Physics Β· Core roadmap Β· Major topic 34 of 43
Computational Physics, Modelling & Scientific Computing
Computational skills support nearly every modern subfield: solving equations numerically, interpreting data and testing models.
38 items insideAsked 1Γ in the RPSC/RSSB papers we analysed, most recently 2012PHY-34
Also look at: Computer: Computational Science and Interdisciplinary ComputingScience & Technology: Mathematics, Statistics and Data FoundationsComputer: Computational Thinking, Problem Solving and DesignChemistry: Computational, Theoretical & Quantum Chemistry
- 1
Historical evolution β key dates
- Electronic computers enable new numerical physics.1940s
- Monte Carlo methods are formalized in early computational research.1949
- Numerical weather prediction and computational fluid dynamics expand.1950sβ1960s
- Molecular dynamics and lattice simulations grow.1970s onward
- High-performance computing reshapes simulations.1990sβ2000s
- GPUs, machine learning and open scientific workflows expand.2010sβ2020s
- 2
Programming foundations
- Variables, types, control flow, functions, arrays and plotting
- Units-aware code; tests; debugging; notebooks; version control
5 more points
- Units-aware code
- Tests
- Debugging
- Notebooks
- Version control
- 3
Numerical methods
- Floating-point arithmetic; truncation/roundoff; stability and convergence
3 more points
- Floating-point arithmetic
- Truncation/roundoff
- Stability and convergence
- Root finding, interpolation, differentiation and numerical integration
- ODE/PDE solvers; finite difference, finite volume and finite element methods
2 more points
- ODE/PDE solvers
- Finite difference, finite volume and finite element methods
- Floating-point arithmetic; truncation/roundoff; stability and convergence
- 4
Simulation and scale
- Monte Carlo; molecular dynamics; N-body and lattice models
3 more points
- Monte Carlo
- Molecular dynamics
- N-body and lattice models
- CFD, plasma, climate and astrophysical simulations
- Parallel computing, GPUs, HPC, cloud and performance profiling
- Monte Carlo; molecular dynamics; N-body and lattice models
- 5
Data and scientific AI
- Parameter estimation; inverse problems; uncertainty quantification
3 more points
- Parameter estimation
- Inverse problems
- Uncertainty quantification
- Visualisation, reproducible pipelines and open data
- Machine learning, surrogate models and physics-informed neural networks
- Parameter estimation; inverse problems; uncertainty quantification
