ABCDEFGHIJKLMNOPQRSTUVWXYZAAABACADAEAFAGAHAIAJAKAL
1
SoftwareConsortiumPartnerContact email (correspondant technique)To Be benchmarked,
To Be benchmarked (2024)		CommentsRepositoryLicenseParallelismData Résilence,
Résilience		DevOpsPCWP1WP2WP3WP4WP5WP6WP7Interfaced with
2
GeoSLLNL,Stanford,TEInria BXSOstefano.frambati@totalenergies.frNOT YEThttps://github.com/orgs/GEOSXOSS:: LGPL v*C++discretisation Elt fini/Elt Spectraux/Vol Finis,
couplage multiphysique		ROM: RB, POD, ...,
PINN		PBI deterministiciterative methodsGeophysics, Energy,
proxy-apps
3
Freefem++Sorbonne USorbonne U,
Inria PARIS		frederic.hecht@sorbonne-universite.fr,
frederic.hecht@sorbonne-universite.fr, pierre-henri.tournier@sorbonne-universite.fr, pierre.jolivet@sorbonne-universite.fr,
frederic.hecht@sorbonne-universite.fr,
pierre-henri.tournier@sorbonne-universite.fr, pierre.jolivet@sorbonne-universite.fr,
frederic.hecht@sorbonne-universite.fr,
pierre-henri.tournier@sorbonne-universite.fr
pierre.jolivet@sorbonne-universite.fr		CPUDSL pour les éléments finis. Le parallélisme est basé sur MPI.https://github.com/FreeFem/FreeFem-sourcesOSS:: LGPL v*C++MPIVTK,
in-house format,
HDF5,
Gmsh and asssociated formats		Continuous integration,
Test - Unit,
Test - Validation,
Container - Docker,
Packages - Debian		PC1 - ExaMA,
PC2 - WP1 - High-level approaches for developing efficient and composable parallel software,
PC3 - WP2 - Entrées sorties, stockage,
PC1 - ExaMA,
PC2 - WP1 - High-level approaches for developing efficient and composable parallel software,		cG, dG/hdG, adaptation de maillage, maillage non structuré, couplage multiphysiqueDDM, Couplage multi-physique Algebrique, multi-rhs, krylov reuseiterative methodsGeophysics, Energy,
Health,
Environment, Energy,
Aero,
mini-apps		PETSc,
MMG/ParMMG,
Scotch,
MUMPS,
HPDDM
4
Feel++Feel++ ConsortiumUnistra,
Inria Grenoble,
CNRS		christophe.prudhomme@cemosis.fr, vincent.chabannes@cemosis.frCPUGitHub - feelpp/feelpp: Feel++: Finite Element Embedded Language and Library in C++OSS:: LGPL v*,
OSS:: GPL v*		C++17,
Python		MPI,
Parallelism - C++17 and after,
Task based - C++		Json,
YAML,
HDF5,
Data-management system,
VTK,
in-house format,
VTK,
Ensight,
Gmsh and asssociated formats		checkpoint restartContinuous integration,
Container - Docker,
Container - Singularity,
Packages - Debian,
Packages - Ubuntu,
Packages - Fedora,
Packages - Spack,
Test - Unit,
Test - Verification,
Test - Validation		PC2 - WP3 - Runtime Systems,
PC2 - WP4 - Portable, scalable numerical building blocks and software,
PC1 - ExaMA,
PC2 - WP1 - High-level approaches for developing efficient and composable parallel software,
PC2 - WP2 - Just-in-Time code optimization with continuous feedback loop		maillage non structuré,
adaptation de maillage,
cG,
dG/hdG,
parallelisation en temps,
couplage multiphysique,
couplage multi-échelle,
inhouse,
interface,
discretisation Elt fini/Elt Spectraux/Vol Finis		ROM-DA: GEIM PBDW, ...,
ROM: NIRB,
ROM: RB, POD, ...		DDM, Couplage multi-physique AlgebriqueDA-stochastic: Ensemble iterative methodsinterfaceHealth, Environment, Energy, Physics,
mini-apps		OpenTurns,
MMG/ParMMG,
PETSc,
Salome,
Dymola/OpenModelica/FMU,
HPDDM
5
ScimbaUNISTRA, INRIAUnistraemmanuel.franck@inria.fr,
emmanuel.franck@inria.fr,,
emmanuel.franck@inria.fr,

emmanuel.franck@inria.fr		​PythonGPU NN/Autoencoder,
PINN		pytorch
6
MMG/ParMMGInriaInria BXSO,
U Grenoble Alpes,
Sorbonne U		CPUC,
Fortran		MPIin-house formatContinuous integrationmaillage non structuré,
adaptation de maillage
7
CGALInriaInria CApierre.alliez@inria.frCPUPas encore de code GPU.,
Pas encore de code GPU.
