scholarly journals Always-On Sub-Microwatt Spiking Neural Network Based on Spike-Driven Clock- and Power-Gating for an Ultra-Low-Power Intelligent Device

2021 ◽  
Vol 15 ◽  
Author(s):  
Pavan Kumar Chundi ◽  
Dewei Wang ◽  
Sung Justin Kim ◽  
Minhao Yang ◽  
Joao Pedro Cerqueira ◽  
...  
Keyword(s):  
Neural Network ◽  
Power Consumption ◽  
Low Power ◽  
Power Dissipation ◽  
State Of The Art ◽  
Spiking Neural Network ◽  
Power Gating ◽  
Ultra Low Power ◽  
Fine Grained ◽  
Input Activity

This paper presents a novel spiking neural network (SNN) classifier architecture for enabling always-on artificial intelligent (AI) functions, such as keyword spotting (KWS) and visual wake-up, in ultra-low-power internet-of-things (IoT) devices. Such always-on hardware tends to dominate the power efficiency of an IoT device and therefore it is paramount to minimize its power dissipation. A key observation is that the input signal to always-on hardware is typically sparse in time. This is a great opportunity that a SNN classifier can leverage because the switching activity and the power consumption of SNN hardware can scale with spike rate. To leverage this scalability, the proposed SNN classifier architecture employs event-driven architecture, especially fine-grained clock generation and gating and fine-grained power gating, to obtain very low static power dissipation. The prototype is fabricated in 65 nm CMOS and occupies an area of 1.99 mm2. At 0.52 V supply voltage, it consumes 75 nW at no input activity and less than 300 nW at 100% input activity. It still maintains competitive inference accuracy for KWS and other always-on classification workloads. The prototype achieved a power consumption reduction of over three orders of magnitude compared to the state-of-the-art for SNN hardware and of about 2.3X compared to the state-of-the-art KWS hardware.

scholarly journals Adiabatic superconducting cells for ultra-low-power artificial neural networks

2016 ◽  
Vol 7 ◽  
pp. 1397-1403 ◽  
Author(s):  
Andrey E Schegolev ◽  
Nikolay V Klenov ◽  
Igor I Soloviev ◽  
Maxim V Tereshonok
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Artificial Neural Networks ◽  
Low Power ◽  
Power Dissipation ◽  
Radial Basis Function Networks ◽  
Activation Functions ◽  
Ultra Low Power ◽  
Network Algorithms ◽  
Low Power Dissipation

We propose the concept of using superconducting quantum interferometers for the implementation of neural network algorithms with extremely low power dissipation. These adiabatic elements are Josephson cells with sigmoid- and Gaussian-like activation functions. We optimize their parameters for application in three-layer perceptron and radial basis function networks.

scholarly journals Subthreshold Region based Linear Feedback Shift Register

2019 ◽  
Vol 8 (6S2) ◽  
pp. 817-821
Keyword(s):  
Power Consumption ◽  
Low Power ◽  
Power Dissipation ◽  
High Speed ◽  
System Level ◽  
Linear Feedback ◽  
Portable Devices ◽  
Subthreshold Region ◽  
Flip Flop ◽  
Ultra Low Power

Growing demand for portable devices and fast increases in complexity of chip cause power dissipation is an important parameter. Power consumption and dissipation or generations of more heat possess a restriction in the direction of the integration of more transistors. Several methods have been proposed to reduce power dissipation from system level to device level. Subthreshold circuits are widely used in more advanced applications due to ultra low-power consumption. The present work targets on construction of linear feedback shift registers (LFSR) in weak inversion region and their performance observed in terms of parameters like power delay product (PDP). In CMOS circuits subthreshold region of operation allows a low-power for ample utilizations but this advantage get with the penalty of flat speed. For the entrenched and high speed applications, improving the speed of subthreshold designs is essential. To enhance this, operate the devices at maximum current over capacitance. LFSR architectures build with various types of D flip flop and XOR gate circuits are analyzed. Circuit level Simulation is carried out using 130 nm technologies.

