How nanosecond magnetization dynamics during spin-Hall switching of in-plane MTJs enables a cryogenic memory cell with superconducting line drivers

Invited

Abstract

A scarcity of advantageous thermal fluctuations typically prevents magnetization reversal of in-plane magnetic memory elements from being both fast and reliable, a problem that is only exacerbated at lower temperatures. We show that three-terminal spin-Hall effect (SHE) memories are not subject to such constraints. By virtue of the combined action of spin-torque and an Oersted field torque from the metallic write channel, these devices can be driven at fast switching times (< 1-2 ns) and low error rates ( < 10-6) that persist down to T=4 K.

The preservation of these switching characteristics makes SHE memories a perfect candidate for integration with low temperature computing schemes. One such scheme, based on single-flux-quantum (SFQ) operation of Josephson-Junction (JJ) circuits, could reduce the enormous power consumption of today’s large-scale computing facilities. This technology has historically lacked a robust and dense memory, such as the SHE three terminal devices, that can be operated at low temperatures.

Here we demonstrate the operation of a memory cell that can be interfaced directly to SFQ logic. The amplitudes of SFQ pulses are insufficient to directly enact switching, so we must instead implement a line driver that can be triggered by SFQ pulses. Advances in SHE materials have brought write currents to around 120 μA, while further advances coupled with device scaling should yield currents an order of magnitude lower. Such currents are easily sourced by superconducting nano-cryotrons (nTrons), which replace access transistors in our cryogenic memory architecture. We show bipolar switching of SHE memory elements with nTrons devices at T=4 K, taking the first steps on the way to an SFQ-compatible main memory.

Presenters

  • Graham Rowlands

    Raytheon BBN Technologies, BBN Technology - Massachusetts

Authors

  • Graham Rowlands

    Raytheon BBN Technologies, BBN Technology - Massachusetts

  • Emily Toomey

    Massachusetts Institute of Technology

  • Andrew Wagner

    Raytheon BBN Technologies

  • Guilhem Ribeill

    Raytheon BBN Technologies

  • Leonardo Ranzani

    Raytheon BBN Technologies

  • Minh-Hai Nguyen

    Cornell University

  • Shengjie Shi

    Cornell University

  • Sriharsha Aradhya

    Cornell University

  • Andrew Dane

    Massachusetts Institute of Technology

  • Karl Berggren

    Electrical Engineering and Computer Science, Massachusetts Inst of Tech-MIT, Massachusetts Institute of Technology

  • Robert Buhrman

    Cornell University, School of Applied and Engineering Physics, Cornell University

  • Thomas Ohki

    Raytheon BBN Technologies