Vlasiator

This new Vlasiator release adds:

• Full support for GPU computation on HPC environments to Vlasiator, targeting both Nvidia (CUDA) and AMD (HIP) platforms.
• A subgrid-scale wave-particle scattering model (implemented as velocity space diffusion) has been added for magnetosheath waves.
• Vorticity and temperature / pressure anisotropy can now be taken into account for mesh refinement criteria.
• Improved the code's runtime resilience by including a mechanism to write frequent recovery files using a circular buffer scheme.
• The code is now ported (or portable) to all EuroHPC machines, as of December 2025.
• In addition, there are a large number of bugfixes to time-varying inflow, system boundary handling, ionosphere stability and overall quality of life.

More detailed list of what's Changed:

• Added blocks gather code for FALLBACK vectors by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/444
• Vec8Simple by @ykempf in https://github.com/fmihpc/vlasiator/pull/445
• Some openacc experiments in the acceleration solver. by @ursg in https://github.com/fmihpc/vlasiator/pull/446
• Support of MPI ranks sharing GPUs. by @ykempf in https://github.com/fmihpc/vlasiator/pull/447
• Restructured OpenACC acceleration solver to parallelize over columns. by @ursg in https://github.com/fmihpc/vlasiator/pull/448
• Cudasiator mutator, split plm/ppm/pqm etc back into .cuh files by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/595
• Cudasiator realf vectorclass by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/597
• Implementation of CUDA streams (one per thread) by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/616
• Cudasiator by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/790
• Vlasiator gpu by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/799
• Added testpackage script for mahti_cuda by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/803
• ARCH fixes to LUMI-G compilation, squelch some warnigns, added NOMAD … by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/805
• Cudasiator pre-NOMAD-Hackathon fixes by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/806
• Cudasiator preparation for the NOMAD hackathon by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/808
• Fix project initialization fixing due to bogus free by @ursg in https://github.com/fmihpc/vlasiator/pull/809
• Deactivated xnack on LUMI, added trial config for all datareducers by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/810
• Add -march=znver3 to lumi gpu makefiles by @ursg in https://github.com/fmihpc/vlasiator/pull/811
• Don't NACK in the lumi gpu testpackage script by @ursg in https://github.com/fmihpc/vlasiator/pull/813
• Recapacitate splitVectors in acceleration if they are too small. by @ursg in https://github.com/fmihpc/vlasiator/pull/812
• Added method for applying reservation and call it after LB. Also swit… by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/814
• Stop your damn timers! by @lkotipal in https://github.com/fmihpc/vlasiator/pull/815
• Fix timer by @lkotipal in https://github.com/fmihpc/vlasiator/pull/816
• Fix arch-interface issues in DRO and add HIP kernel launch error checking by @hokkanen in https://github.com/fmihpc/vlasiator/pull/817
• Fix kernel launch issue and dynamic shared memory type size issue by @hokkanen in https://github.com/fmihpc/vlasiator/pull/819
• Fix issue with restart_read testpackage test in GPU branch by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/835
• Gpu fix restarts, update to dev by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/837
• Prefer --offload-arch over the deprecated --amdgpu-target and fix commit parsing in Makefile. by @sfantao in https://github.com/fmihpc/vlasiator/pull/844
• Use multiple handlers of device mesh wrapper to avoid relocatable code device builds by @sfantao in https://github.com/fmihpc/vlasiator/pull/868
• Avoid rdc device code mk2 by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/872
• Vlasiator gpu fix cpu compilation and running by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/922
• Purge all remaining references to velocity mesh refinement by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/923
• Vlasiator gpu bring up to date with dev by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/924
• Vlasiator gpu fix timevarying by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/928