Analytical Modeling of D.C. Parameters of Double Gate Junctionless MOSFET in Near and Subthreshold Regime for RF Circuit Application

2020 ◽  
Vol 10 (4) ◽  
pp. 457-470 ◽  
Author(s):  
Dipanjan Sen ◽  
Savio J. Sengupta ◽  
Swarnil Roy ◽  
Manash Chanda ◽  
Subir K. Sarkar
Keyword(s):  
Power Consumption ◽  
Low Power ◽  
Circuit Design ◽  
Power Dissipation ◽  
Supply Voltage ◽  
Dominant Factor ◽  
Double Gate ◽  
Cmos Inverter ◽  
Ultra Low Power ◽  
Voltage Swing

Aims:: In this work, a Junction-Less Double Gate MOSFET (JLDG MOSFET) based CMOS inverter circuit is proposed for ultra-low power applications in the near and sub-threshold regime operations. Background:: D.C. performances like power, delay and voltage swing of the proposed Inverter have been modeled analytically and analyzed in depth. JLDG MOSFET has promising features to reduce the short-channel effects compared to the planner MOSFET because of better gate control mechanism. So, proposed Inverter would be efficacious to offer less power dissipation and higher speed. Objective:: Impact of supply voltage, temperature, High-k gate oxide, TOX, TSI on the power, delay and voltage swing of the Inverter circuits have been detailed here. Methods: Extensive simulations using SILVACO ATLAS have been done to validate the proposed logic based digital circuits. Besides, the optimum supply voltage has been modelled and verified through simulation for low voltage operations. In depth analysis of voltage swing is added to measure the noise immunity of the proposed logic based circuits in Sub & Near-threshold operations. For ultra-low power operation, JLDG MOSFET can be an alternative compared to conventional planar MOSFET. Result:: Hence, the analytical model of delay, power dissipation and voltage swing have been proposed of the proposed logic based circuits. Besides, the ultra-low power JLDG CMOS inverter can be an alternative in saving energy, reduction of power consumption for RFID circuit design where the frequency range is a dominant factor. Conclusion:: The power consumption can be lowered in case of UHF, HF etc. RF circuits using the Double Gate Junction-less MOSFET as a device for circuit design.

scholarly journals Ultra-Low Power and High-Throughput SRAM Design to Enhance AI Computing Ability in Autonomous Vehicles

Electronics ◽  
2021 ◽  
Vol 10 (3) ◽  
pp. 256
Author(s):  
Youngbae Kim ◽  
Shreyash Patel ◽  
Heekyung Kim ◽  
Nandakishor Yadav ◽  
Kyuwon Ken Choi
Keyword(s):  
Integrated Circuits ◽  
Power Consumption ◽  
Low Power ◽  
High Throughput ◽  
Autonomous Vehicles ◽  
Field Effect Transistors ◽  
State Of The Art ◽  
Electrical Performance ◽  
Large Area ◽  
Ultra Low Power

Power consumption and data processing speed of integrated circuits (ICs) is an increasing concern in many emerging Artificial Intelligence (AI) applications, such as autonomous vehicles and Internet of Things (IoT). Existing state-of-the-art SRAM architectures for AI computing are highly accurate and can provide high throughput. However, these SRAMs have problems that they consume high power and occupy a large area to accommodate complex AI models. A carbon nanotube field-effect transistors (CNFET) device has been reported as a potential candidates for AI devices requiring ultra-low power and high-throughput due to their satisfactory carrier mobility and symmetrical, good subthreshold electrical performance. Based on the CNFET and FinFET device’s electrical performance, we propose novel ultra-low power and high-throughput 8T SRAMs to circumvent the power and the throughput issues in Artificial Intelligent (AI) computation for autonomous vehicles. We propose two types of novel 8T SRAMs, P-Latch N-Access (PLNA) 8T SRAM structure and single-ended (SE) 8T SRAM structure, and compare the performance with existing state-of-the-art 8T SRAM architectures in terms of power consumption and speed. In the SRAM circuits of the FinFET and CNFET, higher tube and fin numbers lead to higher operating speed. However, the large number of tubes and fins can lead to larger area and more power consumption. Therefore, we optimize the area by reducing the number of tubes and fins without compromising the memory circuit speed and power. Most importantly, the decoupled reading and writing of our new SRAMs cell offers better low-power operation due to the stacking of device in the reading part, as well as achieving better readability and writability, while offering read Static Noise Margin (SNM) free because of isolated reading path, writing path, and greater pull up ratio. In addition, the proposed 8T SRAMs show even better performance in delay and power when we combine them with the collaborated voltage sense amplifier and independent read component. The proposed PLNA 8T SRAM can save 96%, while the proposed SE 8T SRAM saves around 99% in writing power consumption compared with the existing state-of-the-art 8T SRAM in FinFET model, as well as 99% for writing operation in CNFET model.

