Kernels
BabelStream
Our great sponsors
| Kernels | BabelStream | |
|---|---|---|
| 6 | 1 | |
| 401 | 309 | |
| 0.5% | 5.5% | |
| 7.2 | 8.0 | |
| 11 days ago | 7 days ago | |
| C | C++ | |
| GNU General Public License v3.0 or later | GNU General Public License v3.0 or later |
Stars - the number of stars that a project has on GitHub. Growth - month over month growth in stars.
Activity is a relative number indicating how actively a project is being developed. Recent commits have higher weight than older ones.
For example, an activity of 9.0 indicates that a project is amongst the top 10% of the most actively developed projects that we are tracking.
Kernels
-
Can you give me some proof that storing multidimansional data into a 1d array is the standard and best way to do it?
https://github.com/ParRes/Kernels/tree/default/C1z has some examples I’ve tested in the past. 2d is in the filenames of the relevant ones.
-
Fortran on GPU
I've evaluated all of these against each other. One presentation is https://www.nvidia.com/en-us/on-demand/session/gtcspring22-s41620/ (sorry, you have to register - it's not my preference). The performance numbers there are based on code derived from https://github.com/ParRes/Kernels/tree/default/FORTRAN (the code differences are not interesting). Another comparison is found in https://github.com/jeffhammond/nwchem-tce-triples-kernels, which is more complicated in some ways.
-
Cross Platform Computing Framework?
If you want to learn by viewing code side by side, https://github.com/ParRes/Kernels/tree/default/Cxx11 might be useful. I haven’t kept up with my RAJA ports because they kept making breaking changes in the API a few years ago (should be stable now).
-
Co-Array MPI issue.
Try https://github.com/ParRes/Kernels/tree/default/FORTRAN coarray programs. Those were written by people who know what they’re doing and have been proven to execute correctly before. That might help you understand if your implementation is broken.
-
I am in grad school and starting a CFD class soon. I am proficient in Python and Matlab, but the course requires Fortran. How rough of a time will I have coding difficult concepts in a new language? I’m hoping my logic skills will overcome any syntax issues I run into, but wanted to ask
https://github.com/ParRes/Kernels has examples of the same thing written in Fortran, MATLAB/Octave and Numpy, if it helps.
-
Small Open Source HPC Code Recommendations
You absolutely went to take a look at the Parallel Research Kernels (PRK) repo at https://github.com/ParRes/Kernels .
BabelStream
-
Fortran on GPU
I wrote the Fortran version of BabelStream (https://github.com/UoB-HPC/BabelStream/pull/135) to compare all of these, which might be useful as well. There is also MiniWeather, which supports all OpenMP, OpenACC and StdPar for GPU: https://github.com/mrnorman/miniWeather/tree/main/fortran.
What are some alternatives?
grbl-L-Mega - An open source, embedded, high performance g-code-parser and CNC milling controller written in optimized C that will run on an Arduino Mega2560. Forked from GRBL modified for use on a lathe with spindle sync threading
nwchem-tce-triples-kernels - NWChem TCE CCSD(T) loop-driven kernels for performance optimization experiments
analisis-numerico-computo-cientifico - Análisis numérico y cómputo científico
bedrock-clump - A GPU accelerated program to search the minecraft world for the largest connected clump of layer 5 bedrock
miniMD - MiniMD Molecular Dynamics Mini-App
gpu-kernel-runner - Runs a single CUDA/OpenCL kernel, taking its source from a file and arguments from the command-line
ompi - Open MPI main development repository
ParallelReductionsBenchmark - Thrust, CUB, TBB, AVX2, CUDA, OpenCL, OpenMP, SyCL - all it takes to sum a lot of numbers fast!
john - John the Ripper jumbo - advanced offline password cracker, which supports hundreds of hash and cipher types, and runs on many operating systems, CPUs, GPUs, and even some FPGAs
miniWeather - A parallel programming training mini app simulating weather-like flows
computecpp-sdk - Collection of samples and utilities for using ComputeCpp, Codeplay's SYCL implementation
alpaka - Abstraction Library for Parallel Kernel Acceleration :llama: