Projects / Programmes
High performance computing algorithms in theoretical physics
Code |
Science |
Field |
Subfield |
1.02.00 |
Natural sciences and mathematics |
Physics |
|
Code |
Science |
Field |
P190 |
Natural sciences and mathematics |
Mathematical and general theoretical physics, classical mechanics, quantum mechanics, relativity, gravitation, statistical physics, thermodynamics |
P210 |
Natural sciences and mathematics |
Elementary particle physics, quantum field theory |
P220 |
Natural sciences and mathematics |
Nuclear physics |
P260 |
Natural sciences and mathematics |
Condensed matter: electronic structure, electrical, magnetic and optical properties, supraconductors, magnetic resonance, relaxation, spectroscopy |
excited baryonic states, weak decays, Lattice chromodynamics, standard model, three-body problem, computational physics, ionization, polymers, lipide bilayers, biological membranes, proteins, gels, colloids, liquid crystals, chirality, biophysics, relaxor ferroelectrics, perovskite ferroelectrics, computer networks, strongly correlated electrons, high-temperature superconductivity, quantum wires, surface reconstruction, quantum chaos, high-performance computing, optimization
Researchers (20)
Organisations (2)
Abstract
The proposed interdisciplinary project deals with the study of several algorithms from the following partly overlapping wider fields of theoretical physics: theory of superconductivity, three-body problem, nonlinear systems and biophysics. The purpose of the project is to increase the efficiency of the group"s own code packages on available hardware. This implies better analytical and mathematical algorithms as well as code optimization. For example, in the theory of superconductivity we shall study code optimization for exact Hamiltonian diagonalization, which is a memory-intensive problem. In the three-body problem we shall poerform precise calculations in ionization, and generalizations of the new quasilinearization method. The latter is a processor-intensive problem, parallelizable using directives alone. Problems related to complex systems and nonlinear dynamics typically belong to the class allowing message-passing treatment on hardware with many processors. Thre expected results will be generally applicable, despite the fact that they stem from well-defined physical problems. They will also improve the efficiency of usage of existing hardware, which has been judiciously built out of memory and network-intensive and processor-intensive components.