Pasar al contenido principal

Resultados relacionados

Resultados totales (Incluyendo duplicados): 12
Encontrada(s) 1 página(s)

Digibug. Repositorio Institucional de la Universidad de Granada
Publicación Artículo científico (JournalArticle)

On the feasibility and usefulness of high-performance computing in probabilistic volcanic hazard assessment: An application to tephra hazard from Campi Flegrei

  • Martínez Montesinos, Beatriz
  • Titos Luzón, Manuel Marcelino
  • Sandri, Laura
  • Rudyy, Oleksandr
  • Cheptsov, Alexey
  • Macedonio, Giovanni
  • Folch, Arnau
  • Barsotti, Sara
  • Selva, Jacopo
  • Costa, Antonio
The research leading to these results has been supported by the European Commission’s CORDIS [ChEESE (grant no. 823844)]. Part of this work was supported by European Community’s Horizon 2020 POP project under grant agreements (no. 824080). Multi-year PRACE Project Access “Volcanic ash hazard and forecast” (ID 2019215114). This work was granted access to the HPC/AI resources of TGCC under the allocation 2020-RA2020235114 made by GENCI., The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/feart.2022.941789/full#supplementary-material, For active volcanoes, knowledge about probabilities of eruption and impacted areas becomes valuable information for decision-makers to develop short- and long-term emergency plans, for which probabilistic volcanic hazard assessment (PVHA) is needed. High-resolution or spatially extended PVHA requires extreme-scale high-performance computing systems. Within the framework of ChEESE (Center of Excellence for Exascale in Solid Earth; www.cheese-coe.eu), an effort was made to generate exascale-suitable codes and workflows to collect and process in some hours the large amount of data that a quality PVHA requires. To this end, we created an optimized HPC-based workflow coined PVHA_HPC-WF to develop PVHA for a volcano. This tool uses the Bayesian event tree methodology to calculate eruption probabilities, vent-opening location(s), and eruptive source parameters (ESPs) based on volcano history, monitoring system data, and meteorological conditions. Then, the tool interacts with the chosen hazard model, performing a simulation for each ESP set or volcanic scenario (VS). Finally, the resulting information is processed by proof-of-concept-subjected high-performance data analytics (HPDA) scripts, producing the hazard maps which describe the probability over time of exceeding critical thresholds at each location in the investigated geographical domain. Although PVHA_HPC-WF can be adapted to other hazards, we focus here on tephra (i.e., lapilli and ash) transport and deposition. As an application, we performed PVHA for Campi Flegrei (CF), Italy, an active volcano located in one of the most densely inhabited areas in Europe and under busy air traffic routes. CF is currently in unrest, classified as being in an attention level by the Italian Civil Protection. We consider an approximate 2,000 × 2,000 × 40 km computational domain with 2 km grid resolution in the horizontal and 40 vertical levels, centered in CF. To explore the natural variability and uncertainty of the eruptive conditions, we consider a large number of VSs allowing us to include those of low probability but high impact, and simulations of tephra dispersal are performed for each of them using the FALL3D model. Results show the potential of HPC to timely execute a vast range of simulations of complex numerical models in large high-resolution computational domains and analyze great volumes of data to obtain quality hazard maps., Para los volcanes activos, el conocimiento de las probabilidades de erupción y de las áreas potencialmente afectadas constituye una información clave para los responsables de la toma de decisiones a la hora de desarrollar planes de emergencia a corto y largo plazo, lo que requiere la aplicación de evaluaciones probabilísticas del peligro volcánico (PVHA). Las evaluaciones PVHA de alta resolución o en dominios espaciales extensos demandan sistemas de computación de alto rendimiento a escala extrema. En el marco del proyecto ChEESE (Center of Excellence for Exascale in Solid Earth), se desarrollaron códigos y flujos de trabajo adaptados a entornos exascalares con el objetivo de recopilar y procesar, en pocas horas, el gran volumen de datos necesario para una PVHA de calidad. Con este fin, se creó un flujo de trabajo optimizado basado en HPC, denominado PVHA_HPC-WF, para el desarrollo de evaluaciones probabilísticas del peligro volcánico. Esta herramienta emplea la metodología del árbol de eventos bayesiano para calcular las probabilidades de erupción, la localización de los centros eruptivos y los parámetros de la fuente eruptiva, integrando información histórica del volcán, datos de los sistemas de monitoreo y condiciones meteorológicas. Posteriormente, el flujo de trabajo interactúa con el modelo de peligro seleccionado, ejecutando una simulación para cada conjunto de parámetros eruptivos o escenario volcánico. La información resultante es procesada mediante técnicas de análisis de datos de alto rendimiento, generando mapas de peligro que describen la probabilidad temporal de superar umbrales críticos en cada punto del dominio geográfico analizado. Aunque PVHA_HPC-WF es adaptable a otros tipos de peligros volcánicos, en este trabajo se centra en el transporte y la deposición de tefra (lapilli y ceniza). Como caso de aplicación, se realizó una PVHA para Campi Flegrei (Italia), un volcán activo situado en una de las zonas más densamente pobladas de Europa y bajo importantes rutas de tráfico aéreo, que actualmente se encuentra en estado de agitación y en nivel de atención según la Protección Civil italiana. Se consideró un dominio computacional aproximado de 2.000 × 2.000 × 40 km, con una resolución horizontal de 2 km y 40 niveles verticales, centrado en Campi Flegrei. Para explorar la variabilidad natural y la incertidumbre de las condiciones eruptivas, se evaluó un gran número de escenarios volcánicos, incluyendo aquellos de baja probabilidad pero alto impacto, ejecutándose para cada uno simulaciones de dispersión de tefra mediante el modelo FALL3D. Los resultados ponen de manifiesto el potencial de la computación de alto rendimiento para ejecutar de forma eficiente un amplio conjunto de simulaciones complejas en dominios extensos y de alta resolución, así como para analizar grandes volúmenes de datos y obtener mapas de peligro de alta calidad en tiempos compatibles con la toma de decisiones., European Commission’s CORDIS [ChEESE (823844)], European Community’s Horizon 2020 POP (824080), PRACE Project Access (ID 2019215114), GENCI TGCC 2020-RA2020235114
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Artículo científico (JournalArticle)

