Internal Training events

From cheese2

2026

Workshop - Rethinking scientific applications for exascale and emerging architectures - the Centres of Excellence challenge

  • Date: 27/01/2026
  • Time: 10:00 AM - 05:30 PM
  • Location: Krakow, Poland (HIPEAC).
  • Registration: here
  • Description: The workshop will focus on the challenges of adapting or rewriting legacy codes for exascale and emerging heterogeneous architectures. Participants will share experiences in developing and modernising algorithms for next-generation platforms. Key strategies and potential solutions to improve performance will be discussed.
  • More information: here


Joint SeisSol and ExaHyPE User Workshop, Training and Hackathon

  • Date: 11-12/06/2026
  • Time: 08:30 AM
  • Location: Leibniz Supercomputing Centre (LRZ), Boltzmannstr. 1, 85748 Garching bei München, Germany.
  • Registration: here
  • Description: In celebration of 20 years of SeisSol and 10 years of ExaHyPE, we are happy to announce: the first Joint SeisSol+ExaHyPE User Workshop; a combined SeisSol+ExaHyPE User Training; the first SeisSol+ExaHyPE Hackathon.
  • More information: here

2025

Computational Methods in Geophysics

  • Date: 31/10/2025
  • Time: 9 AM
  • Location: Barcelona Supercomputing Center, Spain
  • Registration: here
  • Description: This course aims at training students in numerical methods applied to solve partial differential equations in geophysics. Today many researchers use community codes (e.g. developed by CHEESE) often with little training in the underlying mathematical methods. The goal is to provide a basic understanding of a variety of numerical methods used in geophysics today. While we use the elastic (acoustic) wave equation (mostly in 1D) to illustrate the concepts, the knowledge can be easily transferred to other systems in geophysics. The course is “practical” in the sense that we provide the link from the mathematical algorithm to the actual code in Python using Jupyter notebooks. The course format will be lectures interspersed (time-permitting) with simple theoretical exercises, comprehension questions, and Jupyter notebook exercises. The methods discussed will be the finite-difference method, the pseudospectral method, linear and spectral elements (and - if there is interest - the finite volume and the discontinuous Galerkin method). Note that the purpose is NOT to discuss any specific community codes, rather to provide information on what is under the hood and how to avoid errors and - for example - which methods are favorable for complex, geometries, topography, problems with difficult boundary conditions, parallel computers, etc.
  • More information: here

Internal ChEESE Webinar: TALP: Tracking Application Live Performance

  • Date: 24/04/2025
  • Time: 2 PM
  • Location: Online
  • Registration: here
  • Description: In this talk, Marta García-Gasulla, from BSC, will introduce TALP (Tracking Application Live Performance), a module within the DLB library (https://pm.bsc.es/dlb). TALP is a lightweight tool designed to measure the performance efficiency of parallel applications at runtime. As a low-overhead profiling tool, TALP enables the retrieval of performance metrics at the end of execution, as well as the ability to query them during runtime. We will explore its various working modes, from fully transparent operation requiring no code modifications to an advanced mode where a custom API allows users to annotate specific code regions for detailed performance metrics. Additionally, we will demonstrate how TALP can be seamlessly integrated with CI/CD platforms such as GitLab, enabling continuous performance monitoring and ensuring application efficiency throughout the development lifecycle.
  • Watch online

Internal ChEESE Webinar: Overview of the European Processor Initiative (EPI) – RISC-V Vector CPUs for High-Performance Computing

  • Date: 13/03/2025
  • Description: This internal ChEESE seminar aims to provide a comprehensive overview of the EPI project, an introduction to RISC-V, and insights into vector supercomputing. The European Processor Initiative (EPI) is a project dedicated to developing a general-purpose processor and an accelerator, alongside the necessary software layers for their integration into the High Performance Computing (HPC) ecosystem. The Barcelona Supercomputing Center is contributing to the development of a RISC-V-based accelerator targeted at HPC applications, leveraging the RISC-V vector extension.
This talk aims to provide a comprehensive overview of the EPI project, an introduction to RISC-V, and insights into vector supercomputing. Special emphasis will be placed on the RISC-V vector extensions (RVV), with a particular focus on implementations utilizing large vectors. Participants will gain an understanding of how RVV compares with other vector architectures and explore a design approach that utilizes vectors up to 16-kbit wide. Ultimately, the talk aims to present the methodologies, tools, and libraries available for vectorization, while addressing the accompanying challenges and limitations.
  • Speaker: Dr. Filippo Mantovani, Barcelona Supercomputing Center
  • Watch online

