Scalable Emulation of Sign-Problem–Free Hamiltonians with Room-Temperature p -bits

Abstract

A special class of many-body quantum systems can be simulated by probabilistic algorithms running on digital computers, but generating correlated random numbers is computationally expensive. The authors propose an asynchronous probabilistic coprocessor that uses a slightly modified cell structure in the emerging magnetoresistive RAM technology, which should accelerate such probabilistic algorithms by several orders of magnitude, in terms of sampling speed and energy. This approach complement existing efforts to simulate quantum systems by using scalable, room-temperature building blocks.