Search

In this work, we present a novel 8T static random access memory (SRAM)-based compute-in-memory (CIM) macro for processing neural networks with high energy efficiency. The proposed 8T bitcell is free from disturb issues thanks to the decoupled read channels by adding two extra transistors to the standard 6T bitcell. A 128 $\times $ 128 8T SRAM array offers massively parallel binary multiply and accumulate (MAC) operations with 64 $\times $ binary inputs (0/1) and 64 $\times $ 128 binary weights (+1/–1). After parallel MAC operations, 128 column-based neurons generate 128 $\times $ 1–5 bit outputs in parallel. The proposed column-based neuron comprises 64 $\times $ bitcells for dot-product, 32 $\times $ bitcells for analog-to-digital converter (ADC), and 32 $\times $ bitcells for offset calibration. The column ADC with 32 $\times $ replica SRAM bitcells converts the analog MAC results (i.e., a differential read bitline (RBL/RBLb) voltage) to the 1–5 bit output code by sweeping their reference levels in 1–31 cycles (i.e., $2^{N}$ –1 cycles for $N$ -bit ADC). The measured linearity results [differential nonlinearity (DNL) and integral nonlinearity (INL)] are +0.314/–0.256 least significant bit (LSB) and + 0.27/–0.116 LSB, respectively, after offset calibration. The simulated image classification results are 96.37% for Mixed National Institute of Standards and Technology database (MNIST) using a multi-layer perceptron (MLP) with two hidden layers, 87.1%/82.66% for CIFAR-10 using VGG-like/ResNet-18 convolutional neural networks (CNNs), demonstrating slight accuracy degradations (0.67%–1.34%) compared with the software baseline. A test chip with a 16K 8T SRAM bitcell array is fabricated using a 65-nm process. The measured energy efficiency is 490–15.8 TOPS/W for 1–5 bit ADC resolution using 0.45-/0.8-V core supply. [ABSTRACT FROM AUTHOR]

This article (Colonnade) presents a fully digital bit-serial compute-in-memory (CIM) macro. The digital CIM macro is designed for processing neural networks with reconfigurable 1–16 bit input and weight precisions based on bit-serial computing architecture and a novel all-digital bitcell structure. A column of bitcells forms a column MAC and used for computing a multiply-and-accumulate (MAC) operation. The column MACs placed in a row work as a single neuron and computes a dot-product, which is an essential building block of neural network accelerators. Several key features differentiate the proposed Colonnade architecture from the existing analog and digital implementations. First, its full-digital circuit implementation is free from process variation, noise susceptibility, and data-conversion overhead that are prevalent in prior analog CIM macros. A bitwise MAC operation in a bitcell is performed in the digital domain using a custom-designed XNOR gate and a full-adder. Second, the proposed CIM macro is fully reconfigurable in both weight and input precision from 1 to 16 bit. So far, most of the analog macros were used for processing quantized neural networks with very low input/weight precisions, mainly due to a memory density issue. Recent digital accelerators have implemented reconfigurable precisions, but they are inferior in energy efficiency due to significant off-chip memory access. We present a regular digital bitcell array that is readily reconfigured to a 1–16 bit weight-stationary bit-serial CIM macro. The macro computes parallel dot-product operations between the weights stored in memory and inputs that are serialized from LSB to MSB. Finally, the bit-serial computing scheme significantly reduces the area overhead while sacrificing latency due to bit-by-bit operation cycles. Based on the benefits of digital CIM, reconfigurability, and bit-serial computing architecture, the Colonnade can achieve both high performance and energy efficiency (i.e., both benefits of prior analog and digital accelerators) for processing neural networks. A test-chip with $128 \times 128$ SRAM-based bitcells for digital bit-serial computing is implemented using 65-nm technology and tested with 1–16 bit weight/input precisions. The measured energy efficiency is 117.3 TOPS/W at 1 bit and 2.06 TOPS/W at 16 bit. [ABSTRACT FROM AUTHOR]

Gesture recognition has increasingly become one of the most popular human–machine interaction techniques for smart devices. Existing gesture recognition systems suffer from either excessive power consumption or large size, limiting their applications for ultralow-power wearable devices. This article presents an accurate area-efficient and low-power real-time gesture recognition system for smart wearable devices. The proposed work utilizes an accurate peak-based gesture classification engine with less memory and a low-resolution and low-power on-chip image sensor for achieving high area efficiency and low power. In addition, the feature extraction architecture removes fixed-pattern noises from the low-power on-chip image sensor for accuracy improvement and employs parallelism for recognition speed enhancement. Thus, the proposed system accomplishes accurate real-time gesture recognition for eight motion hand gestures with an average recognition accuracy of 90.6% and latency of 4.228 ms. Measurement results of a test chip fabricated in 65-nm CMOS demonstrate that the proposed system consumes $137.0~\mu \text{W}$ at 30 frames/s while occupying only 1.78 mm2, which achieves the lowest power and smallest area among the recently reported gesture recognition systems. [ABSTRACT FROM AUTHOR]

