Optimization of memory usage in quantum computing simulation

• Ano de defesa: 2018
• Autor(a) principal: Nascimento, Mateus Moreira Silveira do
• Orientador(a): Reiser, Renata Hax Sander
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Pelotas
• Programa de Pós-Graduação: Programa de Pós-Graduação em Computação
• Departamento: Centro de Desenvolvimento Tecnológico
• País: Brasil
• Palavras-chave em Português: Quantum computing; Quantum simulation; Memory pattern; D-GM environment; Quantum transformations; Quantum states; Computação quântica; Simulação quântica; Padrões de memória; Ambiente D-GM; Transformadas quânticas; Estados quânticos
• Área do conhecimento CNPq: CNPQ::CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO
• Link de acesso: http://guaiaca.ufpel.edu.br/handle/prefix/4355

Quantum simulation is the most common method to study quantum computing due to the lack of quantum hardware. Even so, simulators and simulations should be properly built to use the max computational power possible. One of the main problems of these assumptions comes from the memory, representing a bottleneck not just for quantum simulations, but for computer science research as a whole. In this dissertation, we present a methodology with the intent to reduce the usage of memory, being able to mitigate the effects of this disparity of performance between memory and others computers peripherals. Through this work, we show how we used our methodology within the D-GM simulator in order to improve the second and show the viability of improving memory usage without having to focus on hardware potential. Furthermore, we are able to achieve improvements about memory, and overall speed of quantum circuits simulation, thus showing our methodology can be applied not only for quantum simulation, but researches which need to develop better memory access pattern. Finally, regarding the D-GM simulator, we discuss how we made better functions to controlled operators, such as dense, primary and secondary diagonal patterns; even though we could not fix memory performance issues, we managed to enhance quantum simulations.