Computing

Scientific Computing

Computational physics, micromagnetics simulations, numerical modelling, and high-performance computing resources for researchers and engineers.

Micromagnetics

Computational micromagnetics tools and numerical modelling methods. Solve the Landau-Lifshitz-Gilbert equation with MuMax3 and OOMMF.

MuMax3OOMMFGNUPlotFEMMonte Carlo

Explore

HPC & Simulation

High-performance computing infrastructure, CUDA programming, GPU-accelerated workflows, and parallel computing paradigms for large-scale simulations.

CUDAGPU ComputingParallel WorkflowsBenchmarks

Explore

OpenFOAM

Open-source CFD toolbox for solving Navier-Stokes, heat transfer, multiphase flow, and turbulence using the finite volume method.

Finite VolumeNavier-StokesTurbulenceMultiphase

Explore

Gmsh

3D finite element mesh generator with built-in CAD engine. Structured, unstructured, and hybrid meshing with OpenCASCADE support.

CADMeshingDelaunayOpenCASCADE

Explore

FEniCS

Solve PDEs using the finite element method with near-mathematical notation. Automated code generation from UFL variational forms.

FEMUFLDOLFINxPETSc Backend

Explore

deal.II

C++ finite element library with hp-adaptive refinement, multigrid solvers, and massively parallel computing on 100K+ cores.

hp-AdaptivityMultigridC++PETSc / Trilinos

Explore

PETSc

Portable, Extensible Toolkit for Scientific Computation. Scalable linear/nonlinear solvers, time integrators, and optimisation for distributed PDEs.

KSPSNESTSMPI Distributed

Explore

Performance Benchmarks & Profiling

Sample Data

*Note: This is sample graph data illustrating expected metrics for our HPC clusters when solving nonlinear PDEs (10M+ DOFs) using the FEniCS/PETSc backend.

Strong Scaling (Time to Solution)

100s75s50s25s0s

1 Node2 Nodes4 Nodes8 Nodes16 Nodes

Memory Allocation Profile (GB)

128G96G64G32G0G

Mesh

Assembly

Krylov

Precond.

KSP Solver Convergence (Residual Norm)

1e01e-31e-61e-91e-12

020406080100 iters