SCAPE Lab

The SCAlable PErformance (SCAPE) laboratory is directed by Professor Kirk W. Cameron. Our main research focus is the design, analysis and improvement of scalable systems and applications using emerging technologies. Our current projects include:

High-end computing performance modeling and analysis

This project involves creating formal mathematical and statistical techniques to model, predict, and explain the why’s and how’s of performance on real systems. In this project we attempt to identify performance bottlenecks in high-end systems and applications using techniques such as analytical models and hardware counters.

High-performance, power-aware computing

This project involves creating software and hardware techniques to measure and control emergent power-aware capabilities on real systems. In this project we attempt to identify code phases and control power-aware scheduling to improve the energy efficiency of systems without negatively impacting performance.

High-performance systems and applications

The theme of these topics is usually creation of a new hardware or software technique that will aid the HPC community. For example, we have successfully deployed the PBPI (Parallel Bayesian Phylogenetic Inference) toolkit that allows biologists to identify phylogenetic trees faster than previously possible. This toolkit is now in use on IBM’s Blue Gene machine – the world’s fastest supercomputer.