Des algorithmes en parallème mémoire partagée,
Pas encore de code GPU.
Des algorithmes en parallèle (mémoire partagée, Intel TBB),
Pas encore de code GPU.,
Des algorithmes en parallèle (mémoire partagée, Intel TBB)		https://github.com/CGALOSS:: GPL v*,
OSS:: LGPL v*		C++maillage non structuré
8
HawenInriaInria BXSOflorian.faucher@inria.frYeshttps://gitlab.com/ffaucher/hawenOSS:: GPL v*FortranMPI,
Multithread-OpenMP		in-house format,
VTK,
Gmsh and asssociated formats		dG/hdG,
couplage multiphysique		multi-rhs, krylov reusePBI deterministiciterative methodsSensitivity analysisGeophysics, Energy,
Astrophysics,
mini-apps		MUMPS
9
UranieCEACEArudy.chocat@cea.fr, jean-baptiste.blanchard@cea.frCPUhttps://sourceforge.net/projects/uranie/OSS:: LGPL v*C++,
Python		MPI,
Multithread,
GPU 		Json,
ASCII,
ROOT,
SQL		Continuous integration,
Test - Unit,
Test - Verification,
Test - Validation		NN/Autoencoderiterative methods,
Robust optimisation,
metaheuristic		Sensitivity analysis,
U propagation,
Robust optimisation		Salome
10
SalomeEDF,CEA​
11
Dymola/OpenModelica/FMUModelica & FMI​C++couplage multiphysique,
couplage multi-échelle
12
PETScArgonne National LaboratorySorbonne UINDIRECTPETSc is benchmarked via Feel++ and FreeFEM++ at leastOSS:: 2-clause BSDC,
C++,
Fortran,
Python,
Julia		MPI,
Multithread-OpenMP,
GPU 		XML,
YAML,
HDF5,
Json,
in-house format,
VTK,
Gmsh and asssociated formats,
MED		Continuous integration,
Packages - Debian,
Packages - Ubuntu,
Packages - Fedora,
Packages - Other,
Packages - Spack,
Test - Unit,
Test - Verification,
Test - Validation		PC1 - ExaMA,
PC2 - WP4 - Portable, scalable numerical building blocks and software		DDM,
Couplage multi-physique Algebrique,
multi-precision,
calcul tensoriel,
multi-rhs, krylov reuse,
randomization,
low-rank,
interface		Freefem++,
Feel++,
MMG/ParMMG,
MUMPS,
Scotch,
PaStIX,
HPDDM
13
ScotchInria BXSOINDIRECTBenchmark via Maphys++C
14
StarPUInria BXSOINDIRECTC
15
MUMPSMumps TechnologiesSorbonne UINDIRECThttp://mumps-solver.org/OSS: Cecill-*FortranMPI,
Multithread-OpenMP		multi-precision,
multi-rhs, krylov reuse,
low-rank		Scotch
16
PaStIXInria, Université Bordeaux, BINPInria BXSOINDIRECTOSS: Cecill-*CMultithread,
MPI		Continuous integration,
Packages - GUIX-HPC,
Container - Singularity,
Test - Verification		PC2 - WP3 - Runtime Systems,
PC2 - WP1 - High-level approaches for developing efficient and composable parallel software		direct solver,
low-rank		Scotch
17
qr_mumpsCNRSINDIRECT
Benchmark via Maphys++, mais sera benckmarké en stand alone dans PC2,
Benchmark via Maphys++, mais sera
benckmarké en stand alone dans PC2
https://gitlab.com/qr_mumps/qr_mumps,
https://gitlab.com/qr_mumps/qr_mumps		OSS: Cecill-*FortranTask based - RuntimeContinuous integration,
Packages - GUIX-HPC,
Container - Singularity,
Test - Verification		PC2 - WP1 - High-level approaches for developing efficient and composable parallel software,
PC2 - WP3 - Runtime Systems		svd, eigen solver,
low-rank		StarPU,
Scotch
18
ScalFMMInria BXSOolivier.coulaud@inria.frNOT YEThttps://gitlab.inria.fr/solverstack/ScalFMMOSS: Cecill-*C++Task based - Runtime,
Multithread-OpenMP,
GPU ,
MPI		Continuous integration,
Packages - GUIX-HPC,
Container - Singularity,
Test - Verification		PC2 - WP1 - High-level approaches for developing efficient and composable parallel software,
PC2 - WP3 - Runtime Systems		dense / H matrixStarPU
19
FabulousInria BXSOgilles.marait@inria.frINDIRECThttps://gitlab.inria.fr/solverstack/fabulousOSS: Cecill-*C++Continuous integration,
