-
- 
-
- 
- -
-
- 
-
- 
- -
-
- -10.5281/zenodo.8408733-blue.svg){kind=icon}
ARTEMIS (Adaptive mesh Refinement Time-domain ElectrodynaMIcs\ + \ Solver) is a high-performance coupled electrodynamics\u2013micromagnetics solver\ + \ for full physical modeling of signals in microelectronic circuitry. The overall\ + \ strategy couples a finite-difference time-domain (FDTD) approach for Maxwell\u2019\ + s equations to a magnetization model described by the Landau\u2013Lifshitz\u2013\ + Gilbert (LLG) equation. The algorithm is implemented in the Exascale Computing\ + \ Project (ECP) software framework, AMReX, which provides effective scalability\ + \ on manycore and GPU-based supercomputing architectures. Furthermore, the code\ + \ leverages ongoing developments of the Exascale Application Code, WarpX, which\ + \ is primarily being developed for plasma wakefield accelerator modeling. Our\ + \ temporal coupling scheme provides second-order accuracy in space and time by\ + \ combining the integration steps for the magnetic field and magnetization into\ + \ an iterative sub-step that includes a trapezoidal temporal discretization for\ + \ the magnetization. The performance of the algorithm is demonstrated by the excellent\ + \ scaling results on NERSC multicore and GPU systems, with a significant (59\xD7\ + ) speedup on the GPU using a node-by-node comparison. The utility of our code\ + \ is validated by performing simulations of transmission lines, rectangle electromagnetic\ + \ waveguides, magnetically tunable filters, on-chip coplanar waveguides and resonators,\ + \ magnon-photon coupling circuits, and so on.
\n\n\ +git\
+ \ clone git@github.com:AMReX-Codes/amrex.git
git clone git@github.com:AMReX-Microelectronics/artemis.git
Make sure that the AMReX and Artemis are cloned in the same location in their\
+ \ filesystem. Navigate to the Exec folder of Artemis and execute\nmake -j\
+ \ 4.
\nYou can turn on and off the LLG equation by specifying USE_LLG\
+ \ during compilation.
\nThe following command compiles Artemis without LLG\n\
+ make -j 4 USE_LLG=FALSE
\nThe following command compiles Artemis\
+ \ with LLG\nmake -j 4 USE_LLG=TRUE
\nThe default value of USE_LLG\
+ \ is TRUE.
Example input scripts are\
+ \ located in Examples directory.
You can run the following to simulate an air-filled X-band\ + \ rectangle waveguide:
\n\nmake\
+ \ -j 4 USE_LLG=FALSE
\nmpirun -n 4 ./main3d.gnu.TPROF.MTMPI.OMP.GPUCLOCK.ex\
+ \ Examples/Waveguide/inputs_3d_empty_X_band
make -j 4 USE_LLG=FALSE USE_GPU=TRUE
\nmpirun -n 4 ./main3d.gnu.TPROF.MTMPI.CUDA.GPUCLOCK.ex\
+ \ Examples/Waveguide/inputs_3d_empty_X_band
You can run the following to simulate an X-band magnetically\ + \ tunable filter:
\n\nmake\
+ \ -j 4 USE_LLG=TRUE
\nmpirun -n 8 ./main3d.gnu.TPROF.MTMPI.OMP.GPUCLOCK.ex\
+ \ Examples/Waveguide/inputs_3d_LLG_filter
make -j 4 USE_LLG=TRUE USE_GPU=TRUE
\nmpirun -n 8 ./main3d.gnu.TPROF.MTMPI.CUDA.GPUCLOCK.ex\
+ \ Examples/Waveguide/inputs_3d_LLG_filter
Refer to the following link for several\ + \ visualization tools that can be used for AMReX plotfiles.
\n\n\nYou can extract the data in numpy\
+ \ array format using yt (you can refer to this for installation and usage of
import yt\nds = yt.load('./plt00001000/')\
+ \ # for data at time step 1000\nad0 = ds.covering_grid(level=0, left_edge=ds.domain_left_edge,\
+ \ dims=ds.domain_dimensions)\nE_array = ad0['Ex'].to_ndarray()\nACCESS-OM3 repository for spack configuration information
+shared spack configs repo
' stargazers_count: 0 - subscribers_count: 5 + subscribers_count: 4 topics: [] - updated_at: 1712101237.0 -AMReX-Codes/pyamrex: + updated_at: 1639176281.0 +C2SM/spack-c2sm: data_format: 2 - description: GPU-Enabled, Zero-Copy AMReX Python Bindings including AI/ML + description: Repository for c2sm spack config and repo files filenames: - - docs/spack.yaml - full_name: AMReX-Codes/pyamrex - latest_release: '24.05' - readme: ' -
-
-
-
-
-
-
-
-
-
-
The Python binding pyAMReX bridges the compute in AMReX block-structured codes - and data science: +
Spack is the package manager used by C2SM and MeteoSwiss to install and deploy + software on supercomputers, local machines and the cloud.
- it provides zero-copy application GPU data access for AI/ML, in situ analysis, - application coupling and enables rapid, massively parallel prototyping. + - pyAMReX enhances the Block-Structured - AMR Software Framework AMReX and its applications. +Infos about c2sm-supported software and machines
- +pyAMReX can be installed with package managers or from source.
+Please see the manual and our test - cases for detailed examples.
+Use AMReX objects and APIs from Python:
+import
- amrex.space3d as
- amr
+
+
- small_end = amr.IntVect()
+ If you are new to CMake, this short tutorial from the HEP Software foundation is the - perfect place to get started with it.
+If you just want to use CMake to build the project, jump into sections 1. - Introduction, 2. Building with CMake and 9. Finding Packages.
+pyAMReX depends on the following popular third party software.
+Optional dependencies include:
+ -spack-c2sm + v0.18.1.3 [deprecated]
-spack-c2sm + v0.18.1.2 [deprecated]
spack-c2sm + v0.18.1.1 [deprecated]
- pytest 6.2+: for - running unit testsOptional CUDA-capable dependencies for tests include:
+General infos about spack
macOS/Linux:
- -spack env activate -d .
+ With spack v0.18 we suggest local/individual spack instances and the use of
+ spack environments.
- # optional:
+ A user clones the spack repo
- # spack add cuda
+ git clone --depth 1 --recurse-submodules
+ --shallow-submodules -b v0.20.1.5 https://github.com/C2SM/spack-c2sm.git
- spack installgets spack in the command line
-(in new terminals, re-activate the environment with spack env activate
- -d . again)
.
