Computational paradigm for nanoelectronics: self-assembled quantum dot cellular neural networks

Bandyopadhyay, Supriyo; Karahaliloglu, K.; Balkir, S.; Pramanik, Sandipan

doi:10.1049/ip-cds:20041175

Cited by 20 publications

(14 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The cellular automata proposed by Ikebe and Asai 108 is proposed to be realized as a CMOS circuit, therefore being easily compatible with at least nanoscale CMOS. The nanodot proposals by Morie et al 42 and Bandyopadhyay et al 40 are also compatible with CMOS due to having been produced on silicon.…”

Section: Comparison Of the Implementation Approachesmentioning

confidence: 96%

“…At best, some of the design principles or simulation parameters have backing from actual measurements: for example, in a 2005 paper by Bandyopadhyay et al 40 a self-assembled quantum dot cellular neural network is proposed where the inter-dot resistance, negative differential resistance, and dot capacitance have been measured for manufactured quantum dots.…”

Section: Architectures Implementing Machine Learningmentioning

confidence: 99%

“…Approaches using nanodots are the most advanced in the area of physical realization, Bandyopadhyay et al 40 having produced a partial prototype allowing the measurement of some parameters. Many approaches with cellular automata and cellular nonlinear networks can be completely regular.…”

Section: Comparison Of the Implementation Approachesmentioning

confidence: 99%

“…In nanotechnology, CNNs have been proposed to be physically implemented with quantum dots (QD, e.g., Refs. [62,40]), single electron transistors (SETs), 63 49 and resonant tunneling diodes (RTDs). 64 65 Nanoscale implementation of cellular automata by semiconductor QDs was proposed by Lent et al 66 More recent proposals of these Quantum Cellular Automata (QCA) are implementations by magnetic nanodots 67 and molecules.…”

Section: Cellular Neural Network and Cellular Automatamentioning

confidence: 99%

See 3 more Smart Citations

Machine Learning: How It Can Help Nanocomputing

Uusitalo¹,

Peltonen²,

Ryhänen³

2011

Jnl of Comp & Theo Nano

View full text Add to dashboard Cite

We examine possibilities for making advances in nanocomputing by bringing in ideas from the field of machine learning. The potential from combining machine learning with nanocomputing seems to be underutilized. We review three complementary approaches. Firstly, machine learning can be used in the different phases of developing complicated nanocomputing devices: in modeling, designing, constructing, and programming the devices. Secondly, machine learning methods implemented by nanocomputing hardware can be a competitive solution especially for specialized application areas like sensory information processing; working towards such implementations advances nanocomputing by guiding development of the nanocomponents and architectures required for such applications. Thirdly, nanotechnology enabled quantum computing can significantly increase our capacity to solve NP-complete optimization problems; although this increase is not specific to machine learning, several such problems occur in machine learning and artificial intelligence, hence solving such problems is a useful goal that partly motivates development of quantum computing. The main value of this paper is to provide new ideas for researchers working on nanocomputing, nanoarchitectures, development and design of nanoprocessors and other nanocomponents, or nanomanufacturing.

show abstract

Section: Comparison Of the Implementation Approachesmentioning

confidence: 96%

Section: Architectures Implementing Machine Learningmentioning

confidence: 99%

Section: Comparison Of the Implementation Approachesmentioning

confidence: 99%

Section: Cellular Neural Network and Cellular Automatamentioning

confidence: 99%

See 2 more Smart Citations

Machine Learning: How It Can Help Nanocomputing

Uusitalo¹,

Peltonen²,

Ryhänen³

2011

Jnl of Comp & Theo Nano

View full text Add to dashboard Cite

show abstract

“…Finally, it must be noted that though in this paper the NDR devices are realized with the intraband double barrier RTD, the CNN architecture proposed can be constructed with a host of other NDR devices [22], [23].…”

mentioning

confidence: 99%

Tunneling-Based Cellular Nonlinear Network Architectures for Image Processing

Mazumder

Ebong

2009

IEEE Trans. VLSI Syst.

View full text Add to dashboard Cite

Abstract-The resonant tunneling diode (RTD) has found numerous applications in high-speed digital and analog circuits due to the key advantages associated with its folded back negative differential resistance (NDR) current-voltage (I-V) characteristics as well as its extremely small switching capacitance. Recently, the RTD has also been employed to implement high-speed and compact cellular neural/nonlinear networks (CNNs) by exploiting its quantum tunneling induced nonlinearity and symmetrical I-V characteristics for both positive and negative voltages applied across the anode and cathode terminals of the RTD. This paper proposes an RTD-based CNN architecture and investigates its operation through driving-point-plot analysis, stability and settling time study, and circuit simulation. Full-array simulation of a 128 128 RTD-based CNN for several image processing functions is performed using the Quantum Spice simulator designed at the University of Michigan, where the RTD is represented in SPICE simulator by a physics based model derived by solving Schrödinger's and Poisson's equations self-consistently. A comparative study between different CNN implementations reveals that the RTD-based CNN can be designed superior to conventional CMOS technologies in terms of integration density, operating speed, and functionality.Index Terms-Resonant tunneling diode (RTD), cellular neural/ nonlinear network (CNN), full array simulation, settling time analysis.

show abstract

Light‐Stimulatable Molecules/Nanoparticles Networks for Switchable Logical Functions and Reservoir Computing

Viero

Guérin

Vladyka

et al. 2018

Adv Funct Materials

View full text Add to dashboard Cite

The fabrication and electron transport properties of nanoparticles selfassembled networks (NPSAN) of molecular switches (azobenzene derivatives) interconnected by Au nanoparticles are reported, and optically driven switchable logical operations associated to the light-controlled switching of the molecules are demonstrated. The switching yield is up to 74%. It is also demonstrated that these NPSANs are prone to light-stimulable reservoir computing. The complex nonlinearity of electron transport and dynamics in these highly connected and recurrent networks of molecular junctions exhibits rich high harmonics generation (HHG) required for reservoir computing approaches. Logical functions and HHG are controlled by the isomerization of the molecules upon light illumination. These results, without direct analogs in semiconductor devices, open new perspectives to molecular electronics in unconventional computing.paves the way for the processing of multi-input signals through a device that acts as a reservoir computer.The authors declare no conflict of interest.

show abstract

Computational paradigm for nanoelectronics: self-assembled quantum dot cellular neural networks

Cited by 20 publications

References 16 publications

Machine Learning: How It Can Help Nanocomputing

Machine Learning: How It Can Help Nanocomputing

Tunneling-Based Cellular Nonlinear Network Architectures for Image Processing

Light‐Stimulatable Molecules/Nanoparticles Networks for Switchable Logical Functions and Reservoir Computing

Contact Info

Product

Resources

About