报告题目：Combining the Best of Two Worlds: Near-Term Quantum Computing and Quantum Monte Carlo
报告人：黄翼飞 博士
报告人单位：字节跳动量子计算研究员
主持人：瞿岸 研究员
报告时间：5月22日（周三）下午2:00
地点：闵行校区光学大楼B225会议室
报告人简介：
Yifei Huang is currently a researcher scientist at Bytedance Research, interested in second-quantized quantum Monte Carlo methods as well as NISQ and fault-tolerant quantum applications in quantum chemistry. Previously he obtained his PhD at Tufts University working with Prof. Peter Love on classical simulation of quantum computation and bachelor's degree in quantum optics from University of Science and Technology of China.
报告人摘要：
In this talk I will describe a new type of hybrid quantum algorithms, which we dub quantum computing quantum Monte Carlo, namely combining quantum Monte Carlo with quantum algorithms. I will talk about previous effort on this topic by the Google Quantum AI team as well as its drawbacks. Then I will introduce the basic procedure of Full Configuration Interaction Quantum Monte Carlo (FCIQMC) and our new algorithm that suppresses the sign problem of FCIQMC by taking advantage of the states prepared from a quantum device. I argue that FCIQMC is a better fit when incorporating into quantum algorithms than AFQMC. This hybrid algorithm helps reducing the depth of quantum circuits, thus minimizes the noise effect. But more importantly, this hybrid framework could potentially push forward the limit of FCIQMC and therefore providing a viable path towards practical quantum advantage.