scholarly journals Design and Analysis of Low Power Dual Edge Triggered Mechanism Flip-Flop Employing Power Gating Methodology

2020 ◽  
Vol 6 (01) ◽  
pp. 26-31
Author(s):  
Prof. Virendra Umale
Keyword(s):  
Power Consumption ◽  
Low Power ◽  
Reduction Technique ◽  
Power Gating ◽  
Flip Flop ◽  
Ultra Low Power ◽  
Efficient Power

The advancement of battery operated designs has abundantly increases the memory elements and registers to be operated in ultra-low power. That is the this paper we have proposed a design of CT_C DET flip-flop with power gating technique which is the most efficient power consuming reduction technique.  The design of the power gating technique involves the pull-up transistor in the Vdd of the circuit and pull-down transistor in the ground terminal. This power gating technique reduces the power consumption by more than 40% than that of the existing design.

scholarly journals Robust neuromorphic coupled oscillators for adaptive pacemakers

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Renate Krause ◽  
Joanne J. A. van Bavel ◽  
Chenxi Wu ◽  
Marc A. Vos ◽  
Alain Nogaret ◽  
...  
Keyword(s):  
Neural Network ◽  
Low Power ◽  
Coupled Oscillators ◽  
Cardiac Pacemaker ◽  
Processing System ◽  
Spiking Neural Network ◽  
Ultra Low Power ◽  
Mixed Signal ◽  
Surface Ecg ◽  
Electronic Hardware

AbstractNeural coupled oscillators are a useful building block in numerous models and applications. They were analyzed extensively in theoretical studies and more recently in biologically realistic simulations of spiking neural networks. The advent of mixed-signal analog/digital neuromorphic electronic circuits provides new means for implementing neural coupled oscillators on compact, low-power, spiking neural network hardware platforms. However, their implementation on this noisy, low-precision and inhomogeneous computing substrate raises new challenges with regards to stability and controllability. In this work, we present a robust, spiking neural network model of neural coupled oscillators and validate it with an implementation on a mixed-signal neuromorphic processor. We demonstrate its robustness showing how to reliably control and modulate the oscillator’s frequency and phase shift, despite the variability of the silicon synapse and neuron properties. We show how this ultra-low power neural processing system can be used to build an adaptive cardiac pacemaker modulating the heart rate with respect to the respiration phases and compare it with surface ECG and respiratory signal recordings from dogs at rest. The implementation of our model in neuromorphic electronic hardware shows its robustness on a highly variable substrate and extends the toolbox for applications requiring rhythmic outputs such as pacemakers.

An Ultra-Low-Power Five-Input Majority Gate in Quantum-Dot Cellular Automata

2020 ◽  
Vol 29 (11) ◽  
pp. 2050176
Author(s):  
Feifei Deng ◽  
Guangjun Xie ◽  
Shaowei Wang ◽  
Xin Cheng ◽  
Yongqiang Zhang
Keyword(s):  
Cellular Automata ◽  
Power Consumption ◽  
Low Power ◽  
Power Dissipation ◽  
Single Layer ◽  
Full Adder ◽  
Majority Gate ◽  
Ultra Low Power ◽  
Quantum Dot Cellular Automata ◽  
Number Of Cells

Quantum-dot cellular automata (QCA) is a highly attractive alternative to CMOS for future digital circuit design, relying on its high-performance and low-power-consumption features. This paper analyzes and compares previously published five-input majority gates. These designs do not perform well in terms of physical properties, especially concerting power consumption. Therefore, an ultra-low-power five-input majority gate in one layer is proposed, which uses a minimum number of cells and smaller area, and achieves the expected highly polarized output compared with previous designs. In order to evaluate its practicability, a new one-bit coplanar full-adder is proposed. The analysis results show that this full-adder performs well compared with existing multilayer and single-layer designs. The number of cells of the proposed design is reduced by 7.14% to get the same area and clock delay compared with the best coplanar full-adder. In addition, its power dissipation is also reduced by 9.28% at 0.5[Formula: see text], 11.09% at 1[Formula: see text] and 12.66% at 1.5[Formula: see text] in terms of average energy dissipation compared with the best single-layer design. QCADesigner tool is used to verify the simulation results of the proposed designs and QCAPro tool is used to evaluate the power dissipation of all considered designs.

Sign in / Sign up

Export Citation Format

Share Document