View: |
Part 1: Document Description
|
Citation |
|
---|---|
Title: |
Experiences Porting Shared and Distributed Applications to Asynchronous Tasks: A Multidimensional FFT Case-study |
Identification Number: |
doi:10.18419/darus-4094 |
Distributor: |
DaRUS |
Date of Distribution: |
2024-04-02 |
Version: |
1 |
Bibliographic Citation: |
Strack, Alexander; Taylor, Christopher; Diehl, Patrick; Pflüger, Dirk, 2024, "Experiences Porting Shared and Distributed Applications to Asynchronous Tasks: A Multidimensional FFT Case-study", https://doi.org/10.18419/darus-4094, DaRUS, V1 |
Citation |
|
Title: |
Experiences Porting Shared and Distributed Applications to Asynchronous Tasks: A Multidimensional FFT Case-study |
Identification Number: |
doi:10.18419/darus-4094 |
Authoring Entity: |
Strack, Alexander (Universität Stuttgart) |
Taylor, Christopher (Tactical Computing Labs LLC) |
|
Diehl, Patrick (Louisiana State University) |
|
Pflüger, Dirk (Universität Stuttgart) |
|
Distributor: |
DaRUS |
Access Authority: |
Strack, Alexander |
Access Authority: |
Pflüger, Dirk |
Access Authority: |
Scientific Computing |
Depositor: |
Strack, Alexander |
Date of Deposit: |
2024-03-13 |
Holdings Information: |
https://doi.org/10.18419/darus-4094 |
Study Scope |
|
Keywords: |
Computer and Information Science, HPX, Multidimensional FFT, Asynchronous Many-Task Runtime Models, Shared-Memory Computing, Distributed-Memory Computing |
Abstract: |
<p>The source code and benchmark scripts related to "Experiences Porting Shared and Distributed Applications to Asynchronous Tasks: A Multidimensional FFT Case-study".</p> <p>This paper conducts a case study of the multidimensional Fast Fourier Transform to identify which applications will benefit from the asynchronous many-task model. Our basis is the popular FFTW library. We use the asynchronous many-task model HPX and a one-dimensional FFTW backend to implement multiple versions using different HPX features and highlight overheads and pitfalls during migration. Furthermore, we add an HPX threading backend to FFTW. The case study analyzes shared memory scaling properties between our HPX-based parallelization and FFTW with its pthreads, OpenMP, and HPX backends. The case study also compares FFTW's MPI+X backend to a purely HPX-based distributed implementation.</p> <p>Find more information how to run the code and scripts in the <a href="https://darus.uni-stuttgart.de/file.xhtml?fileId=286287">README.md</a>.</p> |
Methodology and Processing |
|
Sources Statement |
|
Data Access |
|
Other Study Description Materials |
|
Label: |
CMakeLists.txt |
Notes: |
text/plain |
Label: |
compile.sh |
Notes: |
application/x-sh |
Label: |
distributed_benchmark.sh |
Notes: |
application/x-sh |
Label: |
LICENSE |
Notes: |
text/plain; charset=US-ASCII |
Label: |
README.md |
Notes: |
text/markdown |
Label: |
shared_benchmark.sh |
Notes: |
application/x-sh |
Label: |
install_fftw_hpx.sh |
Notes: |
application/x-sh |
Label: |
install_fftw_omp_mpi.sh |
Notes: |
application/x-sh |
Label: |
install_fftw_threads_mpi.sh |
Notes: |
application/x-sh |
Label: |
fftw_hpx.cpp |
Notes: |
text/x-c |
Label: |
fftw_mpi_omp.cpp |
Notes: |
text/x-c |
Label: |
fftw_mpi_threads.cpp |
Notes: |
text/x-c |
Label: |
fft_hpx_loop.cpp |
Notes: |
text/x-c |
Label: |
fft_hpx_loop_shared.cpp |
Notes: |
text/x-c |
Label: |
fft_hpx_task_agas.cpp |
Notes: |
text/x-c |
Label: |
fft_hpx_task_agas_shared.cpp |
Notes: |
text/x-c |
Label: |
fft_hpx_task_naive_shared.cpp |
Notes: |
text/x-c |
Label: |
fft_hpx_task_shared.cpp |
Notes: |
text/x-c |
Label: |
fft_hpx_task_sync_shared.cpp |
Notes: |
text/x-c |
Label: |
vector_2d.hpp |
Notes: |
application/octet-stream |