Supplementary Software for Comparison of Ensemble-Based Data Assimilation Methods for Sparse Oceanographic Data

This release represent the supplementary software for the paper Comparison of Ensemble-Based Data Assimilation Methods for Sparse Oceanographic Data (Section 4) by Florian Beiser, Håvard Heitlo Holm, and Jo Eidsvik.

It contains the software for setting up and running data-assimilation and drift trajectory forecasting experiment, as well as Jupyter Notebooks for post-processing and visualization of the results. The folder gpu_ocean/papers/DAComparison contains the code specific for this publication.

Supplementary Software for Coastal Ocean Forecasting on the GPU using a Two-Dimensional Finite-Volume Scheme

This release represent the supplementary software for the paper Coastal Ocean Forecasting on the GPU using a Two-Dimensional Finite-Volume Scheme by André Rigland Brodtkorb and Håvard Heitlo Holm.

It contains the software described in the paper, and Jupyter notebooks for setting up and running and visualizing the numerical experiments. The folder gpu_ocean/papers/realisticSimulations contains the code specific to this work.

Supplementary Software for Data Assimilation for Ocean Drift Trajectories Using Massive Ensembles and GPUs

This release represent the supplementary software for the paper Data Assimilation for Ocean Drift Trajectories Using Massive Ensembles and GPUs by Håvard Heitlo Holm, Martin Lilleeng Sætra and André Rigland Brodtkorb.

It contains the software for setting up and running data-assimilation and drift trajectory forecasting experiments, as well as Jupyter Notebooks for post-processing and visualization of the results. The folder gpu_ocean/demos/MPI_SIR contains the code specific to this work.

Supplementary Software for Massively Parallel Implicit Equal-Weights Particle Filter for Ocean Drift Trajectory Forecasting

This release represent the supplementary software for the paper Massively Parallel Implicit Equal-Weights Particle Filter for Ocean Drift Trajectory Forecasting by Håvard Heitlo Holm, Martin Lilleeng Sætra and Peter Jan van Leeuwen.

It contains the software for setting up and running data-assimilation and drift trajectory forecasting experiment, as well as Jupyter Notebooks for post-processing and visualization of the results.

All figures and results presented in the paper can be reproduced from the notebooks provided here, combined by the supplementary data published under DOI 10.5281/zenodo.3457538.

Supplementary Material to Evaluation of Selected Finite-Difference and Finite-Volume Approaches to Rotational Shallow-Water Flow

This release represent the supplementary material for the paper Evaluation of Selected Finite-Difference and Finite-Volume Approaches to Rotational Shallow-Water Flow by Holm, Brodtkorb, Broström, Christensen and Sætra, and contains the numerical schemes and test cases used in the paper.

All figures and results presented in the paper can be reproduced from the notebooks provided here.

eval-findiff-finvol-rotSWE

This release represent the supplementary material for the paper Evaluation of Selected Finite-Difference and Finite-Volume Approaches to Rotational Shallow-Water Flow by Holm, Brodtkorb, Broström, Christensen and Sætra, and contains the numerical schemes and test cases used in the paper.

All figures and results presented in the paper can be reproduced from the notebooks provided here.

cpuCOP-main-cuda-tuned-sub

The main version of the GPU Ocean code used to obtain the submitted results of the paper "Comparing performance and usability of CUDA and OpenCL with Python", by Brodtkorb, Holm and Sætra.

The paper compares four releases of the GPU Ocean code:

• cpuCOP-opencl-original-sub: The original OpenCL code
• cpuCOP-cuda-original-sub: The original CUDA code, ported from original OpenCL
• cpuCOP-main-cuda-tuned-sub: The tuned CUDA code, and scripts for running the experiments
• cpuCOP-opencl-tuned-sub: The tuned OpenCL code, ported from tuned CUDA code.

(bold marks this release)

cpuCOP-opencl-tuned-sub

The tuned OpenCL version of the GPU Ocean code used to obtain the submitted results of the paper "Comparing performance and usability of CUDA and OpenCL with Python", by Brodtkorb, Holm and Sætra.

The paper compares four releases of the GPU Ocean code:

• cpuCOP-opencl-original-sub: The original OpenCL code
• cpuCOP-cuda-original-sub: The original CUDA code, ported from original OpenCL
• cpuCOP-main-cuda-tuned-sub: The tuned CUDA code, and scripts for running the experiments
• cpuCOP-opencl-tuned-sub: The tuned OpenCL code, ported from tuned CUDA code.

(bold marks this release)

cpuCOP-opencl-original-sub

The original untuned OpenCL version of the GPU Ocean code used to obtain the submitted results of the paper "Comparing performance and usability of CUDA and OpenCL with Python", by Brodtkorb, Holm and Sætra.

The paper compares four releases of the GPU Ocean code:

• cpuCOP-opencl-original-sub: The original OpenCL code
• cpuCOP-cuda-original-sub: The original CUDA code, ported from original OpenCL
• cpuCOP-main-cuda-tuned-sub: The tuned CUDA code, and scripts for running the experiments
• cpuCOP-opencl-tuned-sub: The tuned OpenCL code, ported from tuned CUDA code.

(bold marks this release)

cpuCOP-cuda-original-sub

The original untuned CUDA version of the GPU Ocean code used to obtain the submitted results of the paper "Comparing performance and usability of CUDA and OpenCL with Python", by Brodtkorb, Holm and Sætra.

The paper compares four releases of the GPU Ocean code:

• cpuCOP-opencl-original-sub: The original OpenCL code
• cpuCOP-cuda-original-sub: The original CUDA code, ported from original OpenCL
• cpuCOP-main-cuda-tuned-sub: The tuned CUDA code, and scripts for running the experiments
• cpuCOP-opencl-tuned-sub: The tuned OpenCL code, ported from tuned CUDA code.

(bold marks this release)