• fix gpu translation DZ access from pencils by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/929
• Vlasiator_gpu update on LUMI to 23.09 by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/936
• Fix compute-time loop by @hokkanen in https://github.com/fmihpc/vlasiator/pull/938
• Updated Mahti_cuda makefile to use gcc-10.4.0 and cuda-12.1.1 by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/937
• Reverting submodule urls to https instead of ssh, for easier accessib… by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/945
• Gpu update - new block adjustment algorithm and plenty of other improvements by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/958
• Gpu remove map add by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/962
• Fix GPU branch crash on several MPI Tasks by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/965
• Optimizations to translation memory ordering, and fix to broken ifdefs by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/972
• Vlasiator gpu hashinator update by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/979
• GPU: Update Hashinator interface to use templated prefetch deactivations by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/978
• Removed block offset evaluation kernel in ACC; included functionality… by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/980
• Add OMP support for ARCH host backend by @hokkanen in https://github.com/fmihpc/vlasiator/pull/969
• Vlasiator gpu branch, fix dynamic AMR operations by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/981
• Purged config references to the GPUBLOCKS / CUDABLOCKS parameter and … by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/983
• Gpu single trans kernel launch dev by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/1021
• Split cell transfer into a number of passes in load balance. by @ykempf in https://github.com/fmihpc/vlasiator/pull/1017
• Different trans thread guide by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/1032
• Update vlasiator to use updated fsgrid constructor by @cscjlan in https://github.com/fmihpc/vlasiator/pull/1019
• Lumi 2403 makefile by @JonasSuni in https://github.com/fmihpc/vlasiator/pull/1031
• GCC 13.2.0 carrington by @ykempf in https://github.com/fmihpc/vlasiator/pull/1035
• Update DCCRG commit by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/1037
• Caching of DCCRG <-> FSgrid coupling by @veetihaaja in https://github.com/fmihpc/vlasiator/pull/1012
• Better documentation for alpha1 constituent weighting by @lkotipal in https://github.com/fmihpc/vlasiator/pull/1001
• Fix check vlasiator cfg script by @ykempf in https://github.com/fmihpc/vlasiator/pull/1024
• Add jemalloc purge calls to after shrink_to_fit and deallocation of r… by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/1027
• Makefile (and small library build script fix) for Meluxina. by @ursg in https://github.com/fmihpc/vlasiator/pull/1041
• Use curl instead of wget in build_libraries.sh by @ursg in https://github.com/fmihpc/vlasiator/pull/1047
• Make chunk data script use 5 GB chunks, for Allas by @JonasSuni in https://github.com/fmihpc/vlasiator/pull/1046
• Correct EGradPe/derivatives communications in subcycling by @ykempf in https://github.com/fmihpc/vlasiator/pull/1038
• fsgrid RAII by @lkotipal in https://github.com/fmihpc/vlasiator/pull/918
• No L2 NOT_SYSBOUNDARY EXB nudge by @ykempf in https://github.com/fmihpc/vlasiator/pull/1050
• GradPe fix cleanup by @ykempf in https://github.com/fmihpc/vlasiator/pull/1051
• Build testpackage data with new gcc version on carrington. by @ursg in https://github.com/fmihpc/vlasiator/pull/1054
• Fixed typo in documentation. by @ykempf in https://github.com/fmihpc/vlasiator/pull/1056
• Reverse direction when stepping back to hit exact radius when tracing. by @ykempf in https://github.com/fmihpc/vlasiator/pull/1057
• New amr criteria by @lkotipal in https://github.com/fmihpc/vlasiator/pull/1022
• Multipop pressure anistropy by @lkotipal in https://github.com/fmihpc/vlasiator/pull/1052
• Ghost translation dev merge by @ursg in https://github.com/fmihpc/vlasiator/pull/1060
• Ghost translation by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/1006
• Dccrg RAII by @lkotipal in https://github.com/fmihpc/vlasiator/pull/953
• Multipop pressure anistropy by @lkotipal in https://github.com/fmihpc/vlasiator/pull/1061
