v24.08

AMReX: release 24.08
openPMD-api: release 0.15.2

What's Changed

• Build back on AMReX development after release 24.06 by @MaxThevenet in #1124
• Add a relative tolerance to SALAME by @AlexanderSinn in #1125
• Add inputs for scaling benchmarks to examples by @AlexanderSinn in #1128
• CI: CCache for macOS by @ax3l in #1129
• Read laser profile from parser by @huixingjian in #1122
• Clean in-situ diagnostic with TupleSplit and tupleToArray by @AlexanderSinn in #1126
• Remove amrex::Gpu::notInLaunchRegion() checks by @AlexanderSinn in #1127
• Fix compilation without openpmd by @AlexanderSinn in #1132
• Add fix for slow GPU aware MPI to JSC booster script by @AlexanderSinn in #1133
• Add pre_register_memory option by @AlexanderSinn in #1134
• Fix unused parameter warning by @AlexanderSinn in #1138
• Remove particles with negative psi_inv by @AlexanderSinn in #1142
• HPMG: Compute finest level directly on the field by @AlexanderSinn in #1137
• Update release number and AMReX build version by @MaxThevenet in #1145

Full Changelog: v24.06...v24.08

v24.06

AMReX: release 24.06
openPMD-api: release 0.15.2

What's Changed

• [mini] Revert amrex branch to build hipace-dev on amrex-dev by @MaxThevenet in #1078
• Improve Maxwell documentation to setup and run by @MaxThevenet in #1076
• Add option to use HPMG to solve Poisson equations by @AlexanderSinn in #1063
• Fix MPI buffer max_leading_slices by @AlexanderSinn in #1079
• Add Beam Spin tracking by @AlexanderSinn in #1071
• Put input parameters for MPI in a separate category by @AlexanderSinn in #1080
• Reverse order of MPI ranks by @AlexanderSinn in #1081
• [mini] Add doc entry for use_previous_rng by @MaxThevenet in #1084
• Make CPU-GPU memory copy highly asynchronous by @AlexanderSinn in #1082
• Add in-situ diagnostic for spin by @AlexanderSinn in #1086
• Limit size in GiB of communications buffer by @AlexanderSinn in #1085
• NVCC 11.5 fixes by @AlexanderSinn in #1089
• Use 63 bits for Beam ID by @AlexanderSinn in #1090
• correct the direction of the spatial rotation by @huixingjian in #1091
• Put Laser abs(a) in separate Field by @AlexanderSinn in #1088
• Clean-up TinyProfiler output by @AlexanderSinn in #1087
• Fix Spin in-situ diagnostic for NVCC 11.5 by @AlexanderSinn in #1094
• Simplify tiling in plasma deposition by @AlexanderSinn in #1093
• Fix Plasma tiling in SALAME by @AlexanderSinn in #1096
• Fix SALAME overloaded beam removal by @AlexanderSinn in #1097
• Don't use Laser FFT array in Laser MG solver by @AlexanderSinn in #1095
• Fix index bug in FFTPoissonSolverPeriodic by @AlexanderSinn in #1102
• Fix adaptive time step for non-electron beams by @AlexanderSinn in #1092
• Enable initial pulse front tilt set for lasers by @huixingjian in #1100
• Remove outer_depos_loop by @AlexanderSinn in #1101
• Use correct geometry for Laser diagnostics by @AlexanderSinn in #1103
• Fix ReorderParticles for Beam by @AlexanderSinn in #1104
• Sort plasma after initialization when running on CPU by @AlexanderSinn in #1107
• Refactor AnyFFT by @AlexanderSinn in #1099
• CI: Update ROCm/HIP by @ax3l in #1113
• Doc: Fix Legal References by @ax3l in #1115
• CMake: No Deprecation Warnings by @ax3l in #1118
• Separate grid for Laser by @AlexanderSinn in #1106
• Fix tiling without OMP by @AlexanderSinn in #1108
• Amend Notices for Non-Endorsement by @ax3l in #1114
• Initialize MR level with beam from file by @AlexanderSinn in #1117
• Use less than or equal for min_density with beam by @AlexanderSinn in #1119
• Add warning if beam is injected with zero particles by @AlexanderSinn in #1120
• Fix openPMD crash if beam has no particles by @AlexanderSinn in #1121
• Improve performance of FFTDirichletExpanded by @AlexanderSinn in #1111
• Improve performance of FFTDirichletFast by @AlexanderSinn in #1112
• Prepare for release 24.06 by @MaxThevenet in #1123

