Search

Stochastic computing (SC) is a promising technology that can be used for low-cost hardware designs. However, SC suffers from its long latency. Although parallel processing can efficiently shorten the latency, duplicated stochastic number generators (SNGs) are necessary, which cause substantial hardware overhead. This paper proposes a scalable SNG based on the spin-Hall-effect (SHE), which is capable of generating multiple independent stochastic streams simultaneously. The design takes advantages of the efficient charge-to-spin conversion from the Spin-Hall material and the intrinsic stochasticity of nanomagnets. Compared to previous spintronic SNGs, the SHE-SNG can reduce the area by $1.6\times -7.8\times $ and the power by $4.9\times -13\times $ while increasing the degree of parallelism from 1 to 16. Compared to CMOS-based SNGs, the proposed SNG obtained $24\times -120\times $ and $53\times $ reduction in terms of area and power, respectively. Finally, three benchmarks were implemented, and the results indicate that SC implementations with the proposed SHE-SNG can achieve $1.2\times -29\times $ reduction of hardware resources compared to implementations with previous CMOS- and spintronic-based designs while scaling the degree of parallelism from 1 to 64. [ABSTRACT FROM AUTHOR]

Spin transfer torque (STT)-based switching of magnetic random access memories (MRAMs) has stimulated considerable research interest in recent years. The nonvolatility of STT-MRAMs, high areal density, and low static power dissipation makes them a strong contender for possible replacement of conventional silicon-based memories. However, a major bottleneck associated with STT-MRAMs is their high write current requirement. Recently, it has been demonstrated that the high write current of the STT switching mechanism can be improved by using other magnetic switching techniques like voltage-controlled magnetic anisotropy, spin Hall effect, and magnetoelectric effect. Interestingly, each of these new physics increases the design space to be explored and presents interesting tradeoffs with respect to read and write mechanisms. It is, therefore, important to analyze these devices with respect to their readability, writability, and areal density in a unified framework that is based on the state-of-the-art experimental results. Toward that end, in this paper, we present an overview of various physics underlying these magnetic memories and discuss associated bit-cell circuitry. Furthermore, we present a holistic comparative analysis based on detailed numerical simulations highlighting the various design tradeoffs. [ABSTRACT FROM AUTHOR]