On the feasibility and usefulness of high-performance computing in probabilistic volcanic hazard assessment: An application to tephra hazard from Campi Flegrei

  • Martínez Montesinos, Beatriz
  • Titos, Manuel
  • Sandri, Laura
  • Rudyy, Oleksandr
  • Cheptsov, Alexey
  • Macedonio, Giovanni
  • Folch, Arnau
  • Barsotti, Sara
  • Selva, Jacopo
  • Costa, Antonio
For active volcanoes, knowledge about probabilities of eruption and impacted areas becomes valuable information for decision-makers to develop short- and long-term emergency plans, for which probabilistic volcanic hazard assessment (PVHA) is needed. High-resolution or spatially extended PVHA requires extreme-scale high-performance computing systems. Within the framework of ChEESE (Center of Excellence for Exascale in Solid Earth; www.cheese-coe.eu), an effort was made to generate exascale-suitable codes and workflows to collect and process in some hours the large amount of data that a quality PVHA requires. To this end, we created an optimized HPC-based workflow coined PVHA_HPC-WF to develop PVHA for a volcano. This tool uses the Bayesian event tree methodology to calculate eruption probabilities, vent-opening location(s), and eruptive source parameters (ESPs) based on volcano history, monitoring system data, and meteorological conditions. Then, the tool interacts with the chosen hazard model, performing a simulation for each ESP set or volcanic scenario (VS). Finally, the resulting information is processed by proof-of-concept-subjected high-performance data analytics (HPDA) scripts, producing the hazard maps which describe the probability over time of exceeding critical thresholds at each location in the investigated geographical domain. Although PVHA_HPC-WF can be adapted to other hazards, we focus here on tephra (i.e., lapilli and ash) transport and deposition. As an application, we performed PVHA for Campi Flegrei (CF), Italy, an active volcano located in one of the most densely inhabited areas in Europe and under busy air traffic routes. CF is currently in unrest, classified as being in an attention level by the Italian Civil Protection. We consider an approximate 2,000 × 2,000 × 40 km computational domain with 2 km grid resolution in the horizontal and 40 vertical levels, centered in CF. To explore the natural variability and uncertainty of the eruptive conditions, we consider a large number of VSs allowing us to include those of low probability but high impact, and simulations of tephra dispersal are performed for each of them using the FALL3D model. Results show the potential of HPC to timely execute a vast range of simulations of complex numerical models in large high-resolution computational domains and analyze great volumes of data to obtain quality hazard maps., The research leading to these results has been supported by the European Commission’s CORDIS [ChEESE (grant no. 823844)]. Part of this work was supported by European Community’s Horizon 2020 POP project under grant agreements (no. 824080). Multi-year PRACE Project Access “Volcanic ash hazard and forecast” (ID 2019215114). This work was granted access to the HPC/AI resources of TGCC under the allocation 2020-RA2020235114 made by GENCI., Peer reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Performance and energy consumption of HPC workloads on a cluster based on Arm ThunderX2 CPU

