Scientific Computing Environments# Contents: Parallel Computing Serial applications Symmetric multiprocessing (SMP) MPI parallel applications Hybrid parallel applications Embarrassingly parallel applications GPU computing Running GPU applications Profiling GPU applications Multi-GPU Multi-host GPU applications Python Jupyter notebooks Jupyter notebook job on a compute node Connect to the jupyter server from a client Running production jobs with Jupyter notebooks Astrophysics Applications and Libraries Rebound Nbody family of codes NGWire - secular dynamics framework Enzo - multi-physics hydrodynamic astrophysical calculations Molecular dynamics LAMMPS GROMACS HOOMD Machine Learning - Deep Learning - Artificial Intelligence jobs Deep learning frameworks Hardware optimized for deep learning Allocating GPU resources Using tensorflow, Keras or pytorch Deep learning jobs tips and best practices Distribued training and inference with torch Distribued training with tensorflow and keras Troubleshooting Large language models Environments Available models in the model library Running inference and evaluating models Fine-tuning Sharding Serving models using ollama Quantizing models The Matlab environment Overview Matlab as a client on Windows Matlab on the compute nodes of the cluster The Ansys Fluent environment Overview Oil and Gas Applications Eclipse - Petrel Abaqus Graphical user interface mode Template Abaqus job (batch mode) Multi-node parallel Abaqus jobs User defined subroutines and functions in Abaqus CST Pari - Computer Algebra Introduction Running Pari (single threaded)