Computational Seismology and Underwater Acoustics: the SPECFEM code

  • Date: 22-23/04/2025
  • Organisers: ChEESE-2P / NuTS / SEASOUNDS
  • Location: Faculté des sciences et des techniques Nantes Université, LPG Building 4, Chem. de la Houssinière, 44322 Nantes, France. [See Building 4]
  • Cost: No registration fee
  • Registration closed: 29 January, 16:00h
  • Programme: read
  • Description: This course will focus on the SPECFEM code, which uses a spectral-element method in the time domain for full-wave propagation modeling. It covers both 2D and 3D versions of SPECFEM, including setup, meshing, and HPC integration. Hands-on sessions and case studies will be provided.
  • More information: here

Workshop: Tackling Software Exascale Challenges: The CoEs in HPC Perspective

  • Date: 22/01/2025
  • Organisers: EoCoE, ChEESE-2P, ESiWACE, POP, MAX, EXCELLERAT P2, MultiScale, Space CoE
  • Location: Barcelona, Spain
  • Programme: read
  • Description: At the HiPEAC25 Conference, the European Centres of Excellence (CoEs) will showcase two workshops addressing the technical challenges of transitioning to exascale computing, while also highlighting the contributions of women researchers and engineers in the field.
  • More information: here

Workshop: From Petascale to Exascale and Beyond: The Centres of Excellence Challenge

  • Date: 20/01/2025
  • Organisers: EoCoE, ChEESE-2P, ESiWACE, POP, MAX, EXCELLERAT P2, MultiScale, Space CoE
  • Location: Barcelona, Spain
  • Programme: read
  • Description: At the HiPEAC25 Conference, the European Centres of Excellence (CoEs) will showcase two workshops addressing the technical challenges of transitioning to exascale computing, while also highlighting the contributions of women researchers and engineers in the field.
  • More information: here

2024

Workshop: How European projects are developing HPC-based software and services to address natural hazards

  • Date: 10/12/2024
  • Location: Online
  • Description: ChEESE CoE, in collaboration with NCC Spain and CASTIEL2 WP3, hosted this virtual workshop to explore the role of HPC in natural hazard management. The event featured experts discussing applications of supercomputing in disaster preparedness, covering topics like volcanic eruptions, floods, wildfires, and earthquakes.
  • More information: [1]
  • Watch on YouTube: here

Artificial Intelligence and Machine Learning for Geosciences

  • Date: 5-7/11/2024
  • Location: Delegació CSIC Catalunya, Carrer de les Egipcíaques, 15, Ciutat Vella, Barcelona
  • Cost: No registration fee
  • Registration closed
  • Programme: here
  • Description: This course introduced AI tools, covering topics from linear regression to deep neural networks. Participants explored classification and regression problems using hybrid CPU-GPU architectures through hands-on tutorials.
  • More information: here

Tsunami and meteotsunami modelling

  • Date: 13-14/05/2024
  • Time: 15:00 – 18:00 CET
  • Registration closed
  • Location: Online
  • Programme: here
  • Description: This course provided an introduction to tsunami and meteotsunami modeling using the Tsunami-HySEA and Meteo-HySEA codes, covering theoretical aspects, numerical simulations, and Python-based pre/post-processing.
  • Watch on YouTube:
* Part 1: here * Part 2: here

Open Webinar: Good Science in EuroHPC Centers of Excellence – Incorporating Gender Analysis in Research

  • Date: 29/04/2024
  • Time: 11:00 CEST
  • Location: Online
  • Registration: Closed
  • Description: This webinar, led by Dr. Débora Zurro, explored the integration of gender perspectives in HPC research within EuroHPC Centers of Excellence.
  • More information: here

Computational Geodynamics – Modeling Geological Processes using European Flagship Codes

  • Date: 25-28/03/2024
  • Location: University of Mainz, Germany
  • Cost: No registration fee (travel, accommodation, and meals not covered)
  • Registration closed
  • Programme: here
  • Description: Participants gained hands-on experience with LaMEM and pTatin3D for simulating geological processes like subduction and rifting, using Julia for model setups on HPC clusters.
  • More information: here

Computational Volcanology – Modeling Volcanic Eruptions using European Flagship Codes

  • Date: 11-14/03/2024
  • Location: Barcelona Supercomputing Center
  • Programme: here
  • Description: This course covered OpenPDAC and FALL3D, providing hands-on training in volcanic flow and ash dispersion simulations.
  • More information: here

2023

EOSC-synergy tools for SQaaS

  • Date: 24/03/2023
  • Time: 10:00 – 12:00 CET
  • Zoom link: here (Access code: 513936)
  • Recording and presentation: here