Full Changelog: v24.03...v24.06

v24.03

AMReX: release 24.03
openPMD-api: release 0.15.2

What's Changed

• Hipace dev builds on amrex dev by @MaxThevenet in #1035
• Doc: Button to Copy Code Blocks by @ax3l in #1039
• [mini] Fix typo in laser pulse length to pulse duration conversion by @MaxThevenet in #1036
• Fix typo in assert by @AlexanderSinn in #1041
• Beam particle sorting by @AlexanderSinn in #1032
• Add option to symmetrize fields by @AlexanderSinn in #1038
• Laser FFT solver for AMD by @AlexanderSinn in #1043
• Set amrex.omp_threads = "nosmt" by @ax3l in #1044
• Add in-situ diagnostics for field energy by @AlexanderSinn in #1030
• Use 64 bit idcpu by @AlexanderSinn in #1047
• CI: Fix ROCm 6.0 hiprand by @ax3l in #1048
• Fix minor formatting issue in documentation by @AlexanderSinn in #1050
• Make plasma ppc a mandatory input parameter by @AlexanderSinn in #1051
• MacOS CI explicitly install python@3.12 by @AlexanderSinn in #1049
• Laser with non-electron plasma by @AlexanderSinn in #1002
• Make laser output LASY compliant by @AlexanderSinn in #1046
• Add zyx geometry for laser input by @AlexanderSinn in #1053
• Clean-up in-situ diagnostic by @AlexanderSinn in #1052
• DepositCurrent: atomicAdd -> lockAdd by @WeiqunZhang in #1059
• Use AMReX version of TypeMultiplier by @AlexanderSinn in #1061
• Add HiPACE_CCACHE Option by @ax3l in #1065
• Add [xuy], [yux], [ux/uz], [uy/uz] to beam in-situ diagnostics by @AlexanderSinn in #1058
• Initial propagation angle of laser enabled by @huixingjian in #1057
• Add laser insitu diagnostics by @AlexanderSinn in #1060
• Use id().is_valid() by @AlexanderSinn in #1066
• Use amre::removeInvalidParticles for the Beam by @AlexanderSinn in #1067
• Fix hpmg for some non-ideal resolutions by @AlexanderSinn in #1062
• Try fix MacOS CI by unlinking Python by @AlexanderSinn in #1073
• Prepare Beam Communication for Runtime Components by @AlexanderSinn in #1068
• Add support for Spin to openPMD Beam output by @AlexanderSinn in #1069
• Use amre::GpuComplex in diagnostics by @AlexanderSinn in #1070
• [mini] Option to specify laser CEP by @MaxThevenet in #1074
• [mini] Fix typos in executable names in doc by @MaxThevenet in #1075
• Prepare for release 24.03 by @MaxThevenet in #1077

New Contributors

• @huixingjian made their first contribution in #1057

Full Changelog: v23.11...v24.03

v23.11

AMReX: release 23.11
openPMD-api: release 0.15.2

WARNING: this release builds on AMReX 23.11, but this was not heavily tested. In case of unexpected problems, this could be a good place to check.

What's Changed

• [mini] build on latest openpmd-api 0.15.2 by @MaxThevenet in #1023
• Add xy_integrated diagnostic type by @AlexanderSinn in #1020
• Initialize fixed weight beam from PDF by @AlexanderSinn in #1024
• Add interpolate_neutralizing_background option by @AlexanderSinn in #1021
• Add option to limit number of buffered beam slices by @AlexanderSinn in #1026
• Beam external field from parser by @AlexanderSinn in #1028
• Fix External AMReX Builds by @ax3l in #1031
• Update amrex version when checking for pre-installed version by @MaxThevenet in #1033
• Update version numbers before release by @MaxThevenet in #1034