+ spack-c2sm/setup-env.shor macOS/Linux:
+activates an environment
-brew update
+ spack env activate <path_to_env>
- brew install ccache cmake libomp mpi4py numpy open-mpi pythonand starts exploring
-Now, cmake --version should be at version 3.20.0 or newer.
spack info <package>
- Or go:
+ spack spec <spec>python3 -m pip install -U pip
+ and building
- python3 -m pip install -U build packaging setuptools wheel
+ spack install <spec>
- python3 -m pip install -U cmake
+ spack dev-build <spec>If you wish to run unit tests, then please install pytest
a package.
-python3 -m pip install -U pytestUpdating spack-c2sm is in the hands of the user.
-Some of our tests depend on optional third-party modules (e.g., pandas,
- cupy, numba, and/or pytorch).
+
git pull
- If these are not installed then their tests will be skipped.
+ git submodule update --recursiveAfter an update we advice to clean
-For example, using the Clang compiler:
+spack uninstall -a
- export
- CC=$(which clang)
+ spack clean -a
- export CXX=$(which
- clang++)
+ rm -rf ~/.spackIf you also want to select a CUDA compiler:
+and rebuild.
- +| - #export AMREX_SPACEDIM=3 + | Command | - #export AMREX_MPI=ON +
|---|---|
| Clone | +git clone --depth 1 --recurse-submodules --shallow-submodules -b <branch/tag>
+ https://github.com/C2SM/spack-c2sm.git |
- python3 -m pip install -U -r requirements.txt
+
| Load | -
- cmake --build build --target pip_install -j 8
+ . spack-c2sm/setup-env.sh autodetects machine or . spack-c2sm/setup-env.sh
+ <machine> forces machineor . spack-c2sm/setup-env.sh
+ unknown uses blank configspack compiler find autodetects compilersspack external find --all
+ autodetects externally installed packages
-
+ |
-
| Update | - python3 -m pytest tests/ +
+ git pullgit submodule update --recursive
- # Run tests from a single file
+ |
- python3 -m pytest tests/test_intvect.py
+
| Clean | - python3 -m pytest tests/test_intvect.py::test_iv_conversions +
+ spack uninstall -a uninstalls all packagesspack clean -a cleans all misc cachesrm -rf ~/.spack removes
+ user scope data |
- # Run all tests, do not capture "print"
- output and be verbose
+
If you are using the pip-driven install, selected AMReX CMake options can be controlled with environment variables:
+Spec syntax: <package>@<version>%<compiler>+<variant> ~<variant>^<sub-package> +<sub-package-variant><compiler flags>
| Environment Variable | - -Default & Values | +- | Description | +Command |
|---|---|---|---|---|
AMREX_OMP |
-
- ON/OFF + | Find | - +- | Enable OpenMP | +
AMREX_GPU_BACKEND |
-
- - - NONE/SYCL/CUDA/HIP | +Info | -On-node, accelerated GPU backend | +spack info <package> |
AMREX_MPI |
+ Spec | -ON/OFF + | - | +Enable MPI | +
AMREX_PRECISION |
-
- SINGLE/DOUBLE - - | +Install | -Precision of AMReX Real type | +spack install <spec> |
AMREX_SPACEDIM |
+ Locate | -"1;2;3" | +- | Dimension(s) of AMReX as a ;-separated list |
+
AMREX_BUILD_SHARED_LIBS |
-
- ON/OFF + | Load env | - +- | Build the core AMReX library as shared library | +
AMREX_SRC |
-
- None | +Activate + env | -Absolute path to AMReX source directory (preferred if set) | +spack env activate <env_name> |
AMREX_REPO |
-
- https://github.com/AMReX-Codes/amrex.git |
+ Deactivate + env | -Repository URI to pull and build AMReX from | +spack deactivate |
AMREX_BRANCH |
+
developmentAMREX_REPO
+
- chipStar (formerly CHIP-SPV) - + is software that allows software written to use the -
AMREX_INTERNALOFF
+ chipStar can use either the Intel Level Zero runtime or an OpenCL
- This repository contains support for building chipStar and its -
PYBIND11_INTERNALNote: most development to date has been done with the Level Zero - ON/OFF
OFF
+ the environment targeting the OpenCL backend to work.
- CMAKE_BUILD_PARALLEL_LEVELPYAMREX_LIBDIRllvm@16.0.2 and then using spack compiler add with the
+ llvm
- chipstar package defined
- llvm package anyway.
- PYAMREX_CCACHEintel-oneapi-compilers
- $ spack install intel-oneapi-compilers@2023
-
+ $ spack compiler add $(spack location
+ -i intel-oneapi-compilers@2023)/compiler/latest/linux
- PYAMREX_IPO
+
- ON/OFF
+ - Clone this repository to the target system.
- Compile with interprocedural/link optimization (IPO/LTO)
+
-
+ $ git clone https://github.com/CHIP-SPV/CHIP-SPV-Spack
-
+
- PYINSTALLOPTIONS- Activate the environment you want to build. E.g., for the
-
None
+ environment that just builds chipStar with Level Zero backend:
- Additional options for pip install, e.g., -v --user
+
-
+ $ cd
+ CHIP-SPV-Spack/Environments/LevelZero
-
+ $ spack env activate .Furthermore, pyAMReX adds a few selected CMake build options:
+ "to concretize" means to let Spack examine the package specifications -| CMake Option | + -Default & Values | +Description | +
|---|---|---|
AMReX_SPACEDIM |
+ for every dependency package, so even though there are some packages
-
+ that must be built with %clang, there are others that may be
- 3, use "1;2;3" for all |
+ built (or re-used from already-installed packages) using Dimension(s) of AMReX as a ;-separated list |
+ as the system''s GCC installation.
-
pyAMReX_CCACHE |
+ configuration options (recommended) or by editing the environment''s
-
+ spack.yaml file or other configuration options that your Spack installation
- ON/OFF |
+ is using. (Use Search and use CCache to speed up rebuilds | + using.) For instance, if you have both
pyAMReX_IPO |
+ following:
-
+ $ spack -c "packages:chipstar:require:''%clang@15.0.7''"
+ concretize -f -U |
+ Compile with interprocedural/link optimization (IPO/LTO) | + is using the compiler version you want,
pyAMReX_INSTALL |
+
-
+ $ spack install |
+ Enable install targets for pyAMReX | + such as
pyAMReX_amrex_src |
+ more information.