High-Performance Seismology CyberTraining

  • Date: 9-12/05/2023
  • Time: 08:45 – 17:00 CET
  • Description: With the rise of data volume and computing power, seismological research requires more advanced skills in data processing, numerical methods, and parallel computing. This training aims to create entry points easing access to high-performance computing and data management for cutting-edge scientific analysis for the seismological community, including training in best practices in data management, software development, and open science. The learning goals of the workshop are: To deploy software on the Cloud Systems and clusters from High-Performance Computing Centers; To adopt software best practices for large-scale computing (e.g., from a script to a full-stack containerized software); To use and re-use tutorials on topical research that require large-scale processing (e.g., ambient noise cross-correlation, wavefield simulations, etc.)
  • More info: here

11th EGU GALILEO CONFERENCE

  • Date: 23-26/05/2023
  • Time: 08:45 – 17:00 CET
  • Description: With the birth of geosciences and then of computational geosciences, people's attitude towards the world has changed. People are no longer passive observers but rather more proactive players in their environment and the knowledge gained from computational geosciences has been used to save lives and prepare for future scenarios. Now, with the Exascale computing era dawning upon us, even more accurate, clearer and faster data will be within our reach. However, computational methods and codes need to be reshaped and honed, and geoscientists need to be prepared for the upcoming challenges. The geoscience community is multidisciplinary and transversal, encompassing Earth observatories (e.g. monitoring networks, data providers, tsunami warning centres, etc), research and academia, model developers and, finally, model end-users and social agents. From a scientific point of view, the entire community is facing scientific breakthrough problems for which researchers need to share their knowledge and experience and make roadmap recommendations in order to contribute to the advancement of this rapidly growing field. The “Solid Earth and Geohazards in the Exascale Era” conference would bring together some of the world's best minds in various branches of computational geosciences to jointly tackle challenges and issues of Exascale computing. Europe is making a huge effort in the strategic global race to Exascale, with large investments in the infrastructure and the application pillars (e.g. Centers of Excellence and national Competence Centers). In 2023 the EuroHPC Joint Undertaking will have deployed 3 Pre-exascale machines (LUMI, Leonardo and MareNostrum-V), and 2 Exascale systems will follow afterwards. From the point of view of science and services, this unprecedented pan-European infrastructure opens a myriad of possibilities but, at the same time, the transition is challenging and requires a joint effort from (geo)scientists, software developers, and data analysts.
  • More info: here


Porting to GPU a Fortran code using intrinsic parallelism

  • Date: 13/06/2023
  • Online
  • Instructor: Giorgio Amati
  • Description: This webinar delve into the world of Fortran intrinsic parallelism and its application to GPU programming. Fortran, a powerful programming language for scientific and engineering computations, offers built-in mechanisms for parallelism, enabling efficient utilization of GPUs. In this session, we will explore the key features, challenges, and benefits of porting Fortran code to GPU architectures.
  • Agenda: Porting to GPU: Demonstrating the conversion of a simple kernel to exploit GPU parallelism · A More Complex Code: Exploring the application of Fortran intrinsic parallelism to a 2D Kinetic CFD code · Porting to GPU: Tackling the challenges of optimizing a 3D CFD code using Fortran intrinsic parallelism.
  • Recording: here

OpenCL GPU Programming for HPC Applications

  • Date: 28/06/2023
  • Instructor: Dr. Moritz Lehmann
  • Description: This workshop introduced OpenCL for GPU computing, covering performance optimization techniques and multi-GPU parallelization.
  • Recording and presentation: here
  • YouTube video: here

Kokkos – A Portable Programming Library for the Exascale Era

  • Date: 07/07/2023
  • Instructor: Rohit Kakodkar (Princeton University)
  • Description: Organized by the Center of Excellence ChEESE, this webinar offers a brief overview of the Kokkos C++ portable programming library. Its purpose is to demonstrate how Kokkos enables the creation of performance-portable code that can be executed on CPUs and GPUs. The webinar focuses on highlighting effective use cases for the Kokkos programming model. Additionally, the webinar shares insights into the experience of porting SPECFEM2D using Kokkos.
  • Presentation: here
  • Recording: here

Device Accelerated Solvers with PETSc

  • Date: 20/12/2023
  • Instructor: Stefano Zampini
  • Description: This talk covered recent developments in GPU-based PDE solvers using PETSc.
  • Registration: here
Retrieved from "https://cheese2.eu/wiki/index.php?title=Internal_Training_events&oldid=2170"