Full Changelog: v23.10...v23.11

v23.10

Dependencies

AMReX: release 23.10
openPMD-api: release 0.15.1

What's Changed

• build on amrex dev again by @MaxThevenet in #996
• Add output_input option by @AlexanderSinn in #997
• Per-slice communication by @AlexanderSinn in #991
• Include SALAME beam profile in first output iteration by @AlexanderSinn in #1001
• fix B field conversion in radiation reactions in normalized units by @SeverinDiederichs in #998
• Communicate lambda0 with laser from file by @AlexanderSinn in #1004
• Add ghost cells to laser arrays for 2^n-1 MG solve by @AlexanderSinn in #1005
• set MG_tolerance_abs to numeric_limits min by @AlexanderSinn in #1007
• Add deposit_rho_individual option by @AlexanderSinn in #1003
• Per-slice fixed_ppc beam initialization by @AlexanderSinn in #1000
• [mini] Add radius to fixed weight beams by @SeverinDiederichs in #1006
• add plasma subcycling by @SeverinDiederichs in #1009
• Per-slice fixed_weight beam initialization by @AlexanderSinn in #1008
• Fix deposit_rho_individual with neutralizing background and SALAME by @AlexanderSinn in #1010
• CMake: Verify HiPACE_..._src paths exit by @ax3l in #1012
• push beam particles on correct slice when subcycling by @AlexanderSinn in #1013
• add weight for plasma insitu diags and radius by @SeverinDiederichs in #1011
• [mini] update LUMI docs by @SeverinDiederichs in #1017
• Plasma initialization for MR by @AlexanderSinn in #1018
• Workaround std::gcd and std::lcm constexpr issue by @AlexanderSinn in #1019
• Add simple string input parameter parsing by @AlexanderSinn in #990
• Faster multigrid solve using shared memory by @AlexanderSinn in #1015
• Prepare for Release 23.10: proper release number and amrex version by @MaxThevenet in #1022

Full Changelog: v23.07...v23.10

v23.07

Dependencies

AMReX: release 23.07
openPMD-api: release 0.15.1

What's Changed

• Replace rho field with rho - jz/c by @AlexanderSinn in #940
• CI: apt retry and new Easyinstall link by @AlexanderSinn in #930
• make CI triggers consistent by @AlexanderSinn in #938
• Fix rocfft 5.5 by @SeverinDiederichs in #942
• Option to symmetrize plasma by @SeverinDiederichs in #943
• [mini] Document manip to build HiPACE++ release on corresponding AMReX release by @MaxThevenet in #944
• Enable 3 MR levels by @AlexanderSinn in #945
• replace static with inline static for member variables by @AlexanderSinn in #946
• Use less than or equal for min_density by @AlexanderSinn in #947
• Don't use tiling in the plasma pusher by @AlexanderSinn in #949
• [mini] Read angularFrequency from mesh instead of from iteration by @MaxThevenet in #955
• [mini] Fix plasma do_symmetrize by @MaxThevenet in #957
• [mini] Fix b0 calculation in collisions by @SeverinDiederichs in #958
• Implement OMP threading in missing functions by @AlexanderSinn in #951
• [mini] Fix formatting for HIPACE_PROFILE output by @MaxThevenet in #956
• Gather Ez in loop over slices to gather from proper array by @MaxThevenet in #948
• [mini] update documentation for Maxwell by @SeverinDiederichs in #960
• [mini] Fix adaptive time step by @MaxThevenet in #959
• Laser file reader can read from arbitrary box and domain by @MaxThevenet in #950
• [mini] Remove unused variables by @MaxThevenet in #961
• Reset plasma id at every call of Init by @MaxThevenet in #963
• Update Beam and Plasma to Pure SoA by @AlexanderSinn in #928
• Laser reader in cylindrical coordinates by @MaxThevenet in #953
• Fix collisions in normalized units by @SeverinDiederichs in #954
• Enable ionization in normalized units by @SeverinDiederichs in #962
• Add [z], [z^2], [uz], [uz^2] and [z*uz] to beam insitu diagnostics by @AlexanderSinn in #964
• Fix bug in intra-species Coulomb collisions by @SeverinDiederichs in #967
• [mini] assert for normalized units when total charge is specified by @SeverinDiederichs in #969
• fix compiler warning: use static_cast for extent by @SeverinDiederichs in #970
• Plasma in-situ diags by @SeverinDiederichs in #968
• add Bz component to beam pusher by @SeverinDiederichs in #973
• Minor improvements for collision module. by @SeverinDiederichs in #972
• Generalize external E and B fields to uniform and linear in all directions by @MaxThevenet in #965
• add classical radiation reactions for beams by @SeverinDiederichs in #974
• fix some issues with the collision module by @SeverinDiederichs in #976
• MR with SALAME and PC by @AlexanderSinn in #971
• don't use managed memory by @AlexanderSinn in #975
• Add IntVect and RealVect overloads to queryWithParser by @AlexanderSinn in #977
• always use PhysConstSI in parser by @SeverinDiederichs in #978
• Cleaning collisions and preparing UpdateMomentumPerez for beams by @SeverinDiederichs in #980
• [mini] remove passing of ibox by @SeverinDiederichs in #982
• Add option to allocate mpi buffers in device memory by @AlexanderSinn in #979
• Remove z component from plasma by @AlexanderSinn in #984
• [mini] clean GetInstance() and use static inline by @SeverinDiederichs in #981
• remove OK assert by @AlexanderSinn in #985
• break loop over time steps for max_time by @SeverinDiederichs in #986
• Fix output logic for insitu diags by @SeverinDiederichs in #987
• Cut box to plasma radius in init by @AlexanderSinn in #988
• Refactoring documentation by @SeverinDiederichs in #989
• Add beam-plasma collisions by @SeverinDiederichs in #983
• [mini] document GPU-aware MPI for various clusters by @SeverinDiederichs in #992
• [mini] reset ID incrementor before starting initializing a beam by @MaxThevenet in #994
• Prepare release 23.07 by @MaxThevenet in #995

