Search NASA⌕ Search

SEARCH · Search NASA

Results for “CMOS”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,009 records · Page 56

PABO: Pseudo Agent-Based Multi-Objective Bayesian Hyperparameter Optimization for Efficient Neural Accelerator Design

The ever increasing computational cost of Deep Neural Networks (DNN) and the demand for energy efficient hardware for DNN acceleration has made accuracy and hardware cost co-optimization for DNNs tremendously important, especially for edge devices. Owing to the large parameter space and cost of evaluating each parameter in the search space, manually tuning of DNN hyperparameters is impractical. Automatic joint DNN and hardware hyperparameter optimization is indispensable for such problems. Bayesian optimization-based approaches have shown promising results for hyperparameter optimization of DNNs. However, most of these techniques have been developed without considering the underlying hardware, thereby leading to inefficient designs. Further, the few works that perform joint optimization are not generalizable and mainly focus on CMOS-based architectures. In this work, we present a novel pseudo agent-based multiobjective hyperparameter optimization (PABO) for maximizing the DNN performance while obtaining low hardware cost. Compared to the existing methods, our work poses a theoretically different approach for joint optimization of accuracy and hardware cost and focuses on memristive crossbar based accelerators. PABO uses a supervisor agent to establish connections between the posterior Gaussian distribution models of network accuracy and hardware cost requirements. The agent reduces the mathematical complexity of the co-optimization problem by removing unnecessary computations and updates of acquisition functions, thereby achieving significant speed-ups for the optimization procedure. PABO outputs a Pareto frontier that underscores the trade-offs between designing high-accuracy and hardware efficiency. Our results demonstrate a superior performance compared to the state-of-the-art methods both in terms of accuracy and computational speed (~100x speed up).

Parsa, Maryam↗

Experimental Demonstration of Ferroelectric Spiking Neurons for Unsupervised Clustering

© 2018 IEEE. We report the first experimental demonstration of ferroelectric field-effect transistor (FEFET) based spiking neurons. A unique feature of the ferroelectric (FE) neuron demonstrated herein is the availability of both excitatory and inhibitory input connections in the compact 1T-1FEFET structure, which is also reported for the first time for any neuron implementations. Such dual neuron functionality is a key requirement for bio-mimetic neural networks and represents a breakthrough for implementation of the third generation spiking neural networks (SNNs) - also reported herein for unsupervised learning and clustering on real world data for the first time. The key to our demonstration is the careful design of two important device level features: (1) abrupt hysteretic transitions of the FEFET with no stable states therein, and (2) the dynamic tunability of the FEFET hysteresis by bias conditions which allows for the inhibition functionality. Experimentally calibrated, multi-domain Preisach based FEFET models were used to accurately simulate the FE neurons and project their performance at scaled nodes. We also implement an SNN for unsupervised clustering and benchmark the network performance across analog CMOS and emerging technologies and observe (1) unification of excitatory and inhibitory neural connections, (2) STDP based learning, (3) lowest reported power (3.6nW) during classification, and (4) a classification accuracy of 93%.

Wang, Z↗

First Demonstration of Vertical 2T-nC FeRAM Hybrid Cell and its Scalability for High-Density 3D Ferroelectric Capacitor Memory

In this article, we perform a comprehensive experimental and modeling study into the scaling of vertical 2T-nC ferroelectric random-access memory (FeRAM) hybrid cell to demonstrate a high performance and high-density 3D capacitor memory. We demonstrate: i) first time successful integration of the vertical 2T-3C FeRAM cell by stacking the vertical metal-ferroelectricmetal (MFM) stack on top of Si CMOS transistors; ii) successful experimental operation of the memory cell, including the quasi-nondestructive read out (QNRO) of the polarization without write back after 106 read cycles; iii) the write bit line (WBL) heavily screens the coupling between neighboring strings, making it a minor concern; V ) aggressive stacking of the WBLs, i.e., number of MFMs in a string, could facilitate the self-boosting during write operation due to ferroelectric linear capacitance (CFE), which allows self-boosted inhibition for Vw/2 scheme and worsens the Vw/3 scheme as disturb increases to intolerable 2Vw/3; v) aggressive horizontal scaling significantly increases the read disturb to cells on neighboring planes due to capacitance between two WBLs (Cz).

