Dr. Michael Manzke

Introductory remarks: I turned 65 this year and could have retired in October, but decided to work another year until October 2027. Consequently, I couldn"t apply for further research funding or take on additional duties. This application process should evaluate whether my achievements over the last 27 years at Trinity warrant the promotion to an associate professor.

I have a somewhat unusual biography. During and after my primary engineering degree, I worked in industry for about 10 years before pursuing a PhD at Trinity at the age of 37. As part of my degree, I spent a year as an intern at Philips. During that time, I also worked on my dissertation. I developed and evaluated a method for measuring high-voltage electrostatic discharge (ESD) on integrated circuits (ICs). I achieved this by measuring these discharges with a scanning electron microscope (SEM). This method allowed Philips to evaluate protective circuits on their ICs.

In the subsequent years, I worked in the semiconductor industry on quartz crystals that generate timing signals for computers, but for most of these years, I worked on large-scale process control systems for the bulk chemical industry, the pharmaceutical industry, and clean rooms in the semiconductor industry.

In 1998, I started as a research assistant and PhD student on an EU project in computer science at Trinity. The project was concerned with a high-speed interconnect for personal computers (PC). This interconnect allowed for the configuration of clusters of PCs that were closely coupled and could share their memory.

In 1999, I successfully applied for a three-year contract as a lecturer, and in 2000 I was appointed as a permanent lecturer in the Department of Computer Science. From the second year of my PhD, I had a full teaching load and supervised PhD students. My many years of experience in advanced control theory for real-time systems influenced my thesis topic. I designed and evaluated a distributed Kalman filter that was implemented over a cluster of PCs, a mathematical algorithm that estimates the true state of a changing compute system by combining a physical prediction with noisy performance counter measurements of the PCs and the interconnect. This allowed me to determine the computational state of the PC cluster in real-time.

Other topics that influenced my research were reconfigurable chips, such as Field-Programmable Gate Arrays (FPGAs), Computer Graphics, and Computer Vision. Interactive Computer Graphics and Vision applications provided my postdocs, my PhD students, and me with problems that are computationally very demanding and need to be solved in real-time.

FPGAs allowed us to design and evaluate novel integrated circuits for Computer Graphics and Vision hardware without the need to manufacture IC chips. The main thrust of my research is focused on hardware and software for high-performance real-time Computer Graphics and Vision systems.

In 2018, for the first time, consumer graphics processing units (GPUs) for Real-time ray tracing became commercially available. Ray tracing can synthesize photorealistic images. These GPUs had integrated circuits that we published in 2013. Previously, we had worked on ray tracing for many years.

More recently, our work addresses the problem of real-time interactive visualisation of 3D scientific data, namely volumetric datasets, which are ubiquitous in many application domains such as Medicine, Biology, Biomechanics, Neuroscience, Fluid Dynamics and Veterinary Medicine. Furthermore, our work focuses on low-power architectures for computer graphics, deep learning, and light fields for volume rendering.

Much of this work was done in collaboration with industry. We also tried to commercialise some of our research.