Packages - GUIX-HPC,
Container - Singularity,
Test - Verification		PC2 - WP1 - High-level approaches for developing efficient and composable parallel software,
PC2 - WP3 - Runtime Systems		krylov solver,
multi-rhs, krylov reuse,
multi-precision
20
Maphys++Inria BXSOgilles.marait@inria.frHYBRIDhttps://gitlab.inria.fr/solverstack/maphys/maphysppOSS: Cecill-*C++,
C,
Fortran		MPI,
Multithread,
GPU 		Continuous integration,
Packages - GUIX-HPC,
Container - Singularity,
Test - Verification		PC2 - WP3 - Runtime Systems,
PC2 - WP1 - High-level approaches for developing efficient and composable parallel software		direct solver,
svd, eigen solver,
krylov solver		PaStIX,
qr_mumps,
MUMPS,
Scotch
21
HPDDMSorbonne U,
Inria PARIS		pierre@joliv.etCPUC++/MPI,
Librairie autonome écrite en C++/MPI ; interfacée avec PETSc, FreeFEM, Feel++, Code_Aster, ...		https://github.com/hpddm/hpddmOSS:: LGPL v*C,
C++,
Fortran,
Python		MPI,
Multithread-OpenMP		in-house formatContinuous integration,
Test - Unit,
Test - Verification		PC1 - ExaMA,
PC2 - WP4 - Portable, scalable numerical building blocks and software		DDM,
multi-precision,
calcul tensoriel,
multi-rhs, krylov reuse,
randomization,
low-rank,
interface		Freefem++,
Feel++,
PETSc,
MUMPS,
PaStIX
22
23
PROMISESorbonne UFabienne.Jezequel@lip6.frNOT YEThttp://promise.lip6.fr/OSS:: LGPL v*PythonContinuous integrationPC1 - ExaMAmulti-precisionCADNA
24
CADNASorbonne UFabienne.Jezequel@lip6.frNOT YEThttp://cadna.lip6.fr/OSS:: LGPL v*C++,
Fortran		MPI,
Multithread-OpenMP,
GPU 		Continuous integrationPC1 - ExaMAmulti-precision
25
SamuraiIP ParisCEA,
IPP		Loic Gouarin​https://github.com/hpc-maths/samuraiOSS::BSDC++,
C++14,
C++17		MPI,
Multithread		HDF5Continuous integration,
Test - Unit,
Test - Verification,
Test - Validation,
Packages - Other		PC1 - ExaMA,
PC2 - WP1 - High-level approaches for developing efficient and composable parallel software,
PC2 - WP3 - Runtime Systems,
PC2 - WP4 - Portable, scalable numerical building blocks and software,
PC3 - WP2 - Entrées sorties, stockage		adaptation de maillagePETSc
26
ZellijUniversité de LilleInria Lilleel-ghazali.talbi@univ-lille.frGPUhttps://github.com/ThomasFirmin/zellijOSS: Cecill-*PythonMPIcheckpoint restartmetaheuristic,
iterative methods
27
pBBUniversité de LilleInria Lillenouredine.melab@univ-lille.frHYBRIDhttps://gitlab.inria.fr/jgmys/permutationbbOSS: Cecill-*C++MPI,
GPU ,
Multithread,
Chapel		checkpoint restartiterative methods
28
TRUST PlatformCEACEApierre.ledac@cea.frHYBRIDPossibilité de benchs CPU-only, GPU-only et vrai HYBRIDhttps://github.com/cea-trust-platformOSS::C++MPI,
GPU 		HDF5,
MED,
VTK		Continuous integration,
Test - Unit,
Test - Verification		PC1, PC2, PC3, PC5Couplage multi-physique
29
MEDCouplingCEA, EDFCEANOT YETPas en première phase de benchmark. A venir ensuite quand couplage mis en place.https://github.com/cea-trust-platformOSS::C++MPIHDF5,
MED		Continuous integration,
Test - Unit,
Test - Verification		PC1, PC2, PC3, PC5Couplage multi-physique
30
ICoCo coupling interfaceCEACEANOT YETPas en première phase de benchmark. A venir ensuite quand couplage mis en place.https://github.com/cea-trust-platform/icoco-couplingOSS::C++MPIMEDPC1, PC2, PC3, PC5Couplage multi-physique
31
EUROPLEXUSCEA, JRC European Commission, EDF, ONERA, Safran Tech, CEANoInutile à benchmarker seul dans ExaMA.Fortran,
C++		MPIMED,
VTK		Continuous integration,
Test - Unit,
Test - Verification		PC1, PC2, PC3, PC5Couplage multi-physique
32