• Update C++ standard from C++17 to C++20 in Makefile.lumi_gnu_2403 by @JonasSuni in https://github.com/fmihpc/vlasiator/pull/1063
• Fix a broken error message in pencil construction. by @ursg in https://github.com/fmihpc/vlasiator/pull/1071
• Slope limiter pass by reference by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/1069
• Limit tracing volume by @ykempf in https://github.com/fmihpc/vlasiator/pull/1055
• Simple matplotlib plotter to help setting VDF parameters. by @ykempf in https://github.com/fmihpc/vlasiator/pull/1073
• Inspection script mechanism to use srun overlap. Works on LUMI at least. by @ykempf in https://github.com/fmihpc/vlasiator/pull/1068
• Node hours report by @ArnaudLalague in https://github.com/fmihpc/vlasiator/pull/1053
• Use ARCH=arch in VLSV build by @ykempf in https://github.com/fmihpc/vlasiator/pull/1075
• Activate report grid memory consumption by @ykempf in https://github.com/fmihpc/vlasiator/pull/1020
• Vlasiator gpu by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/1074
• Warning to check dd_chunk.err in the file handling scripts. by @ykempf in https://github.com/fmihpc/vlasiator/pull/1090
• update DCCRG submodule by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/1089
• Fix order for running test printouts in CI by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/1097
• Add support for multiple boxes where refinement is allowed by @JonasSuni in https://github.com/fmihpc/vlasiator/pull/1095
• Reduce some specific timers that cause a wide spread of phiprof files. by @ykempf in https://github.com/fmihpc/vlasiator/pull/1098
• Reorder memory report lines in logfile. by @ykempf in https://github.com/fmihpc/vlasiator/pull/1094
• GPU improved launch parameters, memory usage, etc. by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/1100
• Fix un/refinement not working if no boxes are defined by @JonasSuni in https://github.com/fmihpc/vlasiator/pull/1105
• CI flag builds in dev by @ykempf in https://github.com/fmihpc/vlasiator/pull/1104
• Periodically overwriting restarts by @ykempf in https://github.com/fmihpc/vlasiator/pull/1092
• Citations file to repo by @ykempf in https://github.com/fmihpc/vlasiator/pull/1067
• Hile makefiles by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/1107
• Improve documentation of EgradPe term in field solver code. by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/1106
• Revert addition of std::format from PR #1094 by @ursg in https://github.com/fmihpc/vlasiator/pull/1101
• Add SIGTERM signal handler, for slurm preemption. by @ursg in https://github.com/fmihpc/vlasiator/pull/1059
• Remove vec from fluxfunction and fix compilation by @ykempf in https://github.com/fmihpc/vlasiator/pull/1029
• Fix refine/coarsen logging by @lkotipal in https://github.com/fmihpc/vlasiator/pull/1036
• Reorder acceleration preparations so that always-on-cpu tasks are sep… by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/1112
• Typo fix by @JonasSuni in https://github.com/fmihpc/vlasiator/pull/1111
• Further improvements to egradpe documentation based on user feedback. by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/1114
• Return to considering has_content blocks of spatial neighbours at all… by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/1115
• Alternative Fieldsolver magnetic field boundary cell parallel approach by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/1110
• Sub-grid model for pitch-angle diffusion by @MaximeII in https://github.com/fmihpc/vlasiator/pull/846
• Fix generate ref data on Carrington by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/1117
• SysBoundary communicate VDFs only in smaller neighborhood by @ykempf in https://github.com/fmihpc/vlasiator/pull/1049
• Fix pop_1dmuspace output reducer which was corrupting some testpackag… by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/1119
• prevent VBC shrink on AMD-G (still not working), switch lumi_hipcc to… by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/1118
• Add CI action for generating reference testpackage data on Ukko DGX by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/1120
• Correct dependency name in generate reference data action by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/1123
• Fix gpu moments by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/1124
• Bin pencils by @lkotipal in https://github.com/fmihpc/vlasiator/pull/1096