42 ENGINEERING↗

Foundry-fabricated silicon source of broadband polarization entanglement

We present a CMOS-fabricated photonic integrated circuit tailored to the generation of ultrabroadband polarization-entangled photons. This fully on-chip source leverages a novel design combining polarization splitter-rotators and a bidirectionally pumped microring resonator to produce high-fidelity polarization Bell states over 115 frequency-bin pairs spaced 38.4 GHz apart.

Miloshevsky, Alex↗

Investigating the Interplay between Energy Efficiency and Resilience in High Performance Computing

Energy efficiency and resilience are two crucial challenges for HPC systems to reach exascale. While energy efficiency and resilience issues have been extensively studied individually, little has been done to understand the interplay between energy efficiency and resilience for HPC systems. Decreasing the supply voltage associated with a given operating frequency for processors and other CMOS-based components can significantly reduce power consumption. However, this often raises system failure rates and consequently increases application execution time. In this work, we present an energy saving undervolting approach that leverages the mainstream resilience techniques to tolerate the increased failures caused by undervolting.

failures↗

A Novel Charge Sensitive Pre-Amplifier Structure for Biological Temperature Readout Applications

Charge sensitive pre-amplifier architectures have been studied for decades for various applications, most of which adopt a gain stage with capacitor feedback to integrate the signal and a reset path to provide DC bias. For the biological temperature readout front-end, we propose a novel structure without the gain stage, which implies a strong potential for fully integrated high frequency applications. The proposed design provides 0.94 mV/pC with high conversion linearity while consuming only 1.4 micro watts of power with buffer under a 1.2-V supply rail. The prototype occupying a chip area of 0.038 square millimeter is fabricated in 130 nm standard CMOS process.

Wang, Hangfeng↗

An 85 nW IoT Node-Controlling SoC for MELs Power-Mode Management and Phantom Energy Reduction

This paper presents an ultra-low power (ULP) node-controlling system-on-chip (SoC) used for power-mode management and phantom energy reduction of miscellaneous electric loads (MELs). The SoC is powered from a single 2.5 V voltage supply enabled by the integrated power management unit (PMU) and can control up to 16 MELs due to the on-chip 16-channel correlator and the 32b RISC-V microprocessor. To further reduce the system power consumption, two clock domains have been adopted for the correlator and the processor separately. Fabricated in 65-nm CMOS, the measured minimum power consumption of the proposed SoC is only 85 nW at 0.45 V voltage supply and 1 kHz clock frequency. The measured maximum operating frequency can go up to 148 kHz with a 0.55 V supply. An application experiment successfully demonstrates that the SoC controls the power modes of MELs from wake-up to cut-off to save the average power and phantom energy.

power-mode management, phantom energy, phantom pow↗

A 194nW Energy-Performance-Aware loT SoC Employing a 5.2nW 92.6% Peak Efficiency Power Management Unit for System Performance Scaling, Fast DVFS and Energy Minimization

A self-powered IoT system-on-chip (SoC) reduces power to sub-μw and employs multiple power-management techniques to trade-off ultra-low power (ULP), higher performance, smaller energy harvester footprint, and longer operating lifetime. Minimum Energy Point Tracking (MEPT) [1]–[4] keeps an SoC operating at the minimum energy point (MEP) to enhance system lifetime. Previous sample-and-hold MEPT schemes need frequent voltage comparisons and a high-frequency clock that increases power [2]. Current-ratio-based MEPT relies on specialized CMOS technology for body-bias tuning [3]. A switched-capacitor-based MEPT can achieve energy minimization at a targeted performance [4], but it uses a 30MHz clock witμW power consumption and low power efficiency. For ULP IoT applications, SoCs need to have ultra-low quiescent power, high efficiency for energy delivery, performance scaling based on available energy, and energy minimization to increase system lifetime. In this work, we propose an ULP IoT SoC with a triple-mode power management unit (PMU) that integrates energy-performance scaling, event-driven fast DVFS, and MEPT features to improve the system energy efficiency, as shown in Fig. 13.8.1. This work achieves a minimum 194nW power consumption for the SoC and 5.2nW quiescent power for the PMU with a 92.6% peak efficiency and >10 4 dynamic range. The timing waveform in Fig. 13.8.1 (bottom), demonstrates the transition of the three modes including energy aware (EA), performance aware (PA), and MEPT based on event priority and input voltage level which reflects the energy availability. As such, the system energy consumption and performance could be well-balanced based on both the input and output conditions.

