PhD in Computer Engineering

Steven Harris

Building the future of heterogeneous computing

Doctoral researcher at Washington University in St. Louis, advised by Dr. Roger D. Chamberlain and Dr. Christopher D. Gill. I design and optimize high-performance computing systems that push the boundaries of hardware capabilities—from FPGAs to GPU clusters to emerging heterogeneous architectures.

Heterogeneous Computing FPGA Design Real-Time Systems High-Performance Computing OpenCL / HLS
01.

About Me

I'm a PhD holding Computer Engineering at Washington University in St. Louis, working at the intersection of heterogeneous computing, real-time systems, and embedded architecture.

My research focuses on understanding how we can better design computing systems that leverage the diversity of modern processors—CPUs, GPUs, FPGAs—to achieve unprecedented performance, energy efficiency, and predictability for mission-critical applications.

Before pursuing my PhD, I gained extensive experience in IT infrastructure and healthcare technology at Washington University School of Medicine, where I continue to serve as Director of Information Technology.

Advisors

10+
Publications
4+
Conference Talks
IEEE
HPEC 2020, 2022, 2026
IPDPSW 2025, 2026
SC25
Supercomputing
02.

Research Interests

⚡

Heterogeneous Architectures

Designing and optimizing computing systems that intelligently combine CPUs, GPUs, FPGAs, and custom accelerators. Exploring performance portability across diverse hardware platforms.

FPGA GPU HLS OpenCL HARPv2
🎯

Parallel Real-Time Systems

Developing scheduling algorithms and system architectures that ensure predictable timing guarantees while maximizing parallelism in modern multi-core and heterogeneous processors.

Scheduling Multi-core Timing Analysis Mixed-Criticality
🔧

Cyber-Physical Systems

Researching the seamless integration of computation, networking, and physical processes with emphasis on safety-critical embedded systems for defense and aerospace applications.

Embedded Systems IoT Safety-Critical Real-Time
📊

Performance Modeling

Creating analytical models to predict and optimize performance across non-uniform heterogeneous platforms using structural equation modeling and machine learning techniques.

SEM Performance Analysis Optimization Benchmarking
🔄

System Software & Virtualization

Building operating systems, hypervisors, and runtime environments that efficiently manage heterogeneous hardware resources with minimal overhead.

OS Virtualization Runtime Systems Resource Mgmt
🧮

High-Performance Computing

Accelerating computational kernels including matrix operations, linear algebra, and scientific computing on emerging heterogeneous supercomputing platforms.

HPC BLAS/LAPACK Parallel Computing Supercomputing
03.

Publications

Understanding Performance Modeling in a Heterogeneous World Ph.D. Dissertation, Washington University in St. Louis

Steven D. Harris

Doctoral dissertation establishing a relational systems-engineering framework for non-uniform heterogeneous and asymmetric multiprocessing (AMP) architectures. Core contributions include the Selection, Assessment, Classification (SAC) tripartite methodology and the application of Structural Equation Modeling (SEM) to bypass multi-stage Ordinary Least Squares (OLS) error propagation in out-of-order execution pipelines. Introduces the use of the Boltzmann Uniqueness Theorem to mathematically disprove the additive "sum of parts" power assumption, identifying physical "architectural entanglement" and quantifying shared-fabric parasitic power via the Ghost Power Coefficient (γ) across ARM DynamIQ and Intel hybrid platforms. Validates these principles through the Entropy-Bounded Performance Model (EBPM), enforcing hard real-time latency bounds for edge-based biomedical sensing applications.

Performance Models for Non-Uniform Heterogeneous Platforms and their Validation Using Structural Equation Modeling IJNC 2026

Steven D. Harris, Roger D. Chamberlain, Christopher D. Gill

Non-uniform heterogeneous computing architectures are increasingly common in modern computing platforms targeting mission-critical edge-cloud and embedded applications. We present a comprehensive performance modeling framework using structural equation modeling to capture the complex relationships between application characteristics, hardware parameters, and execution performance.

Understanding Performance-Impacting Factors ICCPS 2026

Steven D. Harris, Christopher D. Gill, Roger D. Chamberlain

Analyzes performance-impacting factors in cyber-physical systems.

Abstract DOI
Heterogeneous Non-Linear Performance Modeling of Parameter Interactions IPDPSW 2026

Steven D. Harris, Roger D. Chamberlain, Christopher D. Gill

Examines non-linear performance modeling and parameter interactions in parallel and distributed systems.

Paper DOI
Measure-Theoretic Diagnostics of Architectural Entanglement in Asymmetric Multiprocessing Systems: A Boltzmann Uniqueness Approach Modelling 2026

Steven D. Harris, Christopher D. Gill, Roger D. Chamberlain

Presents a measure-theoretic diagnostic framework utilizing a Boltzmann uniqueness approach to evaluate architectural entanglement in asymmetric multiprocessing systems.

