and separable operator methods.