self-powered IoT system-on-chip (SoC)↗

Energy-Efficient and Secure S-Box circuit using Symmetric Pass Gate Adiabatic Logic

Differential Power Analysis (DPA) attack is considered to be a main threat while designing cryptographic processors. In cryptographic algorithms like DES and AES, S-Box is used to indeterminate the relationship between the keys and the cipher texts. However, S-box is prone to DPA attack due to its high power consumption. In this paper, we are implementing an energy-efficient 8-bit S-Box circuit using our proposed Symmetric Pass Gate Adiabatic Logic (SPGAL). SPGAL is energy-efficient as compared to the existing DPAresistant adiabatic and non-adiabatic logic families. SPGAL is energy-efficient due to reduction of non-adiabatic loss during the evaluate phase of the outputs. Further, the S-Box circuit implemented using SPGAL is resistant to DPA attacks. The results are verified through SPICE simulations in 180nm technology. SPICE simulations show that the SPGAL based S-Box circuit saves upto 92% and 67% of energy as compared to the conventional CMOS and Secured Quasi-Adiabatic Logic (SQAL) based S-Box circuit. From the simulation results, it is evident that the SPGAL based circuits are energy-efficient as compared to the existing DPAresistant adiabatic and non-adiabatic logic families. In nutshell, SPGAL based gates can be used to build secure hardware for lowpower portable electronic devices and Internet-of-Things (IoT) based electronic devices.

Kumar, Dinesh↗

A 4×4 electrooptic silicon photonic switch fabric with net neutral insertion loss

We present a strictly nonblocking 4×4 electrooptic silicon photonic switch fabric with on-chip gain. The switch integrates 12 Mach-Zehnder cells in a 3-stage topology equipped with fast electrooptic phase shifters, and thermooptic phase trimmers. To compensate for the losses of the fabric, a 4-channel GaInAsP/InP semiconductor optical amplifier array is flip-chip attached into an etched cavity in the silicon photonic chip with butt-coupled waveguide interfaces. The chip is wirebonded to a CMOS driver that provides push-pull drive to each elementary Mach-Zehnder cell. We demonstrate an optical switch assembly with net neutral insertion loss in the C-band together with nanosecond-scale reconfiguration time.

42 ENGINEERING↗

Bristol Ridge: A 28-nm $\times$ 86 Performance-Enhanced Microprocessor Through System Power Management

Power management techniques can be effective at extracting more performance and energy efficiency out of mature systems on chip (SoCs). For instance, the peak performance of microprocessors is often limited by worst case technology (Vmax), infrastructure (thermal/electrical), and microprocessor usage assumptions. Performance/watt of microprocessors also typically suffers from guard bands associated with the test and binning processes as well as worst case aging/lifetime degradation. Similarly, on multicore processors, shared voltage rails tend to limit the peak performance achievable in low thread count workloads. In this paper, we describe five power management techniques that maximize the per-part performance under the before-mentioned constraints. Using these techniques, we demonstrate a net performance increase of up to 15% depending on the application and TDP of the SoC, implemented on 'Bristol Ridge,' a 28-nm CMOS, dual-core x 86 accelerated processing unit.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

A 0.2–2 GHz Time-Interleaved Multistage Switched-Capacitor Delay Element Achieving 2.55–448.6 ns Programmable Delay Range and 330 ns/mm 2 Area Efficiency