In Space applications, the scaling of transistors has made integrated circuits (ICs) more susceptible to soft errors, caused by radiation strikes. When a soft error causes a bit flip in a memory device, this event is referred to as a Single Event Upset (SEU). Since SEU errors degrade system performance and eventually lead to system failure, the design of radiation-resilient memory is substantial. This paper presents a radiation resilient SRAM with a self-refresh scheme for lowering the number of errors in each row below a threshold number. The proposed self-refresh operation reads out the stored data and performs single error correction using a simple algorithm during its hold/idle mode. A 4KB SRAM test chip in 65nm CMOS technology demonstrates a significant reduction in errors with the self-refresh operation. When the SRAM test chip was exposed to accelerated proton radiation with an energy level of 39.38 MeV, the self-refresh scheme reduces the number of uncorrectable errors by $25\times $ and $8\times $ lesser for the fluence of $9.82\times 10^{11}$ particles/cm2 and $49.1\times 10^{11}$ particles/cm2, respectively. [ABSTRACT FROM AUTHOR]

This article presents an 8T static random access memory (SRAM) macro with vertical read wordline (RWL) and selective dual split power (SDSP) lines techniques. The proposed vertical RWL reduces dynamic energy consumption during read operation by charging and discharging only selected read bitlines (RBLs). The data-aware SDSP technique combined with vertical write bitlines enhances both the write margin (WM) and the static noise margin (SNM). A 16-kb SRAM test chip fabricated in 65-nm CMOS technology demonstrates the minimum energy consumption of 0.506 pJ at 0.4 V and the minimum operating voltage of 0.26 V. [ABSTRACT FROM AUTHOR]

A single-inductor multiple-output (SIMO) dc–dc converter with output-voltage-aware charge control (OVACC) is presented in this paper. The OVACC scheme reduces the cross regulation by computing the total energy required by all outputs to extract exactly the same amount of energy from the input. Moreover, the proposed design also extends the load capability range of the SIMO converter by supporting the operation in both continuous conduction mode and discontinuous conduction mode while boosting the overall power efficiency. The proposed SIMO (five outputs) dc–dc converter has been implemented in a 0.18 μm 1P6M CMOS process. Measurement results show that the best cross regulation is 0.016 mV/mA when the output current of 250 mA abruptly changes. And no cross regulation is observed when the first output current changes 150 mA, the second output current changes 160 mA, the third and fourth output currents change 180 mA, and the last output current changes 250 mA. Also, the measured peak power efficiency is 86%. [ABSTRACT FROM AUTHOR]

This paper presents a novel 3-D maximum power point tracking (3-D MPPT) system for energy harvesting systems (EHS) within Internet of Things (IoT) smart nodes. The proposed 3-D MPPT utilizes a switch width modulation (SWM) technique for improving power efficiency (PE) at idle (<1 $\mu \text{A}$) and heavy (>300 $\mu \text{A}$) load modes. The SWM eliminates the gate driver/conduction loss tradeoff in a reconfigurable switched capacitor charge pump (SCCP). The proposed SWM technique modulates the SCCP switch resistance in proportion to the load condition, input voltage, and $V_{\mathrm {gs}}$ applied. A cold startup technique is incorporated in the proposed EHS with an energy-aware algorithm for input harvester select purpose. The fabricated test chip in 65-nm CMOS technology can harvest solar and thermal energy from 0.35 V and provides a regulated output voltage at 1 V with the peak efficiency of 88% at 200 $\mu \text{W}$ and PE >60% at 100 nW. [ABSTRACT FROM AUTHOR]

This paper presents a direct digital frequency synthesizer (DDFS) based on the nonlinear DAC with a maximum operating frequency of 2 GHz. This work proposes three design methods to enhance the performance of a DDFS. First, a multi-level momentarily activated bias is proposed to reduce power dissipation in the phase accumulator. Second, a coarse phase-based consecutive fine amplitude grouping scheme is presented to reduce hardware complexity and power consumption in the digital decoder. Third, the mixed-wave conversion topology in the nonlinear DAC is proposed to improve the output spectral purity. The DDFS with 9 bit amplitude resolution is capable of producing a minimum spurious-free dynamic range (SFDR) of 55.1 dBc up to Nyquist frequency at the clock frequency of 2 GHz. The prototype DDFS is fabricated in a 55-nm CMOS. It occupies an active area of 0.1 mm 2 with a total power dissipation of 130 mW. The figure of merit of this DDFS is 8944 GHz · 2 (SFDR/6) /W. [ABSTRACT FROM AUTHOR]