Nanosecond True-Random-Number Generation with Superparamagnetic Tunnel Junctions: Identification of Joule Heating and Spin-Transfer-Torque Effects

A superparamagnetic tunnel junction (SMTJ), acting as a random stochastic noise source, is the counterpart to nonvolatile magnetic random access memory (MRAM). This study investigates the impact of applied current on superparamagnetic fluctuations in the SMTJ, influencing both the state probability and the average fluctuation rate due to the effects of spin-transfer torque and Joule heating. Moreover, this research demonstrates the potential of SMTJs as true random number generators, successfully passing all statistical randomness tests by using logic XOR gates, and achieving an average random bit generation rate of 200 Mbit/s.