English

Two-Tower Quantum Matrix Chain Multiplication: Trading Qubits for Depth

Quantum Physics 2026-07-14 v1

Abstract

Matrix chain multiplication -- computing W=M(0)M(K1)\mathcal{W} = M^{(0)}\cdots M^{(K-1)} where M(k)RPk×Pk+1M^{(k)} \in \mathbb{R}^{P_k \times P_{k+1}} -- arises in scientific computing, machine learning, and graph analysis. Despite the importance of this problem, for chains of distinct matrices, the classical number of operations grows linearly with the chain length KK and polynomially in the matrix dimensions. We present \emph{Two-Tower Matrix Multiplication}, a quantum subroutine that encodes the product W\mathcal{W} of the KK matrices into a quantum state in circuit depth O(maxkpolylog(PkPk+1))\mathcal{O}(\max_{k} \mathrm{polylog} (P_k P_{k+1})), which is independent of~KK within the QRAM-based state-preparation model, whereas the qubit count is O(klogPk)\mathcal{O}\bigl(\sum_{k} \log P_k \bigr); the total gate count remains linear in KK, so the gain is in the circuit depth. The construction interleaves state-preparation operators across two layers; within each layer, all operators act on disjoint registers and execute in parallel. This subroutine can be specialized for the chain-vector case, which computes the product of K1K-1 matrices applied to a vector. We prove the correctness of the subroutine for all KK and provide two implementations using the Qiskit and QCLAB frameworks. The subroutine is applicable to any downstream quantum algorithm that operates on a matrix encoded in the statevector, including norm estimation, graph-matrix powers, linear system solving, and quantum machine learning kernels.

Cite

@article{arxiv.2607.13191,
  title  = {Two-Tower Quantum Matrix Chain Multiplication: Trading Qubits for Depth},
  author = {Giacomo Antonioli and Anna Bernasconi and Alessandro Berti and Gianna M. Del Corso and Alessandro Poggiali},
  journal= {arXiv preprint arXiv:2607.13191},
  year   = {2026}
}