• Rewritten Acceleration solver for GPUs by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/1125
• Vlasiator GPU to propagate WID3 at once in acceleration as well by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/1129
• GPU pitch angle diffusion by @MikaelHuppunen in https://github.com/fmihpc/vlasiator/pull/1128
• Dev alfven double normalization fix by @veetihaaja in https://github.com/fmihpc/vlasiator/pull/1136
• Remove useless loop by @ykempf in https://github.com/fmihpc/vlasiator/pull/1138
• Fix comment typos in parameters.cpp by @OssiLeopold in https://github.com/fmihpc/vlasiator/pull/1134
• Gpu optimization by @MikaelHuppunen in https://github.com/fmihpc/vlasiator/pull/1141
• GPU bin pencils by @MikaelHuppunen in https://github.com/fmihpc/vlasiator/pull/1143
• Vlsvdiff WID fixes by @lkotipal in https://github.com/fmihpc/vlasiator/pull/1151
• Add Hile-G TP build CI job by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/1132
• Ci docker and artifact version updates by @ursg in https://github.com/fmihpc/vlasiator/pull/1165
• At BSC, run risc-v CI on pioneer cluster instead of synth-hca by @ursg in https://github.com/fmihpc/vlasiator/pull/1168
• Encapsulate analysator to venv and let analysator handle its dependencies. by @alhom in https://github.com/fmihpc/vlasiator/pull/1166
• Fix multi-gpu buffer allocations issue by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/1142
• HILE-C CI scripts by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/1156
• Add Mahti gcc_build makefile to support library building scripts, tra… by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/1145
• Scaling results run on Mahti, with extracted data, plotting scripts and resulting plots. by @ykempf in https://github.com/fmihpc/vlasiator/pull/1131
• Phiprof acc chunk fix by @markusbattarbee in https://github.com/fmihpc/vlasiator/pull/1159
• Add width specifier to fscanf to fix stack-buffer-overflow by @kstppd in https://github.com/fmihpc/vlasiator/pull/1121
• Fix incorrect vorticity (Fixes #1148) by @JonasSuni in https://github.com/fmihpc/vlasiator/pull/1149
• Refine report improvements by @lkotipal in https://github.com/fmihpc/vlasiator/pull/1147
• Added warning in inflow.cpp loadFile() by @lassejsc in https://github.com/fmihpc/vlasiator/pull/1195
• Added a GPU memory manager by @MikaelHuppunen in https://github.com/fmihpc/vlasiator/pull/1170
• Add option to disable finite differencing at sysboundaries by @JonasSuni in https://github.com/fmihpc/vlasiator/pull/1190
• Reapply initial state in boundary cells after setProjectBField by @JonasSuni in https://github.com/fmihpc/vlasiator/pull/1193
• Do not reapply solved B components in boundary cells except at simulation start by @JonasSuni in https://github.com/fmihpc/vlasiator/pull/1194
• Copy fsgrid moments to outflow after filtering them by @JonasSuni in https://github.com/fmihpc/vlasiator/pull/1202
• Synchronize mpiGrid and FsGrid after inflow update by @JonasSuni in https://github.com/fmihpc/vlasiator/pull/1204
• Improvements to GPU memory manager by @MikaelHuppunen in https://github.com/fmihpc/vlasiator/pull/1199
• Add a friendly banner at the top of Makefile, for no reason at all. by @ursg in https://github.com/fmihpc/vlasiator/pull/1198
• Implement Vectorclass Supporting SVE Instructions by @lnghrdntcr in https://github.com/fmihpc/vlasiator/pull/1127
• GitHub pages Doxygen generation by @alhom in https://github.com/fmihpc/vlasiator/pull/1205
• Remove MacOS from CI builds, try building on github arm64 instead. by @ursg in https://github.com/fmihpc/vlasiator/pull/1206
• Fixed a bug in GPU memory manager by @MikaelHuppunen in https://github.com/fmihpc/vlasiator/pull/1197

New Contributors

• @sfantao made their first contribution in https://github.com/fmihpc/vlasiator/pull/844
• @cscjlan made their first contribution in https://github.com/fmihpc/vlasiator/pull/1019
• @MaximeII made their first contribution in https://github.com/fmihpc/vlasiator/pull/846
• @MikaelHuppunen made their first contribution in https://github.com/fmihpc/vlasiator/pull/1128
• @OssiLeopold made their first contribution in https://github.com/fmihpc/vlasiator/pull/1134
• @lassejsc made their first contribution in https://github.com/fmihpc/vlasiator/pull/1195
• @lnghrdntcr made their first contribution in https://github.com/fmihpc/vlasiator/pull/1127

Full Changelog: https://github.com/fmihpc/vlasiator/compare/v5.3.1...v5.4.0