Overview

The overarching research goal of eLab is to design, model, and optimize energy-efficient computing systems that connect emerging applications with advanced hardware platforms. Our research spans computer architecture, embedded systems, design automation, machine learning systems, wearable sensing, and human-centered computing. Across these areas, we develop cross-layer methodologies that jointly consider applications, algorithms, architectures, interconnects, power delivery, thermal constraints, and runtime management.

A major focus of our current research is heterogeneous and chiplet-based computing. As computing systems move beyond monolithic chips toward 2.5D/3D integration, chiplets, domain-specific accelerators, and network-on-interposer architectures, new challenges arise in communication, power delivery, thermal behavior, workload mapping, and design-space exploration. eLab develops modeling, simulation, and optimization frameworks to guide the design of scalable, efficient, and reliable heterogeneous architectures.

Our work also advances energy-efficient AI and embedded computing systems, including domain-specific SoCs, AI accelerators, processing-in-memory architectures, runtime scheduling, power and thermal modeling, and communication-centric architecture design. We are particularly interested in efficient acceleration of large language models, hybrid Transformer–state-space models, mixture-of-experts models, and edge AI workloads.

In parallel, eLab continues its long-standing work in wearable, flexible, and human-centered systems. The OpenHealth project provides an open-source hardware/software platform for wearable health monitoring and embedded sensing research. Our Flexible Hybrid Electronics project develops methodologies and tools for Systems-on-Polymer: fully integrated flexible electronic systems capable of sensing, computing, and communication. Our Human-Machine Communication project studies wearable and embedded interfaces that help users convey their intent to smart objects, assistive devices, and internet-of-things systems.

Together, these directions support eLab’s broader mission: advancing energy-efficient, intelligent, and human-centered computing systems through cross-layer design, open-source tools, reproducible research artifacts, and education.

Research Funding: We thank the following agencies and companies for their generous support of our research.

We also thank Intel, Altera University Programs and Texas Instruments for their equipment donations.

The application domains we target, and cross-cutting theory and design technologies employed at eLab are summarized below.

Smart phone ICs Server and data centers for cloud computing  Personal e-health ICs
Low-power, heterogeneous
embedded systems
Many integrated core architectures for HPC Flexible electronics
Energy Efficient
VLSI Design
High Performance
VLSI Design
SoC/Circuit Solutions for Innovative Systems
Digital System Design Network Theory VLSI Design Control Theory Algorithms

Please contact uogras@wisc.edu for latest information.