  • Mantovani, Filippo|||0000-0003-3559-4825
  • Garcia Gasulla, Marta|||0000-0003-3682-9905
  • Gracia, Jose
  • Stafford, Esteban
  • Banchelli Gracia, Fabio
  • Josep Fabregó, Marc
  • Criado Ledesma, Joel
  • Nachtmann, Mathias
In this paper, we analyze the performance and energy consumption of an Arm-based high-performance computing (HPC) system developed within the European project Mont-Blanc 3. This system, called Dibona, has been integrated by ATOS/Bull, and it is powered by the latest Marvell’s CPU, ThunderX2. This CPU is the same one that powers the Astra supercomputer, the first Arm-based supercomputer entering the Top500 in November 2018. We study from micro-benchmarks up to large production codes. We include an interdisciplinary evaluation of three scientific applications (a finite-element fluid dynamics code, a smoothed particle hydrodynamics code, and a lattice Boltzmann code) and the Graph 500 benchmark, focusing on parallel and energy efficiency as well as studying their scalability up to thousands of Armv8 cores. For comparison, we run the same tests on state-of-the-art x86 nodes included in Dibona and the Tier-0 supercomputer MareNostrum4. Our experiments show that the ThunderX2 has a 25% lower performance on average, mainly due to its small vector unit yet somewhat compensated by its 30% wider links between the CPU and the main memory. We found that the software ecosystem of the Armv8 architecture is comparable to the one available for Intel. Our results also show that ThunderX2 delivers similar or better energy-to-solution and scalability, proving that Arm-based chips are legitimate contenders in the market of next-generation HPC systems., The authors thank the support team of Dibona operating at ATOS/Bull. This work is partially supported by the Spanish Government through Programa Severo Ochoa (SEV-2015-0493), the Spanish Ministry of Science and Technology project (TIN2015-65316-P), the Generalitat de Catalunya (2017-SGR-1414), the European Community’s Seventh Framework Programme [FP7/2007-2013] and Horizon 2020 under the Mont-Blanc projects (grant agreements n. 288777, 610402 and 671697), the European POP2 Center of Excellence (grant agreement n. 824080), and the Human Brain Project SGA2 (grant agreement n. 785907)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Dynamic load balance of chemical source term evaluation in high-fidelity combustion simulations

  • Ramirez Miranda, Guillem
  • Mira Martínez, Daniel
  • Pérez Sánchez, Eduardo Javier|||0000-0002-5056-415X
  • Surapaneni, Anurag
  • Borrell Pol, Ricard
  • Houzeaux, Guillaume|||0000-0002-2592-1426
  • Garcia Gasulla, Marta|||0000-0003-3682-9905
This paper presents a load balancing strategy for reaction rate evaluation and chemistry integration in reacting flow simulations. The large disparity in scales during combustion introduces stiffness in the numerical integration of the PDEs and generates load imbalance during the parallel execution. The strategy is based on the use of the DLB library to redistribute the computing resources at node level, lending additional CPU-cores to higher loaded MPI processes. This approach does not require explicit data transfer and is activated automatically at runtime. Two chemistry descriptions, detailed and reduced, are evaluated on two different configurations: laminar counterflow flame and a turbulent swirl-stabilized flame. For single-node calculations, speedups of 2.3x and 7x are obtained for the detailed and reduced chemistry, respectively. Results on multi-node runs also show that DLB improves the performance of the pure-MPI code similar to single node runs. It is shown DLB can get performance improvements in both detailed and reduced chemistry calculations., This work has been supported by the Spanish Government (PID2019-107255GB), by Generalitat de Catalunya (contract 2014-SGR-1051), and by the European POP CoE (GA No 824080) and EXCELLERAT project GA No 823691., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Dynamic resource allocation for efficient parallel CFD simulations