Simulation of radar returns, full-duplex systems, and signal repeaters require hundreds of ns of programmable broadband radio frequency (RF) delay in the signal path to simulate large distances in the case of radar returns, for signal cancellation in full-duplex, and for isolation from reflections in signal repeaters. However, programmable broadband RF delay has been limited to ones of ns due to challenges in miniaturization with low loss and low power consumption. In this work, we present a 0.2–2 GHz digitally programmable RF delay element based on a time-interleaved multistage switched-capacitor (TIMS-SC) approach. The proposed approach enables hundreds of ns of broadband RF delay by employing sample time expansion in multiple stages of switched-capacitor storage elements. Further, the delay element was implemented in a 45 nm SOI CMOS process and achieves a 2.55–448.6 ns programmable delay range with < 0.12% delay variation across 1.8 GHz of bandwidth at maximum delay, 2.42 ns programmable delay steps, and 330 ns/mm 2 area efficiency. Through the proposed approach, the device shows minimal delay change across a -40 °C to 85 °C temperature range and < 0.25 dB gain variation across delay settings. The device achieves 26 dB gain, 7.4 dB noise figure, and consumes 74 mW from a 1 V supply with an active area of 1.36 mm 2 .

42 ENGINEERING↗

Controllable Reset Behavior in Domain Wall–Magnetic Tunnel Junction Artificial Neurons for Task-Adaptable Computation

Neuromorphic computing with spintronic devices has been of interest due to the limitations of CMOS-driven von Neumann computing. Domain wall–magnetic tunnel junction (DW-MTJ) devices have been shown to be able to intrinsically capture biological neuron behavior. Edgy-relaxed behavior, where a frequently firing neuron experiences a lower action potential threshold, may provide additional artificial neuronal functionality when executing repeated tasks. In this letter, we demonstrate that this behavior can be implemented in DW-MTJ artificial neurons via three alternative mechanisms: shape anisotropy, magnetic field, and current-driven soft reset. Using micromagnetics and analytical device modeling to classify the Optdigits handwritten digit dataset, we show that edgy-relaxed behavior improves both classification accuracy and classification rate for ordered datasets while sacrificing little to no accuracy for a randomized dataset. This letter establishes methods by which artificial spintronic neurons can be flexibly adapted to datasets.

42 ENGINEERING↗

Will Stochastic Devices Play Nice With Others in Neuromorphic Hardware?: There’s More to a Probabilistic System Than Noisy Devices

Achieving brain-like efficiency in computing requires a co-design between the development of neural algorithms, brain-inspired circuit design, and careful consideration of how to use emerging devices. The recognition that leveraging device-level noise as a source of controlled stochasticity represents an exciting prospect of achieving brain-like capabilities in probabilistic neural algorithms, but the reality of integrating stochastic devices with deterministic devices in an already-challenging neuromorphic circuit design process is formidable. Here, we explore how the brain combines different signaling modalities into its neural circuits as well as consider the implications of more tightly integrated stochastic, analog, and digital circuits. Further, by acknowledging that a fully CMOS implementation is the appropriate baseline, we conclude that if mixing modalities is going to be successful for neuromorphic computing, it will be critical that device choices consider strengths and limitations at the overall circuit level.

42 ENGINEERING↗

Temperature-Dependent Transport Characteristics of 2D MoS2 Channel FETs Grown Using Salt-Based Precursors

D Transition Metal Dichalcogenides (TMDs), particularly MoS2, are promising candidates for sub-10 nm Gate All Around (GAA) CMOS FETs. Salt-assisted Chemical Vapor Deposition (CVD) enable lateral MoS2 growth at atmospheric pressure and low temperatures. This work analyzes salt-based precursor-driven CVD-grown MoS2 FETs at various temperatures. MoS2 was grown using Ammonium Molybdate salt, sulfurized at 750∘C, and transferred onto p−Si3/SiO2 substrates. At room temperature, threshold voltage (VT) ranged from -35 V to -25 V, with a peak drain current of 1.2μA/μm. As temperature increased above 325K, VT shifted exponentially, and carrier mobility dropped significantly. At 400 K, the gate lost channel control, though gate leakage current remained low. These results are compared with non-salt-based MoS2 growth to assess salt precursor effects.Notice: This manuscript has been authored by UT-Battelle, LLC, under contract DE-AC05-00OR22725 with the US Department of Energy (DOE). The US government retains and the publisher, by accepting the article for publication, acknowledges that the US government retains a nonexclusive, paid-up, irrevocable, worldwide license to publish or reproduce the published form of this manuscript, or allow others to do so, for US government purposes. DOE will provide public access to these results of federally sponsored research in accordance with the DOE Public Access Plan (https://www.energy.gov/doe-public-access-plan).

