The promise of spintronics for unconventional computing

Finocchio, G.; Ventra, Massimiliano Di; Çamsarı, Kerem Y.; Everschor-Sitte, Karin; Amiri, Pedram Khalili; Zeng, Zhongming

doi:10.1016/j.jmmm.2020.167506

Cited by 101 publications

(70 citation statements)

References 94 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Neuromorphic computing is regarded as a form of unconventional computing that emulates a biological brain on different levels of computing hierarchy. [ 1–7 ] Traditionally, computers are based on the von Neumann architecture, [ 8 ] where information is continuously exchanged between memory units and the central processing unit. In contrast, under neuromorphic computing, information is processed in a decentralized manner in proximity to the memory.…”

Section: Introductionmentioning

confidence: 99%

Spin–Orbit Torque‐Induced Domain Nucleation for Neuromorphic Computing

et al. 2021

View full text Add to dashboard Cite

Neuromorphic computing has become an increasingly popular approach for artificial intelligence because it can perform cognitive tasks more efficiently than conventional computers. However, it remains challenging to develop dedicated hardware for artificial neural networks. Here, a simple bilayer spintronic device for hardware implementation of neuromorphic computing is demonstrated. In L11‐CuPt/CoPt bilayer, current‐inducted field‐free magnetization switching by symmetry‐dependent spin–orbit torques shows a unique domain nucleation‐dominated magnetization reversal, which is not accessible in conventional bilayers. Gradual domain nucleation creates multiple intermediate magnetization states which form the basis of a sigmoidal neuron. Using the L11‐CuPt/CoPt bilayer as a sigmoidal neuron, the training of a deep learning network to recognize written digits, with a high recognition rate (87.5%) comparable to simulation (87.8%) is further demonstrated. This work offers a new scheme of implementing artificial neural networks by magnetic domain nucleation.

show abstract

Section: Introductionmentioning

confidence: 99%

Spin–Orbit Torque‐Induced Domain Nucleation for Neuromorphic Computing

et al. 2021

View full text Add to dashboard Cite

show abstract

“…We next demonstrate how the metrics apply to a physical system based on magnetic skyrmions [14]. Skyrmion and other spintronic based reservoirs [7,9,15,16] are of particular interest due to their ability to integrate into existing CMOS devices [17], their nanoscale size, and their responsiveness to low power excitations.…”

Section: B Physical Skyrmion Reservoirmentioning

confidence: 99%

Task Agnostic Metrics for Reservoir Computing

Love¹,

Mulkers²,

Bourianoff³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Physical reservoir computing is a computational paradigm that enables temporal pattern recognition to be performed directly in physical matter. By exciting non-linear dynamical systems and linearly classifying their changes in state, we can create highly energy-efficient devices capable of solving machine learning tasks without the need to build a modular system consisting of millions of neurons interconnected by synapses. The chosen dynamical system must have three desirable properties: non-linearity, complexity, and fading memory to act as an effective reservoir. We present task agnostic quantitative measures for each of these three requirements and exemplify them for two reservoirs: an echo state network and a simulated magnetic skyrmion-based reservoir. We show that, in general, systems with lower damping reach higher values in all three performance metrics. Whilst for input signal strength, there is a natural trade-off between memory capacity and non-linearity of the reservoir's behaviour. In contrast to typical task-dependent reservoir computing benchmarks, these metrics can be evaluated in parallel from a single input signal, drastically speeding up the parameter search to design efficient and high-performance reservoirs.

show abstract

“…The concept of SC is well known and extensively studied. CMOS [7,8], memristor [9][10][11] and spintronics [12][13][14] based SC architectures have already been demonstrated in the past. However, CMOS-based SC architectures require several hundred transistors to generate s-bits, which limits its area and energy efficiency [19].…”

Section: Introductionmentioning

confidence: 99%

A Monolithic Stochastic Computing Architecture for Energy and Area Efficient Arithmetic

Ravichandran

Zheng

Schranghamer

et al. 2022

Preprint

View full text Add to dashboard Cite

As the energy and hardware investments necessary for conventional high-precision digital computing continues to explode in the emerging era of artificial intelligence, deep learning, and Big-data [1-4], a change in paradigm that can trade precision for energy and resource efficiency is being sought for many computing applications. Stochastic computing (SC) is an attractive alternative since unlike digital computers, which require many logic gates and a high transistor volume to perform basic arithmetic operations such as addition, subtraction, multiplication, sorting etc., SC can implement the same using simple logic gates [5, 6]. While it is possible to accelerate SC using traditional silicon complementary metal oxide semiconductor (CMOS) [7, 8] technology, the need for extensive hardware investment to generate stochastic bits (s-bit), the fundamental computing primitive for SC, makes it less attractive. Memristor [9-11] and spin-based devices [12-15] offer natural randomness but depend on hybrid designs involving CMOS peripherals for accelerating SC, which increases area and energy burden. Here we overcome the limitations of existing and emerging technologies and experimentally demonstrate a standalone SC architecture embedded in memory based on two-dimensional (2D) memtransistors. Our monolithic and non-von Neumann SC architecture consumes a miniscule amount of energy < 1 nano Joules for s-bit generation and to perform arithmetic operations and occupy small hardware footprint highlighting the benefits of SC.

show abstract

The promise of spintronics for unconventional computing

Cited by 101 publications

References 94 publications

Spin–Orbit Torque‐Induced Domain Nucleation for Neuromorphic Computing

Spin–Orbit Torque‐Induced Domain Nucleation for Neuromorphic Computing

Task Agnostic Metrics for Reservoir Computing

A Monolithic Stochastic Computing Architecture for Energy and Area Efficient Arithmetic

Contact Info

Product

Resources

About