%0 Journal Article %1 das2018ultralow %A Das, B. %A Schulze, J. %A Ganguly, U. %D 2018 %J IEEE Electron Device Letters %K iht j.schulze.iht journal %N 12 %P 1832-1835 %R 10.1109/LED.2018.2876684 %T Ultra-Low Energy LIF Neuron Using Si NIPIN Diode for Spiking Neural Networks %U https://ieeexplore.ieee.org/document/8496854/ %V 39 %X An energy efficient neuron is essential for spiking neural network (SNN) to operate at low energy to mimic the human brain functionalities in hardware. Several CMOS-based Si transistors, memory devices, spintronic devices have been used as a neuron for SNN. However, the main concern is the energy efficiency for these neurons. In this letter, we experimentally demonstrate a Si-based CMOS compatible asymmetric NIPIN diode as a LIF neuron. First, we demonstrate the LIF neuron characteristics by comparing the spike-frequency (f) versus voltage curve with that of a simple LIF neuron model. This neuron shows a classical ReLU behavior, which is attractive for typical software neuron models. Then, we show an ultra-low energy consumption of<inline-formula><tex-math notation="LaTeX">$210^-17 J$</tex-math></inline-formula>per spike at 10-nm node of this neuron, as NIPIN diode is highly scalable (<inline-formula><tex-math notation="LaTeX">$4F^2$</tex-math></inline-formula>) due to its capacitorless structure. This is the lowest reported energy/spike for any LIF neuron for SNN application. Thus, the NIPIN is suitable for ultra-low energy LIF neuron application for energy efficient SNN.

@article{das2018ultralow, abstract = {An energy efficient neuron is essential for spiking neural network (SNN) to operate at low energy to mimic the human brain functionalities in hardware. Several CMOS-based Si transistors, memory devices, spintronic devices have been used as a neuron for SNN. However, the main concern is the energy efficiency for these neurons. In this letter, we experimentally demonstrate a Si-based CMOS compatible asymmetric NIPIN diode as a LIF neuron. First, we demonstrate the LIF neuron characteristics by comparing the spike-frequency (f) versus voltage curve with that of a simple LIF neuron model. This neuron shows a classical ReLU behavior, which is attractive for typical software neuron models. Then, we show an ultra-low energy consumption of<inline-formula><tex-math notation="LaTeX">$\sim \text {2}\times \text {10}^{-\text {17}} {J}$</tex-math></inline-formula>per spike at 10-nm node of this neuron, as NIPIN diode is highly scalable (<inline-formula><tex-math notation="LaTeX">$\text {4}{F}^{\text {2}}$</tex-math></inline-formula>) due to its capacitorless structure. This is the lowest reported energy/spike for any LIF neuron for SNN application. Thus, the NIPIN is suitable for ultra-low energy LIF neuron application for energy efficient SNN.}, added-at = {2018-11-26T12:49:11.000+0100}, author = {Das, B. and Schulze, J. and Ganguly, U.}, biburl = {https://puma.ub.uni-stuttgart.de/bibtex/2e54c71a945f999824a6cc89ea4323739/ihtpublikation}, doi = {10.1109/LED.2018.2876684}, interhash = {41b0991f14e4321eefe39d69a2fe566c}, intrahash = {e54c71a945f999824a6cc89ea4323739}, issn = {0741-3106}, journal = {IEEE Electron Device Letters}, keywords = {iht j.schulze.iht journal}, month = dec, number = 12, pages = {1832-1835}, timestamp = {2018-11-26T11:49:11.000+0100}, title = {Ultra-Low Energy LIF Neuron Using Si NIPIN Diode for Spiking Neural Networks}, url = {https://ieeexplore.ieee.org/document/8496854/}, volume = 39, year = 2018 }