- None | +Absolute path to AMReX source directory (preferred if set) | + should show the packages that you just built. -
pyAMReX_amrex_internal |
+ update your environment to use the software, including:+ - ON/OFF | +Needs a pre-installed AMReX library if set to OFF
+ spack load chipstar |
+
pyAMReX_amrex_repo |
+ via this environment, https://github.com/AMReX-Codes/amrex.git |
+
- Repository URI to pull and build AMReX from | + -
pyAMReX_amrex_branch |
+ using this Spack repository. For instance, on some systems,
- development |
+ one must load the Repository branch for pyAMReX_amrex_repo
+ able to run programs that use the Intel Level Zero runtime.
- |
+
-
pyAMReX_pybind11_src |
+ None | +Absolute path to pybind11 source directory (preferred if set) | +
pyAMReX_pybind11_internal |
+ '
+ stargazers_count: 0
+ subscribers_count: 5
+ topics: []
+ updated_at: 1687550793.0
+CODARcode/z-checker-installer:
+ data_format: 2
+ description: one-key installation to install gnuplot, sz, zfp, and z-checker, and
+ complete the configuration automatically for the testing.
+ filenames:
+ - libpressio-opt/spack.yaml
+ full_name: CODARcode/z-checker-installer
+ latest_release: 0.8.0
+ readme: '
-
+ (C) 2017-2021 by Mathematics and Computer Science (MCS), Argonne National Laboratory. - ON/OFF |
+ Needs a pre-installed pybind11 library if set to OFF
+ Major authors: Sheng Di, Dingwen Tao, Hanqi Guo - |
+ Other contributors: Robert Underwood, Hengzhi Chen
-
pyAMReX_pybind11_repo |
+ https://github.com/pybind/pybind11.git |
+ Repository URI to pull and build pybind11 from | +
pyAMReX_pybind11_branch |
+ v2.12.0 |
+ Repository branch for pyAMReX_pybind11_repo
+ |
+
Python_EXECUTABLE |
+ (newest found) | +Path to Python executable | +
As one example, one can also build against a local AMReX copy. +
$HOME/src/amrex, then export
- AMREX_SRC=$HOME/src/amrex.
+ For simplicity, + + the Fedora users need to run the following command for installation:
+ +sudo dnf install -y gcc gcc-c++
+ git cmake zlib-devel libzstd-devel gfortran which xorg-x11-server-Xorg gnuplot
+ libpng-devel findutils unzip latexmk texlive-*
+
+ <!-- required texlive package:
+ "tex(comment.sty)"
+ "tex(pifont.sty)"
+ "tex(natbib.sty)"
+ "tex(amsmath.sty)"
+ "tex(morefloats.sty)"
+ "tex(geometry.sty)"
+ "tex(nopageno.sty)"
+ "tex(subfigure.sty)"
+ "tex(enumitem.sty)"
+ "tex(morefloats.sty)"-->
- Or as a one-liner, assuming your AMReX source directory is located in ../amrex:
+ git clone http://github.com/CODARcode/z-checker-installer
- AMREX_SRC=$PWD/../amrex
- python3 -m pip install -v --force-reinstall .
+ cd z-checker-installer
- Note that you need to use absolute paths for external source trees, because
- pip builds in a temporary directory.
+ ./z-checker-install.shOr build against an AMReX feature branch of a colleague. +
the Ubuntu users need to run the following command for installation:
- Assuming your colleague pushed AMReX tohttps://github.com/WeiqunZhang/amrex/
- in a branch new-feature then
+ sudo sudo apt-get install -y
+ gcc g++ git cmake zlib-devel gfortran gnuplot libpng-devel xorg openbox findutils
+ unzip latexmk texlive-full texlive-fonts-recommends --no-install-recommends
- unset
- AMREX_SRC # preferred if set
+ git clone http://github.com/CODARcode/z-checker-installer
- AMREX_REPO=https://github.com/WeiqunZhang/amrex.git AMREX_BRANCH=new-feature python3
- -m pip install -v --force-reinstall .
+ cd z-checker-installer
- You can speed up the install further if you pre-install AMReX, e.g. with a
- package manager.
+ ./z-checker-install.sh
AMREX_INTERNAL=OFF and add installation prefix of AMReX to the
- environment variable CMAKE_PREFIX_PATH.
+
- Please see the short CMake tutorial that we linked above
- if this sounds new to you.
+ z-checker-install.sh will download latexmk, gnuplot, Z-checker, ZFP, and SZ + and install them one by one automatically, and then add the patches to let ZFP + and SZ fit for Z-checker.
+ +After installation, please download the two testing data sets, CESM-ATM and + MD-simulation (exaalt). The two data sets are available only for the purpose of + research of compression. Please ask for the data by contacting sdi1@anl.gov + if interested.
+ +LibpressioOPT is a library that is able to search for the appropriate error + bound setting based on user-sepcified metric values such as compression ratio + and PSNR. Z-checker itself has some simple built-in algorithms to do this work, + which may not be as accurate as LibpressioOPT. To this end, you also need to install + spack and use spack to install some preliminary libraries. For more details, please + read the z-checker-installer-instruction.pdf in the ./doc/ directory. If you don''t + need LibpressioOPT, you just need to run ''./z-checker-installer.sh'' to install + everything.
+ + - +Then, you are ready to conduct the compression checking. -
This work was supported by the Laboratory Directed Research and Development - Program of Lawrence Berkeley National Laboratory under U.S. Department of Energy - Contract No. DE-AC02-05CH11231.
+ You can generate compression results with SZ and ZFP using the following simple + steps: - + (Note: you have to run z-checker-install.sh to install the software before doing + the following tests) -pyAMReX Copyright (c) 2023, The Regents of the University of California, +
Configure the error bound setting and comparison cases in errBounds.cfg.
- Laboratory Alliance for Sustainable Energy, LLC and Lawrence Livermore +Create a new test-case, by executing "createZCCase.sh [test-case-name]". You + need to replace [test-case-name] by a meaningful name. -
If you have questions about your rights to use or distribute this software, + For example: - please contact Berkeley Lab''s Intellectual Property Office at + [user@localhost z-checker-installer] ./createZCCase.sh CESM-ATM-tylor-data
- IPO@lbl.gov. +Please see the full license agreement in LICENSE.
+
Perform the checking by running the command "runZCCase.sh": runZCCase.sh [data_type]
+ [error-bound-mode] [test-case-name] [data dir] [extension] [dimensions....].
- The SPDX license identifier is BSD-3-Clause-LBNL.
Then, you can find the report generated in z-checker-installer/Z-checker/[test-case-name]/report.
+ + + +Unlike the above one-command checking, the following steps present the generation + of compression results step by step.
+ +Go to zfp/utils/, and then execute "zfp-zc-ratedistortion.sh [data directory] + [dimension sizes....]". The compression results are stored in the compressionResults/ + directory. + + For example, suppose the directory of CESM-ATM data set is here: /home/shdi/CESM-testdata/1800x3600, + then the command is "zfp-zc-ratedistortion.sh /home/shdi/CESM-testdata/1800x3600 + 3600 1800". Note: the data files stored in the directory are also ending with + .dat and the dimension sizes are the same (1800x3600) in this test-case.
+ +Similarly, go to SZ/example/, and then generate compression results by SZ compressor + as follows: "sz-zc-ratedistortion.sh [data directory] [dimension sizes....]". + The compression results are stored in the compressionResults/ directory. + + As for the example CESM-ATM, the test command is "sz-zc-ratedistortion.sh /home/shdi/CESM-testdata/1800x3600 + 3600 1800".
+ +Then, go to Z-checker/examples/ directory, and run the command "./analyzeDataProperty.sh + [data directory] [dimension sizes....]" to generate the data properties based + on the data sets. This step has nothing to do with the compressors. The data analysis + results are stored in the dataProperties/ directory.
+ +Generate the figure files: run the command "./generateReport.sh" simply. The + results of comparing different compressors (such as sz and zfp in this test-case) + are stored in the directory called compareCompressors/.
+ +"createZCCase.sh [test-case-name]" allows you to create a new test-case. This + command will create a new workspace directory in Z-checker, SZ, and zfp respectively. + The compression results will be put in those workspace directories to avoid bing + messed with other test-cases.
+ +For example, if you run the generateReport.sh in the directory ./Z-checker/examples, + it is actually one test case, where the compression results and data analysis + results will be put in the dataProperty/ and compressionResults/ under it. + + For another test case with another set of data or application, you can create + a new workspace directory by the script createZCCase.sh (which calls ./Z-checker/createNewCase.sh).
+ + + +z-checker-update.sh can be used to update the repository (pull the new update + from the server), so that you don''t have to perform the update manually.
+ + + +Web installation allows to install a web server on the local machine, such + that you can visualize the data through a local webpage and other people can view + the data/results via that page if public ip is provided. + + z-checker-web-install.sh
+ + + +Remove sz_f (sz fast mode): + + $ manageCompressor -d sz -m fast -c manageCompressor-sz-f.cfg + + Remove sz_d (sz fast mode): + + $ manageCompressor -d sz -m deft -c manageCompressor-sz-d.cfg + + Remove zfp: + + $ manageCompressor -d zfp -c manageCompressor-zfp.cfg
+ + + +You can generate Z-checker report directly based on an HDF5 file. + + To this end, you need to install HDF5 library before hand, and then compile the + Z-checker/HDF5Reader as follows:
+ +You need to modify Makefile.linux2 by replacing "HDF5PATH = /home/sdi/Install/hdf5-1.10.1-install" + by your HDF5 installation path. + + Then: + + make -f Makefile.linux2
+ +You will find the executable ''testHDF5_CompDecomp'' generated on Z-checker/HDF5Reader/test/ + directory. + + You can use this command to read HDF5 file and generate analysis results.
+ +After that, you can use ''installHDF5Reader.sh'' and ''runZCCase_hdf5.sh'' + to generate the .pdf report. + + More details can be found in testHDF5/README.txt
+ + + +Go to the directory ADIOS2Header, and then do the following steps:
+ +Example: + + testAdios2 -i myVector_cpp.bp -n 2 -v bpFloats bpInts -o [target output directory]
+ +The generated binary data files will be put in the target output directory. + A meta file called ''varInfo.txt'' contains the extracted variables'' information + and it will be put in the target output directory as well. + + varInfo.txt and the binary files can be processed by runZCCase.sh
' - stargazers_count: 31 - subscribers_count: 16 - topics: - - amrex - - python - updated_at: 1716637941.0 -AMReX-Microelectronics/artemis: + stargazers_count: 5 + subscribers_count: 7 + topics: [] + updated_at: 1658448853.0 +CUP-ECS/beatnik: data_format: 2 - description: ARTEMIS (Adaptive mesh Refinement Time-domain ElectrodynaMIcs Solver) - couples the Maxwell's equations implementation in WarpX with classical equations - that describe quantum material behavior (such as, LLG equation for micromagnetics - and London equation for superconducting materials) for quantifying the performance - of next-generation microelectronics. + description: Initial Cabana/Cajita Low/High-order Z-model Interface Solver. Benchmark + for evaluating the performance of algorithms requiring global communication. Beatnik + is also a precursor to potential later a High Performance Parallel Interface solver. filenames: - - Docs/spack.yaml - full_name: AMReX-Microelectronics/artemis - latest_release: null - readme: "ARTEMIS (Adaptive mesh Refinement Time-domain ElectrodynaMIcs\ - \ Solver) is a high-performance coupled electrodynamics\u2013micromagnetics solver\ - \ for full physical modeling of signals in microelectronic circuitry. The overall\ - \ strategy couples a finite-difference time-domain (FDTD) approach for Maxwell\u2019\ - s equations to a magnetization model described by the Landau\u2013Lifshitz\u2013\ - Gilbert (LLG) equation. The algorithm is implemented in the Exascale Computing\ - \ Project (ECP) software framework, AMReX, which provides effective scalability\ - \ on manycore and GPU-based supercomputing architectures. Furthermore, the code\ - \ leverages ongoing developments of the Exascale Application Code, WarpX, which\ - \ is primarily being developed for plasma wakefield accelerator modeling. Our\ - \ temporal coupling scheme provides second-order accuracy in space and time by\ - \ combining the integration steps for the magnetic field and magnetization into\ - \ an iterative sub-step that includes a trapezoidal temporal discretization for\ - \ the magnetization. The performance of the algorithm is demonstrated by the excellent\ - \ scaling results on NERSC multicore and GPU systems, with a significant (59\xD7\ - ) speedup on the GPU using a node-by-node comparison. The utility of our code\ - \ is validated by performing simulations of transmission lines, rectangle electromagnetic\ - \ waveguides, magnetically tunable filters, on-chip coplanar waveguides and resonators,\ - \ magnon-photon coupling circuits, and so on.
\n\n\ -git\
- \ clone git@github.com:AMReX-Codes/amrex.git
git clone git@github.com:AMReX-Microelectronics/artemis.git
Make sure that the AMReX and Artemis are cloned in the same location in their\
- \ filesystem. Navigate to the Exec folder of Artemis and execute\nmake -j\
- \ 4.
\nYou can turn on and off the LLG equation by specifying USE_LLG\
- \ during compilation.
