TY - JOUR
T1 - Limited Stochastic Current for Energy-Optimized Switching of Spin-Transfer-Torque Magnetic Random-Access Memory
AU - Baek, Eunchong
AU - Purnama, Indra
AU - You, Chun Yeol
N1 - Publisher Copyright:
© 2019 American Physical Society.
PY - 2019/12/3
Y1 - 2019/12/3
N2 - The switching of spin-transfer-torque magnetic random-access memory (STT MRAM) in the simple macrospin model is determined by the amplitude and pulse duration of the applied current, and it requires a current that is higher than a critical current, Ic. However, this critical current misses one fundamental physical issue for the commercialization of STT MRAM; the so-called nonswitching probability (PNS) or write soft-error rate (WSER), which is influenced by the stochastic nature of the switching process at finite temperature. Herein, we propose a limited stochastic switching (LSS) current, which is another definition for the critical current with the PNS incorporated. The definition of the LSS current and the analytical expressions are obtained by solving the Fokker-Planck equation with a given specific PNS value. Most importantly, by using the LSS current and optimizing it together with the related pulse-duration time, we find the optimum combination of current amplitude and pulse duration, which may reduce the energy consumption of the STT MRAM by up to 75%.
AB - The switching of spin-transfer-torque magnetic random-access memory (STT MRAM) in the simple macrospin model is determined by the amplitude and pulse duration of the applied current, and it requires a current that is higher than a critical current, Ic. However, this critical current misses one fundamental physical issue for the commercialization of STT MRAM; the so-called nonswitching probability (PNS) or write soft-error rate (WSER), which is influenced by the stochastic nature of the switching process at finite temperature. Herein, we propose a limited stochastic switching (LSS) current, which is another definition for the critical current with the PNS incorporated. The definition of the LSS current and the analytical expressions are obtained by solving the Fokker-Planck equation with a given specific PNS value. Most importantly, by using the LSS current and optimizing it together with the related pulse-duration time, we find the optimum combination of current amplitude and pulse duration, which may reduce the energy consumption of the STT MRAM by up to 75%.
UR - http://www.scopus.com/inward/record.url?scp=85077181436&partnerID=8YFLogxK
U2 - 10.1103/PhysRevApplied.12.064004
DO - 10.1103/PhysRevApplied.12.064004
M3 - Article
AN - SCOPUS:85077181436
SN - 2331-7019
VL - 12
JO - Physical Review Applied
JF - Physical Review Applied
IS - 6
M1 - 064004
ER -