Bøger af Marian Verhelst

This book focuses on the acceleration of emerging irregular sparse workloads, posed by novel artificial intelligent (AI) models and sparse linear algebra. Specifically, the book outlines several co-optimized hardware-software solutions for a highly promising class of emerging sparse AI models called Probabilistic Circuit (PC) and a similar sparse matrix workload for triangular linear systems (SpTRSV). The authors describe optimizations for the entire stack, targeting applications, compilation, hardware architecture and silicon implementation, resulting in orders of magnitude higher performance and energy-efficiency compared to the existing state-of-the-art solutions. Thus, this book provides important building blocks for the upcoming generation of edge AI platforms.

This book proposes probabilistic machine learning models that represent the hardware properties of the device hosting them.

This book discusses the design and implementation aspects of ultra-low power biosignal acquisition platforms that exploit analog-assisted and algorithmic approaches for power savings.The authors describe an approach referred to as "analog-and-algorithm-assisted" signal processing.This enables significant power consumption reductions by implementing low power biosignal acquisition systems, leveraging analog preprocessing and algorithmic approaches to reduce the data rate very early in the signal processing chain.They demonstrate savings for wearable sensor networks (WSN) and body area networks (BAN), in the sensors' stimulation power consumption, as well in the power consumption of the digital signal processing and the radio link. Two specific implementations, an adaptive sampling electrocardiogram (ECG) acquisition and a compressive sampling (CS) photoplethysmogram (PPG) acquisition system, are demonstrated.First book to present the so called, "analog-and-algorithm-assisted" approaches for ultra-low power biosignal acquisition and processing platforms;Covers the recent trend of "beyond Nyquist rate" signal acquisition and processing in detail, including adaptive sampling and compressive sampling paradigms;Includes chapters on compressed domain feature extraction, as well as acquisition of photoplethysmogram, an emerging optical sensing modality, including compressive sampling based PPG readout with embedded feature extraction;Discusses emerging trends in sensor fusion for improving the signal integrity, as well as lowering the power consumption of biosignal acquisition systems.

Smart energy management is indispensable in modern radios. This book describes and applies an energy-driven design strategy to the design of an energy-efficient, highly scalable, pulsed UWB receiver, suitable for low data rate communication and sub-cm ranging.