\nThe following command compiles Artemis without LLG\n\
- make -j 4 USE_LLG=FALSE
\nThe following command compiles Artemis\
- \ with LLG\nmake -j 4 USE_LLG=TRUE
\nThe default value of USE_LLG\
- \ is TRUE.
Example input scripts are\
- \ located in Examples directory.
You can run the following to simulate an air-filled X-band\ - \ rectangle waveguide:
\n\nmake\
- \ -j 4 USE_LLG=FALSE
\nmpirun -n 4 ./main3d.gnu.TPROF.MTMPI.OMP.GPUCLOCK.ex\
- \ Examples/Waveguide/inputs_3d_empty_X_band
make -j 4 USE_LLG=FALSE USE_GPU=TRUE
\nmpirun -n 4 ./main3d.gnu.TPROF.MTMPI.CUDA.GPUCLOCK.ex\
- \ Examples/Waveguide/inputs_3d_empty_X_band
You can run the following to simulate an X-band magnetically\ - \ tunable filter:
\n\nmake\
- \ -j 4 USE_LLG=TRUE
\nmpirun -n 8 ./main3d.gnu.TPROF.MTMPI.OMP.GPUCLOCK.ex\
- \ Examples/Waveguide/inputs_3d_LLG_filter
make -j 4 USE_LLG=TRUE USE_GPU=TRUE
\nmpirun -n 8 ./main3d.gnu.TPROF.MTMPI.CUDA.GPUCLOCK.ex\
- \ Examples/Waveguide/inputs_3d_LLG_filter
Refer to the following link for several\ - \ visualization tools that can be used for AMReX plotfiles.
\n\n\nYou can extract the data in numpy\
- \ array format using yt (you can refer to this for installation and usage of
import yt\nds = yt.load('./plt00001000/')\
- \ # for data at time step 1000\nad0 = ds.covering_grid(level=0, left_edge=ds.domain_left_edge,\
- \ dims=ds.domain_dimensions)\nE_array = ad0['Ex'].to_ndarray()\nshared spack configs repo
- - ' - stargazers_count: 0 - subscribers_count: 4 - topics: [] - updated_at: 1639176281.0 -C2SM/spack-c2sm: - data_format: 2 - description: Repository for c2sm spack config and repo files - filenames: - - upstreams/daint/icon-dsl/spack.yaml - full_name: C2SM/spack-c2sm - latest_release: v0.20.1.9 - readme: 'Spack is the package manager used by C2SM and MeteoSwiss to install and deploy - software on supercomputers, local machines and the cloud.
+Beatnik is a benchmark for global communication based on Pandya and Shkoller''s + 3D fluid interace "Z-Model" in the Cabana/Cajita mesh framework [1]. The goals -
+ of Beatnik are to: -Infos about c2sm-supported software and machines
+Beatnik uses a simple mesh-based representation of the surface manifold as + a Cabana grid 2D mesh in I/J space and a regular block 2D decomposition of this + manifold. The physical position of each element in the mesh is stored as a separate + vector in the nodes of the mesh. This design results in simple and efficient computation + and communication strategies for surface normals, artificial viscosity, and Fourier + transforms elements. However, it complicates methods where the data decomposition + and communication is based on the spatial location of manifold points, requiring + them to either maintain a separate spatial decomposition of the surface or to + continually construct a spatial decomposition. A surface mesh that decomposed + the mesh by spatial location would be an interesting alternative but would have + the opposite issue - communication for surface calculations would be more complex + but the (expensive) far force methods that rely on spatial decompositions (e.g. + distance sort and spatial tree methods like the fast multi-pole method) would + be less expensive.
- + -Beatnik relies on multiple external packages to build, including:
-To ease building Beatnik, the configs/ directory includes Spack configuration + files for building in spack environments on multiple systems and test case run + scripts for a variety of systems. In addition, the latest version of Spack includes + a package description for directly building Beatnik. More information on building + Beatnik can be found in the README.md file in the configs/ directory.
-By default, Beatnik solves a simple multi-mode rocket rig problem sized for + a single serial CPU core with approximately 4GB of memory. It also includes command + line options to change initial problem state, I/O frequency, and to weak-scale + scale up the initial problem to larger number of processes. It also includes problem-specific + command line parameters; setting these parameters accurately generally requires + expertise in fluid interface models. However, we provide several useful examples + drawn from the ZModel papers that recreate the results in those papers.
--x [cuda|threads|serial] - The node-level parallelism/accelerator
+ backend to use-F [write-frequency] - Interval between timesteps when I/O is written
-O [solution order] - Order of solver to use (''high'', ''medium'',
+ or ''low''). ''low'' is the default.-n [i/j mesh dimension ] - Number of points on the interface manifold
+ in the I and J dimensions-I [interface initialization] - Function to use for interface initial
+ condition. Currently only ''cos'' and ''sech2'' are supported.
-m [magnitude] - The maximum magnitude of the initialization function.
spack-c2sm - v0.18.1.3 [deprecated]
- --p [period] - The number of periods of the interface in the initial
+ bounding box
spack-c2sm - v0.18.1.2 [deprecated]
- --a [atwood] - Atwood''s constant for the difference in pressure between
+ the two fluids
spack-c2sm - v0.18.1.1 [deprecated]
+-g [gravity] - Gravitational acceleration in the -Z direction-a [atwood] - Atwood''s constant for the difference in pressure
+ between the two fluids
- General infos about spack
+-M [mu] - Mu, the artificial viscosity constant used in the Z-Model-e [epsilon] - Epsilon, the desingularization constant used in the
+ Z-Model expressed as a fraction of the distance between interface mesh points
-
+
- With spack v0.18 we suggest local/individual spack instances and the use of - spack environments.
+The simplest test case and the one to which the rocketrig example program defaults + is an initial interface distributed according to a cosine function. Simple usage + examples:
-A user clones the spack repo
+git clone --depth 1 --recurse-submodules
- --shallow-submodules -b v0.20.1.5 https://github.com/C2SM/spack-c2sm.gitbin/rocketrig -x serial
- gets spack in the command line
+.
- spack-c2sm/setup-env.shbin/rocketrig
+ -x cuda -n 512
- activates an environment
+spack env activate <path_to_env>and starts exploring
+Another test case is a single-mode rollup test where the intitial interface + is set according to a hyperbolic secant function. This testcase recreates the + the Gaussian perturbation results in Panda and Shkoller''s paper from sections + 2.3 and 2.4. To run this testcase with a high-order model, use the following + command line parameters. Note that this works best with a GPU accelerator, as + the exact high-order far field force solver is very compute intensive and is generally + impractical for non-trivial mesh sizes without GPU acceleration: -
and building
+bin/rocketrig -x cuda -O high -n 64 -I sech2 -m 0.1 -p 9.0 -b free -a 0.15
+ -M 2 -e 2
-
+ Beatnik is being implemented in multiple distinct steps, with associated planned + releases:
-a package.
