Bandwidth

A 24.5–45.2-GHz Low-Jitter Compact Differentially Injection-Locked Clock Multiplier With Folded-Inductor-Based Magnetic-Flux Cancellation

A 24.5–45.2-GHz Low-Jitter Compact Differentially Injection-Locked Clock Multiplier With Folded-Inductor-Based Magnetic-Flux Cancellation 150 150

Abstract:

In this article, we present a differentially injection-locked clock multiplier (ILCM) featuring an ultra-wide frequency tuning range (TR) and low jitter, achieved through a compact folded-inductor-based magnetic-flux cancellation technique. A co-designed series-LC dual-mode quadrature ring oscillator (QRO) and edge-combining frequency doubler operating in the mm-wave band jointly extend the TR …

View on IEEE Xplore

A 129–146-GHz Direct-Digital Modulation FinFET Transmitter With On-Chip Mismatch Calibrations for Beyond-5G Wireless Communications

A 129–146-GHz Direct-Digital Modulation FinFET Transmitter With On-Chip Mismatch Calibrations for Beyond-5G Wireless Communications 150 150

Abstract:

This article presents a D-band direct-digital modulation (DDM) transmitter with on-chip digital calibration blocks for future beyond-5G (B5G) wireless communication. The proposed DDM architecture mitigates the need for complex intermediate frequency (IF) generation and power-hungry digital-to-analog converters (DACs). The transmitter is implemented primarily in TSMC’s 16-nm p-FinFETs, …

View on IEEE Xplore

A 32-Channel 85.4 dB SNDR Time-Multiplexed Neural Recording Front-End Achieving Within-Conversion Artifact Recovery

A 32-Channel 85.4 dB SNDR Time-Multiplexed Neural Recording Front-End Achieving Within-Conversion Artifact Recovery 150 150

Abstract:

This article presents a 32-channel time-multiplexed highly digital neural recording front-end (RFE) that exhibits sub- $8.8~\mu $ s recovery latency from large stimulation artifacts while delivering high-resolution data at the Nyquist rate. The RFE is time-shared across 32 channels for low-frequency electrocorticography (ECoG) recording and across four channels for high-frequency action potentials (…

View on IEEE Xplore

A 15-MHz, 2.7-mm2 Notch-Free Hybrid Magnetic Current Sensor With Feedforward Ripple Cancellation and ±0.8% Local Gain Non-Uniformity

A 15-MHz, 2.7-mm2 Notch-Free Hybrid Magnetic Current Sensor With Feedforward Ripple Cancellation and ±0.8% Local Gain Non-Uniformity 150 150

Abstract:

This article proposes a hybrid magnetic current sensor achieving a 15-MHz bandwidth within a compact 2.7-mm2 area. To mitigate the pole–zero mismatch inherent in the two-stage integrator topology, a dual-output Gm-C integrator with subtractor-based compensation is proposed, achieving a ±0.8% local gain non-uniformity. A wideband feedforward ripple suppression scheme cancels …

View on IEEE Xplore

A 0.5–2.5-GS/s Resettable Ring-VCO-Based ADC Eliminating Quantization-Noise Shaping

A 0.5–2.5-GS/s Resettable Ring-VCO-Based ADC Eliminating Quantization-Noise Shaping 150 150

Abstract:

This article presents a Nyquist-rate Analog-to-digital converter (ADC) operating from 0.5 to 2.5 GS/s based on an open-loop resettable ring VCO (R-RVCO). By inherently embedding the $1 {\,}-{\,}z^{-1}$ transfer function, the R-RVCO eliminates the need for an explicit differentiator, suppresses VCO phase-noise (PN) integration, and avoids quantization-noise (QN) shaping within …

View on IEEE Xplore

Adelia: A 4-nm LLM Processing Unit With Streamlined Dataflow and Dual-Mode Parallelism for Maximizing Hardware Efficiency

Adelia: A 4-nm LLM Processing Unit With Streamlined Dataflow and Dual-Mode Parallelism for Maximizing Hardware Efficiency 150 150

Abstract:

The proliferation of large language models (LLMs) as cross-domain foundation models is fueled by aggressive scaling in both parameter counts and inference-time computation. The emergence of sophisticated reasoning models further accelerates this trend, demanding longer context windows and escalating the computational and memory burdens of inference. A fundamental challenge arises …

View on IEEE Xplore

A 2–18 GHz High-Efficiency CMOS Nonuniform Distributed Power Amplifier With a Novel Reconfigurable Inductive Termination

A 2–18 GHz High-Efficiency CMOS Nonuniform Distributed Power Amplifier With a Novel Reconfigurable Inductive Termination 150 150

Abstract:

This article presents a 2–18 GHz high-efficiency CMOS nonuniform distributed power amplifier (NDPA) with a novel reconfigurable inductive termination technique for ultra-broadband efficiency enhancement. First, the inherent drawback of the degrading efficiency with growing frequency in a conventional non-reconfigurable NDPA architecture with multi-octave bandwidth is studied. A simple and effective reconfigurable …

View on IEEE Xplore

A 56-Gb/s Hybrid Silicon Photonic and 5-nm CMOS 3-D-Integrated Transceiver for Optical Compute I/O

A 56-Gb/s Hybrid Silicon Photonic and 5-nm CMOS 3-D-Integrated Transceiver for Optical Compute I/O 150 150

Abstract:

This work presents a hybrid 3-D-integrated silicon photonic (SiPh) transceiver suitable for realizing chiplet-based optical I/O in future AI/ML ASIC packages. The optical transceiver die stack is composed of two ICs: a SiPh IC (PIC) with micrometer-scale, thermally robust electro-absorption modulators (EAMs), and a 5-nm CMOS electronic IC (…

View on IEEE Xplore

A 0.29 pJ/b 5.27 Tb/s/mm UCIe Advanced Package Link With 2.5-D CoWoS Interposer and a 0.52 pJ/b 0.448 Tb/s/mm UCIe Standard Package Link With Organic Substrate in 3 nm FinFET

A 0.29 pJ/b 5.27 Tb/s/mm UCIe Advanced Package Link With 2.5-D CoWoS Interposer and a 0.52 pJ/b 0.448 Tb/s/mm UCIe Standard Package Link With Organic Substrate in 3 nm FinFET 150 150

Abstract:

This work presents two die-to-die (D2D) wireline transceivers, one compliant with the UCIe advanced package (UCIe-AP) and the other with the UCIe standard package (UCIe-SP) standard, developed in 3 nm FinFET. The Universal Chiplet interconnect express (UCIe)-AP link has 64 RX and 64 TX data lanes in one PHY module and …

View on IEEE Xplore