QX: A High-Performance Quantum Computer Simulation Platform
N. Khammassia, I. Ashrafb, X. Fuc, C.G. Almudeverd and K. Bertelse
QuTech, Computer Engineering Lab., Delft University of Technology Delft, The Netherlands.
an.khammassi@tudelft.nl
bi.ashraf@tudelft.nl
cx.fu-1@tudelft.nl
dc.garciaalmudever-1@tudelft.nl
ek.l.m.bertels@tudelft.nl
ABSTRACT
Quantum computing is rapidly evolving especially after the discovery of several efficient quantum algorithms solving intractable classical problems such as Shor's factoring algorithm. However the realization of a large-scale physical quantum computer is very challenging and the number of qubits that are currently under development is still very low, namely less than 15. In the absence of large size platforms, quantum computer simulation is critical for developing and testing quantum algorithms and investigating the different challenges facing the design of quantum computer hardware. What makes quantum computer simulation on classical computers particularly challenging are the memory and computational resource requirements. In this paper, we introduce a universal quantum computer simulator, called QX, that takes as input a specially designed quantum assembly language, called QASM, and provides, through agressive optimisations, high simulation speeds and large number of qubits. QX allows the simulation of up to 34 fully entangled qubits on a single node using less than 270 GB of memory. Our experiments using different quantum algorithms show that QX achieves significant simulation speedup over similar state-of-theart simulation environment.
