RosneT: a block tensor algebra library for out-of-core quantum computing simulation

Abstract

With recent Quantum Devices showing increasing capabilities to perform controlled operations, further development on Quantum Algorithms may benefit from Quantum Simulations on classical hardware. Among important applications one finds verification and debugging of Quantum Algorithms, and elucidating the frontier for real Quantum Advantage of new devices [1]. Tensor Networks are regarded as an efficient numerical representation of a Quantum Circuit, but exponential growth forces tensors to be distributed among computing nodes. A number of methods and libraries have appeared recently to implement Quantum Simulators with Tensor Networks [2], [3] intended for HPC clusters. In this work we develop a Python library called RosneT using a task-based programming model able to extend all tensor operations into distributed systems, and prepared for existing and upcoming Exascale supercomputers. It is compatible with the Python ecosystem, and offers a simple programming interface for developers.