Project URA-HPC

MOCCA
MOnte Carlo routines for fractional Calculus and linear Algebra is a parallel C++ template library that focus on the efficient computation of matrix functions using Monte-Carlo methods. Documentation.

HighFM
HighFM is a high-performance numerical package with primary focus on the evaluation of matrix functions. Using modern C++ features, HighFM enables users to write high-level code similar to the mathematical notation without sacrificing performance, translating all operations to highly-optimized kernels from well-known BLAS libraries, such as Intel MKL, AMD AOCL or NVIDIA CUDA.

ParKrylovML
Hybrid (both distributed- and shared-memory) parallel implementation of a Krylov method that computes the Mittag-Leffler function of matrices.

Power Grid Simulator
Simulator for power grid of integrated circuits using the parallel Monte Carlo routines of HighFM.

Parallel Finite Element Algorithm
Fully automatic, parallel hp-adaptive finite element algorithm for electromagnetic simulation using a task-based approach with data dependency management.

nach oben