معرفی
Paul H J Kelly is a Professor of Software Technology at Imperial College London, leading the Software Performance Optimisation research group. He serves as co-Director of the Centre for Computational Methods in Science and Engineering and Director of Industrial Liaison for the HiPEDS Centre for Doctoral Training in High-Performance Embedded and Distributed Systems.
Research Interests
His research focuses on:
- Compiler technology for computational science
- Performance portability across heterogeneous architectures
- Domain-specific languages (DSL) for scientific computing
- Optimization of finite element methods and PDE solvers
- Data locality and parallelism trade-offs
- Computer vision algorithms and SLAM systems
He actively collaborates with hardware vendors and application developers in computational science, robotics, and quantum chemistry.
Article Trends
Recent publications emphasize:
- Temporal and spatial tiling for PDEs and stencil computations
- Quantum circuit simulation optimization
- Distributed SLAM systems
- Performance portability frameworks (e.g., Firedrake, Devito)
- Compiler techniques for GPUs and custom accelerators
- Memory hierarchy optimization
Scientific Awards
- Senior Member of the ACM (2021)
- Imperial College Engineering Faculty Teaching Excellence Award (2013)
- Best Robotics Paper at 18th Conference on Robots and Vision (2021)
Student Mentoring
He has mentored numerous PhD and postdoctoral researchers now in academic positions including:
- Luigi Nardi - Assistant Professor at Lund University
- Sajad Saeedi - Assistant Professor at Ryerson University
- Lawrence Mitchell - Assistant Professor at University of Durham
Current students include Renato Salas-Moreno, David Ham, and Miklos Homolya.
۰مقاله منتشرشده