  • Houzeaux, Guillaume|||0000-0002-2592-1426
  • Badia Sala, Rosa Maria|||0000-0003-2941-5499
  • Borrell Pol, Ricard
  • Dosimont, Damien
  • Ejarque Artigas, Jorge
  • Garcia Gasulla, Marta|||0000-0003-3682-9905
  • López Herrero, Víctor
CFD users of supercomputers usually resort to rule-of-thumb methods to select the number of subdomains (partitions) when relying on MPI-based parallelization. One common approach is to set a minimum number of elements or cells per subdomain, under which the parallel efficiency of the code is “known” to fall below a subjective level, say 80%. The situation is even worse when the user is not aware of the “good” practices for the given code and a huge amount of resources can thus be wasted. This work presents an elastic computing methodology that adapts at runtime the resources allocated to a simulation automatically. The criterion to control the required resources is based on a runtime measure of the communication efficiency of the execution. According to some analytical estimates, the resources are then expanded or reduced to fulfill this criterion and eventually execute an efficient simulation., This work has been supported by the Spanish Government (PID2019- 107255GB); by Generalitat de Catalunya (contract 2014-SGR-1051); by the European Commission H2020 project POP CoE (GA n. 824080); by the European Commission H2020 project CompBioMed CoE (GA n. 823712) and by the European Commission and the EuroHPC JU under contract 955558 (eFlows4HPC project)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

A generic performance analysis technique applied to different CFD methods for HPC

  • Garcia Gasulla, Marta|||0000-0003-3682-9905
  • Banchelli Gracia, Fabio
  • Peiró Conde, Kilian
  • Ramírez Gargallo, Guillem
  • Houzeaux, Guillaume|||0000-0002-2592-1426
  • Hassan Saïdi, Ismaïl Ben
  • Tenaud, Christian
  • Spisso, Ivan
  • Mantovani, Filippo|||0000-0003-3559-4825
For complex engineering and scientific applications, Computational Fluid Dynamics (CFD) simulations require a huge amount of computational power. As such, it is of paramount importance to carefully assess the performance of CFD codes and to study them in depth for enabling optimisation and portability. In this paper, we study three complex CFD codes, OpenFOAM, Alya and CHORUS representing two numerical methods, namely the finite volume and finite-element methods, on both structured and unstructured meshes. To all codes, we apply a generic performance analysis method based on a set of metrics helping the code developer in spotting the critical points that can potentially limit the scalability of a parallel application. We show the root cause of the performance bottlenecks studying the three applications on the MareNostrum4 supercomputer. We conclude providing hints for improving the performance and the scalability of each application., This work is partially supported by the Spanish Government through Programa Severo Ochoa (SEV-2015-0493), by the Spanish Ministry of Science and Technology (TIN2015-65316-P), by the Generalitat de Catalunya (2017-SGR-1414), and by the European POP CoE (GA n. 824080)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Hybrid parallelization of molecular dynamics simulations to reduce load imbalance

  • Morillo, Julian
  • Vassaux, Maxime
  • Coveney, Peter V.
  • Garcia Gasulla, Marta|||0000-0003-3682-9905
The most widely used technique to allow for parallel simulations in molecular dynamics is spatial domain decomposition, where the physical geometry is divided into boxes, one per processor. This technique can inherently produce computational load imbalance when either the spatial distribution of particles or the computational cost per particle is not uniform. This paper shows the benefits of using a hybrid MPI+OpenMP model to deal with this load imbalance. We consider LAMMPS (Large-scale Atomic/Molecular Massively Parallel Simulator), a prototypical molecular dynamics simulator that provides its own balancing mechanism and an OpenMP implementation for many of its modules, allowing for a hybrid setup. In this work, we extend the current OpenMP implementation of LAMMPS and optimize it and evaluate three different setups: MPI-only, MPI with the LAMMPS balance mechanism, and hybrid setup using our improved OpenMP version. This comparison is made using the five standard benchmarks included in the LAMMPS distribution plus two additional test cases. Results show that the hybrid approach can deal with load balancing problems better and more effectively (50% improvement versus MPI-only for a highly imbalanced test case) than the LAMMPS balance mechanism (only 43% improvement) and improve simulations with issues other than load imbalance., This work is partially supported by the Spanish Government through Programa Severo Ochoa (SEV-2015-0493), by the Spanish Ministry of Science and Technology (TIN2015-65316-P), by the Generalitat de Catalunya (2017-SGR-1414), and by the European POP CoE (GA n. 824080). This work is also funded as part of the European Union Horizon 2020 research and innovation program under grant agreement nos. 800925 (VECMA project; www.vecma.eu) and 823712 (CompBioMed2 Centre of Excellence; www.compbiomed.eu), as well as the UK EPSRC for the UK High-End Computing Consortium (grant no. EP/R029598/1)., Peer Reviewed
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