+Updating spack-c2sm is in the hands of the user.
+git pull
+ Version 1.0 Features
- git submodule update --recursiveAfter an update we advice to clean
+spack uninstall -a
+ and rebuild.
+| + | Command | +
|---|---|
| Clone | +git clone --depth 1 --recurse-submodules --shallow-submodules -b <branch/tag>
- https://github.com/C2SM/spack-c2sm.git |
+
| Load | +
+ Potential later (e.g. >=2.0) features -. spack-c2sm/setup-env.sh autodetects machine or . spack-c2sm/setup-env.sh
- <machine> forces machineor . spack-c2sm/setup-env.sh
- unknown uses blank configspack compiler find autodetects compilersspack external find --all
- autodetects externally installed packages
+
|
+
| Update | + -
+
- git pullgit submodule update --recursive
+
- |
+
| Clean | +
+ spack uninstall -a uninstalls all packagesspack clean -a cleans all misc cachesrm -rf ~/.spack removes
- user scope data |
+
The general structure of Beatnik and the rocketrig examples were taken from + the ExaMPM proxy application (https://github.com/ECP-copa/ExaMPM) + developed by the ECP Center for Particle Applications (CoPA), which was also available + under a 3-Clause BSD License when used for creating application structure.
-Spec syntax: <package>@<version>%<compiler>+<variant> ~<variant>^<sub-package> +<sub-package-variant><compiler flags>
| + - | Command | +
|---|---|
| Find | + -
+
- spack find lists all installed packages. spack find <spec>
- lists all installed packages that match the spec. |
+ '
+ stargazers_count: 4
+ subscribers_count: 4
+ topics: []
+ updated_at: 1716587778.0
+E4S-Project/e4s:
+ data_format: 2
+ description: E4S for Spack
+ filenames:
+ - environments/23.08/cuda-aarch64/spack.yaml
+ - environments/22.05/rocm.spack.yaml
+ - environments/23.08/oneapi-x86_64/spack.yaml
+ - environments/24.02/amd64-gcc-cpu-ubuntu20.04/spack.yaml
+ - environments/24.05/ppc64-gcc-cuda-ubuntu20.04/spack.yaml
+ - environments/23.05/cuda-x86_64/spack.yaml
+ - environments/24.05/amd64-oneapi-ubuntu22.04/spack.yaml
+ - environments/24.05/arm64-gcc-cuda-ubuntu22.04/spack.yaml
+ - environments/23.08/cuda-x86_64/spack.yaml
+ - environments/24.05/amd64-gcc-cuda-ubuntu22.04/spack.yaml
+ - environments/22.05/oneapi.spack.yaml
+ - environments/21.08/spack.yaml
+ - environments/23.02/cuda-ppc64le/spack.yaml
+ - environments/24.05/amd64-gcc-cpu-ubuntu22.04/spack.yaml
+ - environments/24.05/arm64-gcc-cpu-ubuntu22.04/spack.yaml
+ full_name: E4S-Project/e4s
+ latest_release: v24.05
+ readme: "
| Info | +spack info <package> |
+
| Spec | +
+ 
- spack spec <spec> concretizes abstract spec (unspecfied variant
- = any)Spack is not required to use the default of an - unspecified variant. The default value is only a tiebreaker for the concretizer. + 
- |
+
| Install | +spack install <spec> |
+
| Locate | + It supports many features including Perfectly-Matched Layers (PML), mesh refinement, + and the boosted-frame technique. -
+ WarpX is a highly-parallel and highly-optimized code, which can run
+ on GPUs and multi-core CPUs, and includes load balancing capabilities.
- |
+ WarpX scales to the world''s largest supercomputers and was awarded the 2022 ACM Gordon Bell Prize.
-
| Load env | +
+ 
- spack load <spec> loads run environment |
+
| Activate - env | + -spack env activate <env_name> |
+
| Deactivate - env | +spack deactivate |
+
Our workflow is described in CONTRIBUTING.rst.
+ + + +WarpX Copyright (c) 2018, The Regents of the University of California, + + through Lawrence Berkeley National Laboratory (subject to receipt of any + + required approvals from the U.S. Dept. of Energy). All rights reserved.
+ +If you have questions about your rights to use or distribute this software, + + please contact Berkeley Lab''s Innovation & Partnerships Office at + + IPO@lbl.gov.
+ +Please see the full license agreement in LICENSE.txt.
+
+ Please see the notices in NOTICE.txt.
+
+ The SPDX license identifier is BSD-3-Clause-LBNL.
-
chipStar (formerly CHIP-SPV)
+
- is software that allows software written to use the
+
- (HIP)
+ -10.5281/zenodo.6954922-blue.svg){kind=icon}
- SPIR-V intermediate
- language.
+ 
ImpactX: an s-based beam dynamics code including space charge effects. -
This repository contains support for building chipStar and its + This is the next generation of the IMPACT-Z + code.
- dependencies via the Spack package + - manager. +In order to learn how to install and run the code, please see the online documentation: -
Note: most development to date has been done with the Level Zero + https://impactx.readthedocs.io
- environment, and it is expected that substantial work is needed for +![\"E4S\"](\"https://github.com/E4S-Project/e4s/blob/master/logos/E4S-dark-green.png\"\)
![\"License\"](\"https://img.shields.io/github/license/E4S-Project/e4s\"\)
![\"GitHub](\"https://img.shields.io/github/issues/E4S-Project/e4s.svg\"\)
![\"GitHub](\"https://img.shields.io/github/issues-pr/E4S-Project/e4s\"\)
-10.5281/zenodo.4571577-blue.svg){kind=icon}
-10.1109/SC41404.2022.00008-blue.svg){kind=icon}
-10.18429%2FJACoW--NAPAC2022--TUYE2-blue.svg){kind=icon}
llvm@16.0.2 and then using spack compiler add with the
- llvm
+ Our workflow is described in CONTRIBUTING.rst.
- package''s install location), because thechipstar package defined
+
- in this repository depends on the llvm package anyway.
+ Please see our developer installation section of the documentation for an + easy install of our software dependencies.
-Before you start, you will need a copy of the ImpactX source code:
- of doing this is by installing the Spackintel-oneapi-compilers
+ git clone git@github.com:ECP-WarpX/impactx.git
- package, then registering the location of its compilers with Spack.