Full Changelog: v23.05...v23.07

v23.05

Dependencies

AMReX: release 23.05
openPMD-api: release 0.15.1

What's Changed

• CI: macOS Only One by @ax3l in #881
• On dev branch, we build on top of amrex dev branch by @MaxThevenet in #882
• Allow for arbitrary sized diagnostics grid by @AlexanderSinn in #871
• Reduce memory footprint of plasma initialization by @AlexanderSinn in #885
• Improvements on the adaptive time step calculation by @MaxThevenet in #884
• [mini] Print pusher for which HiPACE++ was compiled by @MaxThevenet in #888
• [mini] When a checksum test fails, also print the whole new json file by @MaxThevenet in #886
• [mini] Remove unused variable in adaptive time step by @MaxThevenet in #889
• [mini] rename to avoid compiler warning by @SeverinDiederichs in #890
• [mini] fix print statement for small DSTs using rocFFT by @SeverinDiederichs in #892
• CI on GPU by @MaxThevenet in #887
• Better defaults for production (in particular adaptive time step) by @MaxThevenet in #893
• [mini] fix concurrency in local GPU CI test by @MaxThevenet in #896
• [mini] Set default subcycling to 10 by @MaxThevenet in #895
• Upgrade to openPMD API v.15.0 by @MaxThevenet in #901
• Adaptbeams by @MaxThevenet in #899
• use serial openPMD by @SeverinDiederichs in #876
• [mini] Better print statement when checksum fails by @MaxThevenet in #905
• openPMD-api: 0.15.1 by @ax3l in #908
• Multiple diagnostic grids by @AlexanderSinn in #898
• Reorder plasma for deposition by @AlexanderSinn in #900
• Make leapfrog and explicit default by @MaxThevenet in #906
• Update doc to point to chat by @MaxThevenet in #910
• Use parser for beam density by @MaxThevenet in #912
• Remove zeta refinement by @AlexanderSinn in #911
• [mini] remove some unused parameters by @SeverinDiederichs in #913
• add LUMI documentation by @SeverinDiederichs in #914
• [mini] Add info on n_cell in doc by @MaxThevenet in #916
• Remove AmrCore by @AlexanderSinn in #915
• Support new AMReX particle interfaces by @AlexanderSinn in #921
• remove /usr/local from brew cache by @AlexanderSinn in #922
• Interpolate current to level 1 by @AlexanderSinn in #920
• Update LUMI doc by @SeverinDiederichs in #925
• Add time to CI ccache restore key by @AlexanderSinn in #926
• Remove NumPy from brew cache by @AlexanderSinn in #929
• Dynamic particle level by @AlexanderSinn in #924
• Fully initialize plasma every timestep by @AlexanderSinn in #927
• Add mailmap by @MaxThevenet in #931
• Update versions of Python packages to build RTD doc by @MaxThevenet in #934
• add config file for readthedocs by @SeverinDiederichs in #935
• macOS CI add HOMEBREW_NO_INSTALLED_DEPENDENTS_CHECK by @AlexanderSinn in #937
• fix a bunch of issues in the documentation by @SeverinDiederichs in #932
• Interpolate jx_beam and jy_beam to level 1 by @AlexanderSinn in #936
• update Perlmutter documentation by @SeverinDiederichs in #939
• update version tags by @SeverinDiederichs in #941