Jones, Andrew [ORNL] (ORCID:0009000233849687)↗

Development of low mass optical readout for high data bandwidth systems.

At Argonne National Laboratory the High Energy Physics and Center for Nanoscale Materials Divisions are working on a project to develop a new generation of detector readout using high speed data transfer optical devices that can be implemented in particle physics or for long distances. Free-space communications devices offer the potential for reductions in mass, power, and cost of data paths for on-board trigger and readout of tracking detectors. The project involves three areas of study: light modulation, the design and construction of MEMS optical devices, and the control systems for maintaining precise laser light positioning. We demonstrate an optical link in air over one meter and with low error rate at 1 Gb/s. We demonstrate steering of an optical beam over a meter with a precision of 5 micrometers utilizing a MEMS mirror and reflected light in the feedback loop. For early testing, light modulation tests with a fiber link using Li-Niobate modulators and a data generation and error checking chip are done at 1 Gb/s. Many companies and universities are developing modulators which will be incorporated into CMOS chips. We are doing radiation hardness studies for one of the materials involved. Laser light will need to be steered on to and kept centered on the detector in the presence of thermal or mechanical motion, etc. This steering will be controlled by MEMS mirrors. Polycrystalline and crystalline silicon based mirror designs are being studied. We review the current status of the project and outline plans for the future development of the system.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Design and Measurement of a Low-Noise 64-Channels Front-End Readout ASIC for CdZnTe Detectors

Cadmium zinc telluride (CdZnTe) detectors, as one of the principal detectors for the next-generation X-ray and γ-ray imagers, have high energy resolution and supporting electrode patterning in the radiation environment at room-temperature. In the present, a number of internationally renowned research institutions and universities are actively using these detector systems to carry out researches of energy spectrum analysis, medical imaging, materials characterization, high-energy physics, nuclear plant monitoring, and astrophysics. As the most important part of the readout system for the CdZnTe detector, the front-end readout application specific integrated circuit (ASIC) would have an important impact on the performances of the whole detector system. In order to ensure the small signal to noise ratio (SNR) and sufficient range of the output signal, it is necessary to design a front-end readout ASIC with very low noise and very high dynamic range. In addition, radiation hardness should be considered when the detectors are utilized in the space applications and high energy physics experiments. In this paper, we present measurements and performances of a novel multi-channel radiation-hardness low-noise front-end readout ASIC for CdZnTe detectors. The readout circuits in each channel consist of charge sensitive amplifier, leakage current compensation circuit (LCC), CR-RC shaper, S-K filter, inverse proportional amplifier, peak detect and hold circuit (PDH), discriminator and trigger logic, time sequence control circuit and driving buffer. All of 64 readout channels' outputs enter corresponding inputs of a 64 channel multiplexer. The output of the mux goes directly out of the chip via the output buffer. The 64-channel readout ASIC is implemented using the TSMC 0.35 μm mixed-signal CMOS technology. The die size of the prototype chip is 2.7 mm x 8 mm. At room temperature, the equivalent noise level of a typical channel reaches 66 e{sup -} (rms) at zero farad for a power consumption of 8 mW per channel. The linearity error is lower than 1% and the overall gain of the readout channel is 165 V/pC. The crosstalk between the channels is less than 3%. By connecting the readout ASIC to a CdZnTe detector, we obtained a γ-ray spectrum, the energy resolution is 5.1% at the 59.5-keV line of {sup 241}Am source. (authors)

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