+ cd impactx#
+ find dependencies & configure
- $ spack install intel-oneapi-compilers@2023
+ cmake -S . -B build
- $ spack compiler add $(spack location
- -i intel-oneapi-compilers@2023)/compiler/latest/linux
-
+ # compile
-
+ cmake --build build -j 4
That''s all!
-
+ ImpactX binaries are now in build/bin/.
-
$ git clone https://github.com/CHIP-SPV/CHIP-SPV-SpackYou can inspect and modify build options after running cmake -S . -B
+ build with either
ccmake buildor by adding arguments with -D<OPTION>=<VALUE> to
+ the first CMake call, e.g.:
cmake -S .
+ -B build -DImpactX_COMPUTE=CUDA -DImpactX_MPI=OFF#
+ find dependencies & configure
-
+ cmake -S . -B build -DImpactX_PYTHON=ON
- - Concretize the active environment. (In Spack terminology,
- "to concretize" means to let Spack examine the package specifications
+ # compile & install
- it has been asked to build, plus the available package repositories,
+ cmake --build build -j 4 --target pip_install
An executable ImpactX binary with the current compile-time options encoded
+ in its file name will be created in build/bin/.
Additionally, a symbolic link named impactx can be found in that
+ directory, which points to the last built ImpactX executable.
$ spack concretize -f -UThe command-line syntax for this executable is:
-We suggest examining the output from running the spack concretize
+
Usage:
+ impactx <inputs-file> [some.overwritten.option=value]...
- command to make sure that Spack''s concretizer has truly decided to
- use the configuration options and especially the compilers that you
+ Mandatory arguments (remove the <>):
- want it to use. Note that the environment and related configuration
+ inputs-file the path to an input file; can be relative
+ to the current
- are purposefully not overly constrained to use the given compiler
+ working directory or absolute.
- for every dependency package, so even though there are some packages
+ Example: input_fodo.in
- that must be built with %clang, there are others that may be
- built (or re-used from already-installed packages) using %gcc such
+ Optional arguments (remove the []):
- as the system''s GCC installation.
+ options this can overwrite any line in an inputs-file
- If Spack''s concretizer didn''t do what you want, you can re-concretize
+ Example: quad1.ds=0.5 sbend1.rc=1.5
- the environment and be more explicit about what you want using command-line
- configuration options (recommended) or by editing the environment''s
+ Examples:
- spack.yaml file or other configuration options that your Spack installation
+ In the current working directory, there is a file "input_fodo.in"
+ and the
- is using. (Use spack config blame to see which configuration files
- Spack is
+ "impactx" executable.
- using.) For instance, if you have both clang@16.0.2 and clang@15.0.7
+ The line to execute would look like this:
- installed and registered as Spack compilers, and you want to build
+ ./impactx input_fodo.in
- using clang@15.0.7, you may have to use a concretize command like
- the
- following:
+ In the current working directory, there is a file "input_fodo.in"
+ and the
- $ spack -c "packages:chipstar:require:''%clang@15.0.7''"
- concretize -f -U
+ executable "impactx" is in a directory that is listed in
+ the "PATH"
- As before, verify from the output of the spack concretize command
- that it
+ environment variable.
- is using the compiler version you want, clang@15.0.7 in this example.
+ The line to execute would look like this:
-
+ impactx input_fodo.in
- - Build the environment.
-
+ In the current working directory, there is a file "input_fodo.in"
+ and the
- $ spack install
+ "impactx" executable. We want to voerwrite the segment length
+ of the beamline
- Spack supports some options for controlling the build and installation,
+ element "quad1" that is already defined in it. We also want
+ to change the
- such as -j to limit the number of processes used for parallel builds,
+ radius of curvature of the bending magnet "sbend1" to a
+ different value than
- useful for being a good citizen on shared systems by not allowing Spack
+ in the file "input_fodo.in".
- to use all available cores (its default). See the Spack documentation for
+ The line to execute would look like this:
- more information.
+ ./impactx input_fodo.in quad1.ds=0.5 sbend1.rc=1.5Assuming all goes well with the build and install, a spack find
+
In order to run our tests, you need to have a few Python packages installed:
-python3
+ -m pip install --upgrade pip
- You can run all our tests with:
-ctest
+ --test-dir build --output-on-failurespack load chipstarFurther options:
-ctest --test-dir build --help
- and module files have been generated for the software you built
+ module load chipstar
+ ctest --test-dir build -N
Note that you may need to modify your environment to be able to run +
ctest --test-dir build
+ -R FODO
- programs produced using chipStar and the H4I libraries built
+ intel_compute_runtime module before being
+
- able to run programs that use the Intel Level Zero runtime.
+ This work was supported by the Laboratory Directed Research and Development + Program of Lawrence Berkeley National Laboratory under U.S. Department of Energy + Contract No. DE-AC02-05CH11231.
- +ImpactX is supported by the CAMPA collaboration, a project of the U.S. Department + of Energy, Office of Science, Office of Advanced Scientific Computing Research + and Office of High Energy Physics, Scientific Discovery through Advanced Computing + (SciDAC) program.
-ImpactX Copyright (c) 2022, The Regents of the University of California, through + Lawrence Berkeley National Laboratory (subject to receipt of any required approvals + from the U.S. Dept. of Energy). -
If you have questions about your rights to use or distribute this software, + please contact Berkeley Lab''s Intellectual Property Office at IPO@lbl.gov.
-Please see the full license agreement in LICENSE.txt.
- module command.
BSD-3-Clause-LBNL.
'
- stargazers_count: 0
- subscribers_count: 5
- topics: []
- updated_at: 1687550793.0
-COSIMA/spack-config:
+ stargazers_count: 24
+ subscribers_count: 6
+ topics:
+ - simulation
+ - beam-dynamics
+ - particle-in-cell
+ - gpu
+ - physics
+ - pic
+ - particle
+ - accelerator
+ - research
+ updated_at: 1716438419.0
+Exawind/exawind-builder:
data_format: 2
- description: Spack configuration installed at /g/data/ik11/spack/ to provide ACCESS-OM3
- dependencies
+ description: Scripts to help building Exawind codes on various systems
filenames:
- - environments/cesm-0_2_0/spack.yaml
- - environments/access-om3-0_x_0/spack.yaml
- - environments/cesm-0_x_0/spack.yaml
- full_name: COSIMA/spack-config
- latest_release: access-om3-v0.2.0
- readme: "\n\
- This repository contains the spack configuration and the spack environments\ - \ used\nby COSIMA to deploy software on gadi.