Full Changelog: v23.03...v23.05

v23.03

Dependencies

AMReX: release 23.03
openPMD-api: release 0.14.5

What's Changed

• plasmas.names is now an optional parameter, default no_plasma, and same for beam and laser by @MaxThevenet in #864
• fix typo found by Sarah by @SeverinDiederichs in #865
• Use only one MultiFab for all field slices by @AlexanderSinn in #868
• Assert that insitu and full diagnostics file prefixes are different by @AlexanderSinn in #869
• Increase tolerance on production runs by @MaxThevenet in #875
• Read in Laser form OpenPMD file by @AlexanderSinn in #866
• Output normalized momentum instead of proper velocity in insitu diagnostics by @AlexanderSinn in #870
• CI: macOS by @ax3l in #874
• CI: Windows by @ax3l in #873
• Use only one MultiFab for all laser slices by @AlexanderSinn in #867
• Update release number and fix dependency by @MaxThevenet in #880

Full Changelog: v23.02...v23.03

v23.02

Dependencies

AMReX: release 23.02
openPMD-api: release 0.14.5

What's Changed

• fix NVCC 11.5+ compile issue by @AlexanderSinn in #833
• update CUDA CI to 11.8 by @SeverinDiederichs in #834
• Mesh Refinement with Explicit Solver by @AlexanderSinn in #835
• [mini] Clearly state in the doc that HiPACE++ is 3D by @MaxThevenet in #840
• Remove numprocs_x/y in all inputs, useless and confusing by @MaxThevenet in #841
• Add production examples by @MaxThevenet in #842
• Upper bound to time step when adaptive by @MaxThevenet in #845
• Bug fix:reset plasma only once the time step is known by @MaxThevenet in #846
• Reset plasma position before push by @AlexanderSinn in #848
• User defined Ez chirp in SALAME algorithm by @AlexanderSinn in #852
• option to focus the beam at arbitrary position by @MaxThevenet in #853
• Remove unnecessary copy of BeamBins by @AlexanderSinn in #855
• Gather ExmBy and EypBx as derivatives of Psi by @AlexanderSinn in #850
• HPMG: Support for 2^n-1 cells per direction by @WeiqunZhang in #854
• amrex::Gpu::Atomic::Inc -> amrex::Gpu::Atomic::Add by @WeiqunZhang in #859
• Add leapfrog pusher for plasma by @AlexanderSinn in #847
• Adjust boundary condition for MR with 2^n-1 HPMG by @AlexanderSinn in #857
• Get correct central point for the phase calculation in laser advance by @MaxThevenet in #856
• Fix AdaptiveTimeStep for low density beams by @AlexanderSinn in #862
• Prevent perfectly centered plasma particle by @AlexanderSinn in #861
• replace 22.12 with 23.02 by @MaxThevenet in #863

Full Changelog: v22.12...v23.02

v22.12

Dependencies

AMReX: release 22.12
openPMD-api: release 0.14.5

What's Changed

• Update version number before release by @MaxThevenet in #812
• Fix invalid CUDA device function error in HPMG by @AlexanderSinn in #813
• [mini] print number of iterations only for PC solver by @SeverinDiederichs in #816
• Automatic Beam loading using a near direct SALAME algorithm by @AlexanderSinn in #814
• add documentation for Perlmutter by @SeverinDiederichs in #817
• Fix errors in I/O by @SeverinDiederichs in #818
• Use ReduceOp instead of AtomicAdd in AdaptiveTimeStep by @AlexanderSinn in #820
• [mini] option to not use the phase for laser envelope advance by @MaxThevenet in #821
• clean up deposition using ParallelFor with compile time optimizations by @AlexanderSinn in #819
• Fix Juwels compile issue by @AlexanderSinn in #825
• [mini] Fix docstring of laser gather function by @AlexanderSinn in #826
• Enable multiple laser pulses by @MaxThevenet in #823
• reset laser slices before DepositNeutralizingBackground by @SeverinDiederichs in #827
• Remove spurious comments in laser handling by @MaxThevenet in #828
• [mini] Enable laser transverse offset by @MaxThevenet in #830
• Add derivative shape factors by @AlexanderSinn in #831
• Udate version number by @MaxThevenet in #832

Full Changelog: v22.11...v22.12