MANTACEA + consortium in development (see EUROPLEXUS)CEAolivier.jamond@cea.frCPUAccès au code et repo à voir avec le contact Olivier Jamond (solutions transitoires à ce stade au niveau CEA)C++MPIMED,
VTK,
Gmsh and asssociated formats,
MFront		Continuous integration,
Test - Unit,
Test - Verification		PC1, PC2, PC3, PC5Couplage multi-physiqueEnergy,
mini-apps,
proxy-apps
33
MFEMCEANOT YETA venir avec la thèse AMR peut-être en 2025https://github.com/mfem/mfemOSS::C++MPI,
Multithread		VTK,
Gmsh and asssociated formats		Continuous integration,
Test - Unit,
Test - Verification		PC1 - ExaMAFE exascale framework,
AMR		wave propagation in heterogeneous media,
PBI deterministic
34
PLEIADESCEA,
EDF,
Framatome		CEANOT YETPlus tard quand PLEIADES-HPC sera prêtFortran,
Python,
C++		MPI,
Multithread		MED,
HDF5,
XML,
MFront		Continuous integration,
Test - Unit,
Test - Verification		PC1 - ExaMA,
PC5 - Demonstrateur		couplage multi-échelle,
couplage multiphysique		Couplage multi-physiquemini-apps
35
tensorflow​Pytorch utilisé par les collègues, mais plus riche et plus stable, mais approche plus complexe ?C++,
C++17,
Python		NN/Autoencoder,
GAN
36
scikit-learn​Sera développé dans PC3 plutôtC++NN/Autoencoder,
GAN
37
pytorch​C++NN/Autoencoder,
GAN		PBI stochastic
38
melissaINRIA​Développé par Bruno Raffin ==> Sera benchmarké dans PC3,
Développé par Bruno Raffin ==> Sera benchmarké dans PC3. On sera utilisateurs seulement.,
Développé par Bruno Raffin ==> Sera benchmarké dans PC3. On sera utilisateurs seulement.		https://gitlab.inria.fr/melissaBSD3python, C/C++, fortran90DA-stochastic: Ensemble
39
crocoInria, IRD, CNRS, Univ Paul Sabatier, SHOM, Ifremer Laurent.Debreu@inria.fr​Softs volumineuxhttps://gitlab.inria.fr/croco-oceanFortranMPI,
GPU ,
Multithread-OpenMP		netcdf,
HDF5		checkpoint restartDA-deterministic: ndVARproxy-apps
40
OpenTURNSAirbus, EDF, IMACS, ONERA, Phimeca​https://openturns.github.ioOSS:: LGPL v*,
OSS:: GNU LGPL version 3		C++,
Python		MPIHDF5Continuous integration,
Test - Unit,
Test - Validation		PC1 - ExaMAiterative methodsSensitivity analysis,
U propagation,
Robust optimisation
41
Dakota​C++Sensitivity analysis,
U propagation,
Robust optimisation
42
wavenetDeepMind​https://github.com/benmoseley/seismic-simulation-wavenetMITpython, pytorch,
python, pytorch, tensorflow		PC1PBI stochastic proxy-apps
43
deeponet​https://github.com/lululxvi/deeponetCC BY-NC-SA 4.0python, matlab,
python, matlab, pytorch, tensorflow		PC1PBI stochastic
44
Arcane FrameworkCEA, IFPENCEAlydie.grospellier@cea.frHYBRIDPossibilité de benchs CPU-only, GPU-only avec le même binaire. https://github.com/arcaneframework/frameworkOSS: apache-2C++,
C#		MPI,
Multithread,
Multithread-TBB,
GPU ,
Task based - Runtime		XML,
HDF5,
Ensight,
Json		checkpoint restartContinuous integration,
Test - Unit,
Test - Verification		maillage non structuré,
adaptation de maillage,
AMR,
discretisation Elt fini/Elt Spectraux/Vol Finis,
couplage multiphysique		Couplage multi-physiqueEnvironment, Energy,
Physics
45
MaHyCoCEACEAjean-philippe.perlat@cea.frHYBRIDPossibilité de benchs CPU-only, GPU-only avec le même binaire.C++GPU , MPI, XMLcheckpoint restartContinuous integrationPC1 - ExaMAmaillage non structuréCouplage multi-physiqueExact methodsPhysicsArcane Framework
46
SpecxInriaInriaberenger.bramas@inria.fr​https://github.com/berenger-eu/specxOSS:: LGPL v*C++17Task based - RuntimeContinuous integration,
Test - Unit,
Test - Verification		PC2 - WP3 - Runtime SystemsFeel++
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100