Data-stationary Architecture to Execute Quantum Algorithms Classically

– This paper presents a data stationary architecture in which each word has an attached address field. Address fields massively update in parallel to record data interchanges. Words do not move until memory is read for post processing. A sea of such cells can test large-scale quantum algorithms, although other programming is possible. 1. INTRODUCTION Quantum algorithms are better than classical algorithms for certain applications, for example function identification [1, 2]. Quantum algorithms generally have three parts, pre processing, logic implementation, and post processing. Pre processing usually forms a non-sparse state vector whose entries are to be processed in parallel. Logic can be interpreted to mean data interchanges within the state vector as specified by the steps in a 'wiring' diagram [3, 4]. Output processing can use various (real time) digital filtering methods, including Hadamard or Fourier transform.