A generic performance analysis technique applied to different CFD methods for HPC

  • Garcia Gasulla, Marta
  • Banchelli Gracia, Fabio
  • Peiró Conde, Kilian
  • Ramírez Gargallo, Guillem
  • Houzeaux, Guillaume
  • Hassan Saïdi, Ismaïl Ben
  • Tenaud, Christian
  • Spisso, Ivan
  • Mantovani, Filippo
For complex engineering and scientific applications, Computational Fluid Dynamics (CFD) simulations require a huge amount of computational power. As such, it is of paramount importance to carefully assess the performance of CFD codes and to study them in depth for enabling optimisation and portability. In this paper, we study three complex CFD codes, OpenFOAM, Alya and CHORUS representing two numerical methods, namely the finite volume and finite-element methods, on both structured and unstructured meshes. To all codes, we apply a generic performance analysis method based on a set of metrics helping the code developer in spotting the critical points that can potentially limit the scalability of a parallel application. We show the root cause of the performance bottlenecks studying the three applications on the MareNostrum4 supercomputer. We conclude providing hints for improving the performance and the scalability of each application., This work is partially supported by the Spanish Government through Programa Severo Ochoa (SEV-2015-0493), by the Spanish Ministry of Science and Technology (TIN2015-65316-P), by the Generalitat de Catalunya (2017-SGR-1414), and by the European POP CoE (GA n. 824080)., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Hybrid parallelization of molecular dynamics simulations to reduce load imbalance

  • Morillo, Julian
  • Vassaux, Maxime
  • Coveney, Peter V.
  • Garcia Gasulla, Marta
The most widely used technique to allow for parallel simulations in molecular dynamics is spatial domain decomposition, where the physical geometry is divided into boxes, one per processor. This technique can inherently produce computational load imbalance when either the spatial distribution of particles or the computational cost per particle is not uniform. This paper shows the benefits of using a hybrid MPI+OpenMP model to deal with this load imbalance. We consider LAMMPS (Large-scale Atomic/Molecular Massively Parallel Simulator), a prototypical molecular dynamics simulator that provides its own balancing mechanism and an OpenMP implementation for many of its modules, allowing for a hybrid setup. In this work, we extend the current OpenMP implementation of LAMMPS and optimize it and evaluate three different setups: MPI-only, MPI with the LAMMPS balance mechanism, and hybrid setup using our improved OpenMP version. This comparison is made using the five standard benchmarks included in the LAMMPS distribution plus two additional test cases. Results show that the hybrid approach can deal with load balancing problems better and more effectively (50% improvement versus MPI-only for a highly imbalanced test case) than the LAMMPS balance mechanism (only 43% improvement) and improve simulations with issues other than load imbalance., This work is partially supported by the Spanish Government through Programa Severo Ochoa (SEV-2015-0493), by the Spanish Ministry of Science and Technology (TIN2015-65316-P), by the Generalitat de Catalunya (2017-SGR-1414), and by the European POP CoE (GA n. 824080). This work is also funded as part of the European Union Horizon 2020 research and innovation program under grant agreement nos. 800925 (VECMA project; www.vecma.eu) and 823712 (CompBioMed2 Centre of Excellence; www.compbiomed.eu), as well as the UK EPSRC for the UK High-End Computing Consortium (grant no. EP/R029598/1)., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Performance and energy consumption of HPC workloads on a cluster based on Arm ThunderX2 CPU

