Fahad Khalid

Complex Systems Summer School 2014 Home Surveys/Evaluations Logistics Payment Instructions Faculty Lecture Slides Panel Questions Panel Responses Participants 1-Minute Summaries Schedule Background Readings Software Projects & Working Groups Project Group Meeting Schedule Sign Up Sheet for Meetings with Miguel and Josh Tutorials Blog Project Presentations Final Papers After Hours Restaurant Guide Resource Guide Photos TShirts

I'm a PhD student with the HPI Research School on Service Oriented Systems Engineering, and member of the Operating Systems and Middleware Group.

Primary research theme is programming models for hybrid CPU-GPU parallel computing. Devising strategies to balance the productivity vs. performance trade-off inherent in programming massively parallel hybrid architectures. Currently working on the architecture-based algorithm decomposition approach, which uses patterns and certain characteristics of a computational kernel to map the kernel onto the most suitable processing device in a hybrid architecture. Also working on developing mathematical foundations for patterns in parallel programming.

Use cases are based on real-world problems in Computational Biology and Geophysics. These use cases have been acquired during collaborative simulation development projects with the Max Planck Institute of Molecular Plant Physiology and the GFZ German Research Center for Geosciences.

I'm currently also supervising a project on developing a GPU-accelerated Path Integral Monte Carlo algorithm for simulating Quantum Annealing in the Random Field Ising Model. The objective is to develop a high performance simulator for Adiabatic Quantum Computing (AQC) chips. Such a simulator will make it possible to design and evaluate programming models for AQC.

Prior to joining the HPI, I worked as software developer for large-scale telecommunications software systems, such as the Ericsson Real-time Charging system. I also worked as system architect for a Provisioning & Service Activation product. My professional experience is in software design, development as well as automated testing. I studied Computer Science in Pakistan, and Artificial Intelligence in Sweden. website

Project Ideas:

1. Cells and Software: Is Evolution a Software Engineer - I've done some work on finding similarities between design principles in software engineering, and those employed at the cellular and molecular level in living cells. The hypothesis is that evolution and software engineers came up with similar solutions to design problems because both were trying to maximize a certain objective function, while operating within certain resource constraints. Is this true? Can we find design patterns that transcend discipline boundaries? If so, we could improve software engineering processes by learning from cells, and perhaps learn more about cells by applying design patterns in software engineering.

2. Future Architectures for Massively Parallel Processing - Our supercomputers today are based on the Von Neumann architecture. This architecture has become a bottleneck in the sense that it is getting harder and harder to write applications that scale to millions of cores. Perhaps we need to think about alternate architectures such as Adiabatic Quantum Computing using Spin Lattices, or Artificial Neural Networks. Both do optimization, so we need to find out how to express general computations as mathematical programs. Also, what are the similarities and differences between AQC and ANN based computation. How should we proceed?