Design methodology

An Extendable Wideband Band-Divided SPnT Switch Topology Based on a Reconfigurable Pole-to-Throw Network

An Extendable Wideband Band-Divided SPnT Switch Topology Based on a Reconfigurable Pole-to-Throw Network 150 150

Abstract:

This article presents an extendable wideband band-divided SP $n$ T switch topology based on a reconfigurable pole-to-throw network (RPTN). Compared to conventions, it mitigates the loss degradation resulting from complex signal routing in advanced RF front-ends. This is enabled by the RPTN, which divides the throws into low-/mid-/high-band (…

View on IEEE Xplore

FSL-HDnn: A 40-nm Few-Shot On-Device Learning Accelerator With Integrated Feature Extraction and Hyperdimensional Computing

FSL-HDnn: A 40-nm Few-Shot On-Device Learning Accelerator With Integrated Feature Extraction and Hyperdimensional Computing 150 150

Abstract:

This article introduces FSL-HDnn, an energy-efficient accelerator that implements the end-to-end pipeline of feature extraction (FE) and on-device few-shot learning (FSL). The accelerator addresses fundamental challenges of on-device learning (ODL) for resource-constrained edge applications through two synergistic modules: a parameter-efficient feature extractor (FE) employing weight clustering and an FSL classifier …

View on IEEE Xplore

A Three-Level Dual-Path Hybrid Buck Converter With Mitigated Capacitive Current and Seamless DVS Across a Wide Output Voltage Range

A Three-Level Dual-Path Hybrid Buck Converter With Mitigated Capacitive Current and Seamless DVS Across a Wide Output Voltage Range 150 150

Abstract:

This article presents a three-level dual-path (TLDP) buck converter tailored for mobile application processors (APs) with a 1.8-V input and 0.4–1.5-V output range. Conventional buck (CB) and multi-level buck (MLB) converters suffer from prohibitive DC resistance ( $R_{\mathrm {DCR}}$ ) losses in compact inductors as load current increases. Although various multi-path …

View on IEEE Xplore

Building Minimum-Power Digital Circuits With Sub-Femtowatt Dynamic Leakage-Suppression Logic Standard Cells

Building Minimum-Power Digital Circuits With Sub-Femtowatt Dynamic Leakage-Suppression Logic Standard Cells 150 150

Abstract:

This article investigates dynamic leakage-suppression (DLS) logic for minimum-power operation in digital circuits. The DLS logic family uses a strong intrinsic subthreshold leakage reduction technique and is an emerging approach for computing applications that prioritize low active power at or beneath the nW-level. We analytically model the key operating characteristics …

View on IEEE Xplore

A Low-Spur Fractional-N DPLL With Analog Pre-Distortion DTC Implementing Second-/Third-Order Calibration

A Low-Spur Fractional-N DPLL With Analog Pre-Distortion DTC Implementing Second-/Third-Order Calibration 150 150

Abstract:

This article presents a low-spur, low-jitter fractional-N digital phase-locked loop (DPLL) employing an analog pre-distortion digital-to-time converter (APD-DTC) to suppress fractional spurs by compensating for both second- and third-order nonlinearities without using complex digital pre-distortion. To support background calibration, a spur-level detection scheme using phase detector gain (PDG) is proposed …

View on IEEE Xplore

A 27-Gb/s/pin Inverter-Based Delay-Matched PAM-3 Transceiver With Pattern-Selective Switching Jitter Compensation for Die-to-Die Interfaces

A 27-Gb/s/pin Inverter-Based Delay-Matched PAM-3 Transceiver With Pattern-Selective Switching Jitter Compensation for Die-to-Die Interfaces 150 150

Abstract:

This article presents a 27-Gb/s/pin inverter-based delay-matched three-level pulse amplitude modulation (PAM-3) transceiver for short-reach die-to-die (D2D) interfaces. The proposed design uses two parallel, identical offset-programmable transimpedance amplifiers (TIAs) to independently generate two finely tunable PAM-3 reference levels with minimal parasitic loading, enabling multi-level signaling within a …

View on IEEE Xplore

A −248-dB FOMref CS-Fusion PLL Architecture Combining the Advantages of PFD-CP PLLs and Sampling PLLs in a Single Loop

A −248-dB FOMref CS-Fusion PLL Architecture Combining the Advantages of PFD-CP PLLs and Sampling PLLs in a Single Loop 150 150

Abstract:

This article introduces a CP-sampling (CS)-fusion PLL architecture that combines the advantages of PFD-CP PLLs (robust and speedy locking) and sampling PLLs (low jitter/power) in a single loop. The PFD sends pulsewidth-modulation (PWM) pulses to a merged $mathbf {G_{m}}$ /CP (GCP) block so that the phase error …

View on IEEE Xplore

STAR-SRAM: 16-bit Floating-Point SRAM-Based Digital Computing-in-Memory Macro in a 28 nm

STAR-SRAM: 16-bit Floating-Point SRAM-Based Digital Computing-in-Memory Macro in a 28 nm 150 150

Abstract:

A digital computing-in-memory (DCIM) macro emerges as a promising building block in a deep neural network (DNN) accelerator. To better support DNN workloads, circuit designers aim to improve three main metrics for macros: energy efficiency, compute density, and weight density. Improvements in those metrics directly translate into reduced energy consumption, …

View on IEEE Xplore

A 6.9-ppm/°C, 0.66-mV/V Line Regulation, 50-mA Multi-Loop Low-Dropout Regulator With Integrated Single-BJT Voltage and Current References

A 6.9-ppm/°C, 0.66-mV/V Line Regulation, 50-mA Multi-Loop Low-Dropout Regulator With Integrated Single-BJT Voltage and Current References 150 150

Abstract:

This article presents a fully integrated output-capacitor-less (OCL) multi-feedback-loop low-dropout regulator (LDO) with an in-built single bipolar junction transistor (BJT)-based voltage and current reference (VCR) for energy-harvesting Internet of Things (IoT) devices. The proposed architecture comprises four loops, which significantly enhance the DC regulation and transient performance of the …

View on IEEE Xplore