\n\nClone this repository and its submodules to some appropriate\
- \ location (e.g.,\n/g/data/ik11/spack/0.21.2):
$ git clone\
- \ --recursive https://github.com/COSIMA/spack-config.git /g/data/ik11/spack/0.21.2Next, create the python virtual environment:
\n$ cd /g/data/ik11/spack/0.21.2\n\
- $ ./bootstrap_venv.shFinally, to\ - \ use this spack installation one just needs to activate the python\nenvironment:
\n\ -$ . /g/data/ik11/spack/0.21.2/venv/bin/activate\n\
- $ which spack\nspack ()\n\
- { \n : this is a shell function\
- \ from: /g/data/ik11/spack/0.21.2/spack/share/spack/setup-env.sh;\n : the real spack script is here: /g/data/ik11/spack/0.21.2/spack/bin/spack;\n\
- _spack_shell_wrapper \"$@\";\n return $?\n} Most\
- \ of the paths in the Spack configurations files in this repository are\nrelative\
- \ to the Spack location (e.g.,\n/g/data/ik11/spack/0.21.2/spack).\
- \ Unfortunately this is not the case for the\nglobal source's mirror defined in\
- \ config/system/mirror.yaml, so the URL in\nthis file needs to be\
- \ updated if usage of a mirror is planned (see next\nsection).
At this\
- \ point it might be worth reviewing the compilers and packages defined in\nconfig/system/compilers.yaml\
- \ and config/system/packages.yaml and update them\nif necessary.
It is recommended that\ - \ all software be installed using spack\nenvironments. Currently the following\ - \ environments are provided (the names\nshould be self-explanatory):
\naccess-om3-0_1_0access-om3-0_2_0access-om3-0_x_0cesm-0_1_0common_tools_and_librariesInstallation of a spack environment is usually quite straightforward,\ - \ but\nbecause this can be a CPU intensive operation and take quite some time,\ - \ it is\nbest to do this in parallel and to use an interactive job.
\nFirst activate the spack environment
\n\ -$ spack env activate <env>where <env> by the actual name of the environment.
As the compute nodes\ - \ do not have internet access, one needs to download all the\nnecessary sources\ - \ from the login node. This is done using a spack mirror.
\n$ spack mirror create\
- \ -d sources -aHere sources is the name of\
- \ a mirror that has already been configured.
Note that the spack environment\ - \ must have been concretized before creating the\nmirror, otherwise spack will\ - \ not know which files need downloading. To\nconcretize an environment, one uses\ - \ the following command:
\n$ spack concretizeThis should not use more than\
- \ a single node. Also, make sure to add gdata/ik11\nand scratch/ik11\
- \ to the storage options.
Once the job has started, because it starts a completely new shell session,\ - \ one\nneeds to activate again both the python and the spack environments:
\n\ -$ . /g/data/ik11/spack/0.21.2/venv/bin/activate\n\
- $ spack env activate <env> Then one can simply do
\n\ -$ spack install In this case, although each individual\ - \ build will use some level of parallelism,\nspack will proceed through the installation\ - \ of the packages sequentially. To\nfully use parallelism one needs to tell spack\ - \ to create a Makefile and use this\nto install the software:
\n$ spack env\ - \ depfile -o Makefile\n$ make -j
In the end, all the packages should be available in some subdirectory of\n\
- /g/data/ik11/spack/0.21.2/opt/ and the corresponding environment\
- \ modules are\ninstalled under a subdirectory of /g/data/ik11/spack/0.21.2/modules.\
- \ The\nactual subdirectories depend on the selected environment.
The way this repository is set up assumes that no update of spack will take\n\ - place within a given instance. This is to make sure all packages available\nwithin\ - \ a given instance are installed with the same version of spack and are\nfully\ - \ reproducible. This means that, in order to update spack, one needs to\ncreate\ - \ a new instance. Most of the process is therefore very similar to what is\ndescribed\ - \ in the Installation instructions\n\ - section, with a few modifications.
\nStart by cloning this repository and\
- \ its submodules to some appropriate location\n(e.g., /g/data/ik11/spack/0.21.2):
$ git clone --recursive https://github.com/COSIMA/spack-config.git /g/data/ik11/spack/0.21.2Next, create and checkout a new branch in the repository, using the spack\n\ - version as branch name:
\n$ cd /g/data/ik11/spack/0.21.2\n\
- $ git checkout -b 0.21.2This is\ - \ not mandatory, but it helps keeping all the different configurations and\nenvironments\ - \ clearly separated. It also allows to update instances that are\nusing different\ - \ versions of spack independently.
\nThen we proceed with updating the spack\ - \ submodule:
\n$\
- \ cd spack\n$ git fetch --tags\n$ git checkout\
- \ v0.21.2\n$ cd ..\n\
- $ git commit spack -m \"Update spack to 0.21.2 tag.\" Note that the git fetch step is only strictly necessary if you\
- \ want to use a\ntag from the spack repository (which is recommended).
Finally,\ - \ just follow the remaining instructions in the Installation\ninstructions section to create the python virtual\nenvironment.
\n\ -This section describes some of the design decisions and\ - \ their rationale.
\n\nIn this repository, spack is configured to install the\ - \ packages to a path of the\nfollowing form:
\nopt/{architecture}/{compiler.name}-{compiler.version}/{name}-{version}-{hash:7}
By adding the architecture, compiler name, and compiler version to the path,\n\ - installations of packages built with different compilers on different\narchitectures\ - \ are allowed to coexist without any clashes. The inclusion of part\nof the spack\ - \ hash also allows for the same version of a given package to be\ncompiled with\ - \ different options. Note that there is nothing specific to the\nspack environments,\ - \ as this is not necessary (two environments that require\nexactly the same package\ - \ built in exactly the same way will share the\ncorresponding installation).
\n\ -By default, the TCL environment\ - \ modules use the following naming scheme:
\n`{name}/{version}-{compiler.name}-{compiler.version}-{hash:7}``
\n\ -This ensures that there are no clashes, but unfortunately the inclusion of\
- \ the\nspack hash makes the use of the modules a bit cumbersome. To avoid the\
- \ use of\nthe hash, one can set the modules installation path and naming scheme\
- \ at the\nlevel of the spack environments. The existing environments include the\
- \ following\ndefinitions in their spack.yaml file:
![\"Build](\"https://github.com/LLNL/UnifyFS/actions/workflows/build-and-test.yml/badge.svg?branch=dev\"\){kind=icon}
![\"Exisiting](\"./host/vscode_remote_jupyter-server.png\"\)
![\"project](\"https://img.shields.io/badge/zulip-join_chat-brightgreen.svg\"\){kind=icon}
![\"SST\"](\"http://sst-simulator.org/img/sst-logo-small.png\"\){kind=logo}