PY4C01

Module PY4C01 Computer Simulation III

Cohort: SS Physics and Astrophysics, SS Theoretical Physics (optional)

Credits: 5

Lecturer: Professors G. Watson, M. Mobius

Duration: Hilary Term, High Performance Computing, 9 lectures, Numberical Methods II, 14 lectures

Assesment: End of Year Exam.

Description: Part I: High Performance Computing: Code analysis performance optimisation including code monitoring, profiling and tuning. Discussion of different machine architectures and related optimisation including superscalar, vector and parallel computing (distributed and shared memory). Introduction to parallel computing definition of scale up and efficiency, problem decomposition and load balancing. Introduction to parallel coding using MPI, including communicators, point to point communications and collective communications.

Part II: Numerical Methods: Fourier series, discrete Fourier transform, Fast Fourier transform, Signal filtering, Windowing, Gaussian integration , Monte Carlo integration, Metropolis algorithm, variational Monte Carlo, Ising model, finite size scaling, simulated annealing, Travelling Salesman problem.