Neil J. Gunther

Content sourced from Wikipedia, licensed under CC BY-SA 3.0.

Neil James Gunther (born 15 August 1950) is an Australian computer science researcher known for his work in performance analysis and capacity planning. He created the open‑source performance modeling tool PDQ (Pretty Damn Quick) and developed the Guerrilla approach to capacity planning and performance analysis. He is also known for the theory of large transients in computer systems and the Universal Scalability Law, which helps explain how throughput grows and then levels off as more processors or users are added.

Gunther’s career spans academia, industry research, and consulting. He taught physics at San Jose State University, then worked at Syncal for NASA/JPL on thermoelectric materials for deep‑space missions, influencing the Galileo project. He joined Xerox PARC in 1982, contributing to VLSI test software and the Dragon multiprocessor, and later devised a Wick‑rotated version of Feynman’s path integral to analyze performance in large systems. In 1990 he moved to Pyramid Technology, leading the Performance Analysis Group and helping raise Unix multiprocessor benchmarks.

In 1994 he founded Performance Dynamics Company to provide performance analysis and capacity planning services. Around 1998 he released PDQ and wrote The Practical Performance Analyst, followed by several other books. He has since explored quantum information system technologies, including a theory of photon bifurcation tested at EPFL, applying path integral ideas to optics and imaging. Gunther also created Barry, a visualization tool for CPU usage data, and later applied similar methods to visualize application and network performance. In 2008 he helped form the PerfViz Google group.

Gunther’s Universal Scalability Law expresses throughput as a function of the number of processors or users, with components representing contention, coherency, and concurrency. It helps practitioners estimate maximum useful capacity and identify where improvements will be most effective. He remains active in the field as a senior member of ACM and IEEE and continues research in quantum information systems, including work at EPFL.