Use of Computers in Physics

Abstract

Computer physics, as a branch of computational science, is critical for today's scientific endeavours. It is a field in which supercomputers are used to do computations based on established physical laws and to simulate conditions that are difficult to achieve with existing technology, or that can only be realised through experiments, or in which large amounts of experimental data must be handled. Varying study domains have different computational resource demands, but computational physics applications in general require extremely high floating point execution power and memory bandwidth, hence such computations can only be done in reasonably large clusters. Computational physics is that the study and implementation of numerical analysis to unravel problems in physics that a quantitative theory already exists. Historically, computational physics was the primary application of recent computers in science, and is now a subset of computational science. It is sometimes considered a sub discipline (or offshoot) of theoretical physics, but others consider it an intermediate branch between theoretical and experimental physics, a neighborhood of study which supplements both theory and experiment [1]. Problems in computational physics are notoriously difficult to solve precisely. This is due to a variety of (mathematical) factors, including the absence of algebraic and/or analytic solvability, complexity, and chaos. Even seemingly simple problems, such as calculating the wave function of an electron orbiting an atom during a strong field (Stark effect), may necessitate a significant amount of effort to formulate a practical algorithm (if one can be found at all); other cruder or brute-force techniques, such as graphical methods or root finding, may also be necessary [2]. On the more advanced end of the spectrum, mathematical perturbation theory is occasionally employed. In addition, the computational cost and computational complexity for many-body problems (and their classical counterparts) tend to grow quickly. A macroscopic system typically features a size of the order of constituent particles, so it's somewhat of a drag. Solving quantum mechanical problems is usually of exponential order within the size of the system and for classical N-body it's of order N-squared. Finally, many physical systems are inherently nonlinear at the best and at the worst chaotic: this suggests it is often difficult to make sure any numerical errors don't grow to the purpose of rendering the 'solution' useless. Due to the broad class of problems computational physics deals, it's an important component of recent research in several areas of physics, namely: accelerator physics, astrophysics, hydraulics (computational fluid dynamics), lattice field theory/lattice gauge theory (especially lattice quantum chromo dynamics), physics (see plasma modelling), simulating physical systems (using e.g. molecular dynamics), engineering computer codes, protein structure prediction, weather prediction, solid state physics, soft condensed matter physics, hypervelocity impact physics etc. Computational solid state physics, for instance, uses density functional theory to calculate properties of solids, a way almost like that employed by chemists to review molecules [3]. Other quantities of interest in solid state physics, like the electronic band structure, magnetic properties and charge densities are often calculated by this and a number of other methods, including the Luttinger-Kohn/k.p method and ab-initio methods. For each major subject of physics, such as computational mechanics and computational electrodynamics, a matching computational branch can be found. Computational mechanics includes Computational Fluid Dynamics (CFD), Solid Mechanics (Computational Solid Mechanics), and Computational Contact Mechanics (Computational Contact Mechanics). Computational magneto hydrodynamics is an area at the intersection of CFD and electromagnetic modelling. The quantum many-body issue naturally leads to computational chemistry, a vast and fast expanding area [4]. VASP software is based on first-principle theory and computes utilising the pseudopotential plane wave approach, and is widely used in computing for solid state physics, materials science, and particle physics, among other fields. It's employed in astronomy, geophysics, and other fields as well, making it one of the most essential scientific applications. For many years, Inspur has undertaken extensive theoretical and algorithmic research on VASP, gaining valuable insights.