Paper DOI
An Entropy-Bounded Thermodynamic Approach for Real-Time Maternal Hemorrhage Monitoring HPEC 2026

Steven D. Harris, Francesca Bonetta-Misteli, Lleyton Martin, Christopher D. Gill, Roger D. Chamberlain, Christine M. O’Brien

Proposes an entropy-bounded thermodynamic model for the real-time monitoring of maternal hemorrhage.

Paper DOI
Performance Modeling of Non-Uniform Heterogeneous Platforms IPDPSW 2025

Steven D. Harris, Roger D. Chamberlain, Christopher D. Gill

Details performance modeling techniques applied to non-uniform heterogeneous architectures.

Paper DOI
Structural Equation Modeling for Heterogeneous Platforms SC25

Steven D. Harris

Proceedings of the ACM/IEEE International Conference for High Performance Computing, Networking, Storage, and Analysis (Supercomputing 2025). Presented research on applying structural equation modeling to understand performance characteristics of heterogeneous computing platforms.

Challenges Designing for FPGAs using High-Level Synthesis IEEE HPEC 2022

Clayton J. Faber, Steven D. Harris, Zhili Xiac, Roger D. Chamberlain, Anthony M. Cabrera

An exploration of the challenges and trade-offs when designing hardware accelerators using High-Level Synthesis (HLS) tools for FPGAs. Provides insights for practitioners working at the intersection of software development and reconfigurable hardware.[cite: 2]

OpenCL Performance on the Intel Heterogeneous Architecture Research Platform IEEE HPEC 2020

Steven Harris, Roger D. Chamberlain, Christopher Gill

Investigates the utility of traditional matrix multiplication optimizations on the Intel HARPv2 heterogeneous platform using OpenCL. Provides targeted insights into performance portability and best practices for emerging heterogeneous architectures.

Investigating Single Precision Floating General Matrix Multiply in Heterogeneous Hardware Washington University Thesis

Steven Harris

Master's thesis exploring SGEMM optimizations on heterogeneous architectures using the Intel HARP2 platform and High Level Synthesis compiler. University Distinguished Thesis Award nominee.

Supporting graceful degradation through elasticity in mixed-criticality federated scheduling WMC 2018

Chris Gill, James Orr, Steven Harris

Introduces a novel multi-criticality elastic task model combining Mixed-Criticality Federated Scheduling with elastic scheduling of parallel real-time tasks for improved resource utilization.

04.

Experience

2019 — Present

Director of Information Technology

Leading IT strategy and infrastructure for one of the nation's top medical research institutions. Overseeing enterprise systems, data architecture, and technology initiatives supporting clinical research and academic operations. Managing enterprise-scale deployments and ensuring HIPAA compliance across healthcare systems.

Healthcare IT Enterprise Architecture Data Security HIPAA Cloud Infrastructure
2019 — Present

PhD Researcher

Conducting cutting-edge research in heterogeneous computing under the guidance of Dr. Roger Chamberlain and Dr. Christopher Gill. Published in top-tier venues including IEEE HPEC and SC (Supercomputing). Collaborating with industry partners on next-generation computing architectures.

Research Heterogeneous Computing FPGA Performance Modeling
2015 — 2019

Senior IT Positions

Washington University School of Medicine

Held various leadership positions in IT, managing enterprise systems, infrastructure upgrades, and digital transformation initiatives for a major research university medical center.

IT Leadership Infrastructure System Administration
05.

Technical Skills

Hardware & Architecture

FPGA Design Verilog/VHDL High-Level Synthesis RTL Design GPU Architecture Hardware Acceleration

Programming Languages

C/C++ Python OpenCL CUDA SystemVerilog Assembly

Frameworks & Tools

Intel HLS Vivado HLS OpenMP MPI CUDA Toolkit ROCm

Research Areas

Real-Time Systems Scheduling Theory Performance Analysis Embedded Systems Cyber-Physical Systems HPC

Operating Systems

Linux RTOS FreeRTOS Bare-metal Windows Server

Industry Skills

Project Management Team Leadership Technical Writing Peer Review STEM Education
06.

Teaching

CSE 560 — Domain Specific Architectures

Fall 2023 | Washington University

CSE 560 — Superscalar Processors

Fall 2023 | Washington University

CSE 560 — Technology & Innovation

Fall 2023 | Washington University

CSE 560 — Domain Specific Accelerators

Fall 2022 | Washington University
✦

Philosophy

"We must look for ways to be an active force in our own lives. We must take charge of our own destinies, design a life of substance and truly begin to live our dreams."

— Les Brown

"The master has failed more times than the beginner has even tried."

— Stephen McCranie

I believe that the most significant breakthroughs in computing come from understanding the fundamental limitations of current systems and daring to reimagine them. Every challenge—be it in real-time scheduling, heterogeneous architecture design, or hardware-software co-optimization—represents an opportunity to build something better. My philosophy is simple: embrace the hard problems, learn from every failure, and never stop pushing the boundaries of what's possible.

07. What's Next?

Let's Build Something Amazing

I'm always interested in research collaborations, speaking opportunities, and discussions about the future of computing. Whether you're from a top tech company, defense contractor, or academic institution—let's talk.

↓ Download Full CV