  • Mantovani, Filippo
  • Garcia Gasulla, Marta
  • Gracia, Jose
  • Stafford, Esteban
  • Banchelli Gracia, Fabio
  • Josep Fabregó, Marc
  • Criado Ledesma, Joel
  • Nachtmann, Mathias
In this paper, we analyze the performance and energy consumption of an Arm-based high-performance computing (HPC) system developed within the European project Mont-Blanc 3. This system, called Dibona, has been integrated by ATOS/Bull, and it is powered by the latest Marvell’s CPU, ThunderX2. This CPU is the same one that powers the Astra supercomputer, the first Arm-based supercomputer entering the Top500 in November 2018. We study from micro-benchmarks up to large production codes. We include an interdisciplinary evaluation of three scientific applications (a finite-element fluid dynamics code, a smoothed particle hydrodynamics code, and a lattice Boltzmann code) and the Graph 500 benchmark, focusing on parallel and energy efficiency as well as studying their scalability up to thousands of Armv8 cores. For comparison, we run the same tests on state-of-the-art x86 nodes included in Dibona and the Tier-0 supercomputer MareNostrum4. Our experiments show that the ThunderX2 has a 25% lower performance on average, mainly due to its small vector unit yet somewhat compensated by its 30% wider links between the CPU and the main memory. We found that the software ecosystem of the Armv8 architecture is comparable to the one available for Intel. Our results also show that ThunderX2 delivers similar or better energy-to-solution and scalability, proving that Arm-based chips are legitimate contenders in the market of next-generation HPC systems., The authors thank the support team of Dibona operating at ATOS/Bull. This work is partially supported by the Spanish Government through Programa Severo Ochoa (SEV-2015-0493), the Spanish Ministry of Science and Technology project (TIN2015-65316-P), the Generalitat de Catalunya (2017-SGR-1414), the European Community’s Seventh Framework Programme [FP7/2007-2013] and Horizon 2020 under the Mont-Blanc projects (grant agreements n. 288777, 610402 and 671697), the European POP2 Center of Excellence (grant agreement n. 824080), and the Human Brain Project SGA2 (grant agreement n. 785907)., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Dynamic resource allocation for efficient parallel CFD simulations

  • Houzeaux, Guillaume
  • Badia Sala, Rosa Maria
  • Borrell Pol, Ricard
  • Dosimont, Damien
  • Ejarque Artigas, Jorge
  • Garcia Gasulla, Marta
  • López Herrero, Víctor
CFD users of supercomputers usually resort to rule-of-thumb methods to select the number of subdomains (partitions) when relying on MPI-based parallelization. One common approach is to set a minimum number of elements or cells per subdomain, under which the parallel efficiency of the code is “known” to fall below a subjective level, say 80%. The situation is even worse when the user is not aware of the “good” practices for the given code and a huge amount of resources can thus be wasted. This work presents an elastic computing methodology that adapts at runtime the resources allocated to a simulation automatically. The criterion to control the required resources is based on a runtime measure of the communication efficiency of the execution. According to some analytical estimates, the resources are then expanded or reduced to fulfill this criterion and eventually execute an efficient simulation., This work has been supported by the Spanish Government (PID2019- 107255GB); by Generalitat de Catalunya (contract 2014-SGR-1051); by the European Commission H2020 project POP CoE (GA n. 824080); by the European Commission H2020 project CompBioMed CoE (GA n. 823712) and by the European Commission and the EuroHPC JU under contract 955558 (eFlows4HPC project)., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Dynamic load balance of chemical source term evaluation in high-fidelity combustion simulations

  • Ramirez Miranda, Guillem
  • Mira Martínez, Daniel
  • Pérez Sánchez, Eduardo Javier
  • Surapaneni, Anurag
  • Borrell Pol, Ricard
  • Houzeaux, Guillaume
  • Garcia Gasulla, Marta
This paper presents a load balancing strategy for reaction rate evaluation and chemistry integration in reacting flow simulations. The large disparity in scales during combustion introduces stiffness in the numerical integration of the PDEs and generates load imbalance during the parallel execution. The strategy is based on the use of the DLB library to redistribute the computing resources at node level, lending additional CPU-cores to higher loaded MPI processes. This approach does not require explicit data transfer and is activated automatically at runtime. Two chemistry descriptions, detailed and reduced, are evaluated on two different configurations: laminar counterflow flame and a turbulent swirl-stabilized flame. For single-node calculations, speedups of 2.3x and 7x are obtained for the detailed and reduced chemistry, respectively. Results on multi-node runs also show that DLB improves the performance of the pure-MPI code similar to single node runs. It is shown DLB can get performance improvements in both detailed and reduced chemistry calculations., This work has been supported by the Spanish Government (PID2019-107255GB), by Generalitat de Catalunya (contract 2014-SGR-1051), and by the European POP CoE (GA No 824080) and EXCELLERAT project GA No 823691., Peer Reviewed, Postprint (author's final draft)