Energy-efficient spintronic based neuromorphic computing system using current mode track and termination circuit

Pegah Shafaghi, Mehdi Dolatshahi, Hooman Farkhani

Publikation: Bidrag til tidsskrift/Konferencebidrag i tidsskrift /Bidrag til avisTidsskriftartikelForskningpeer review

4 Citationer (Scopus)

Abstract

Spintronic nano-devices have shown great potential to reduce the energy consumption of neuromorphic computing systems (NCSs). In the spintronic-based NCSs, the switching or oscillation of a magnetic tunnel junction (MTJ) is a common approach to mimic neuron firing. However, still there is a gap between the performance (operation/sec/Watt/cm 2) of the human brain and NCSs. To mitigate this gap, it is essential to further decrease the energy consumption and the delay of the NCS. The high-energy consumption of the MTJ-based NCS is mostly related to the high current needed to switch the MTJ state. Hence, some previous methods tried to perform real-time tracking of the MTJ state by monitoring the voltage across the MTJ and cut off its current immediately after switching. However, due to the small voltage changes after switching, these methods suffer from high-power consumption. In this article, a new method based on the tracking of the MTJ current (instead of its voltage) and terminating the MTJ current after switching is proposed. Due to the large changes in the MTJ current after switching (about 40%), there is no need to use an amplifier in the proposed common-mode tracking and terminating circuit (CM-TTC). The simulation results in 65-nm CMOS technology confirm that the proposed CM-TTC technique can improve the energy consumption and speed of a typical NCS by 53% and 2X. Moreover, the power consumption, delay, and area overhead of CM-TTC is reduced by 12.8%, 73%, and 95% compared with the best state-of-the-art track and termination circuits.

OriginalsprogEngelsk
TidsskriftIEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems
Vol/bind42
Nummer9
Sider (fra-til)2915-2923
Antal sider9
ISSN0278-0070
DOI
StatusUdgivet - sep. 2023

Fingeraftryk

Dyk ned i forskningsemnerne om 'Energy-efficient spintronic based neuromorphic computing system using current mode track and termination circuit'. Sammen danner de et unikt fingeraftryk.

Citationsformater