47th CERN School of Computing - Latin America 2026

America/Santiago
Lo Contador Campus

Lo Contador Campus

Santiago, Chile
Alberto Pace (CERN), Andrzej Nowicki (CERN), Kristina Gunne (CERN)
Description

Welcome to the 47th School of Computing,
the
1st CERN School of Computing in Latin America !

The school will take place from 11-24th of January 2026 in Santiago, Chile!  This year's school is organized in collaboration with the SAPHIR Millenium Institute and will be hosted at the Lo Contador Campus of Pontificia Universidad Católica de Chile. This event is held with the support of the Ministry of Science, Technology, Knowledge and Innovation of Chile.

The deadline for applications has been extended to October 15th!

Academic Programme

The two-week programme will consist around 50 hours of lectures and hands-on exercises, covering three main themes:

  • Physics computing
  • Software engineering
  • Data technologies. 

Students who pass the final optional exam will receive a diploma from CERN and the School.

Other activities

It's not all study; the social and sport programme is also a vital part of the School. We will have ample opportunities to explore and experience some of the great cultural, historical and natural attractions of Santiago and its surroundings! Even more, it is a golden opportunity to build relations with peers around the world, which we hope will last for a long time also after the school.  

Important dates

  • Wednesday 23 July - applications open
  • Wednesday Oct 15 (midnight UTC+2 / CEST) - deadline for applications
  • Friday Oct 17 - invitations sent to the selected participants
  • Wednesday 19th of November - registration fee payment deadline
  • Sunday 11 January - arrival of students in Santiago
  • Saturday 24 January (morning) - departure

What is the CERN School of Computing?

The CERN School of computing, created in 1970, has been training thousands of worldwide students in Scientific computing architectures and Data Science.

Its mission is to create a common scientific culture in computing among young scientists and engineers involved in particle physics or other sciences, as a strategic direction to promote mobility and to facilitate the development of large computing-oriented transnational projects.

 The school bridges the gap between  Science and Computing by ensuring that:

  • Scientists master computing technologies as the main tool for their research
  • Computer scientists understand the scientific domain of the investigation to deliver computing services that meet the needs of the research project

Who can apply?

The School is aimed at postgraduate (ie. minimum of Bachelor's degree or equivalent) students, engineers, and scientists with a few years of experience in particle physics, computing, or in related fields. We particularly welcome applications from scientists and engineers affiliated with CERN or institutions involved in CERN experiments. We welcome applications from all countries and nationalities, especially from students in Latin America.

Applicants are responsible for ensuring that their home institute or employer covers their registration fee and travel cost, or, failing this, themselves. 

Talented students with proven financial difficulties to have their fee paid by their institute can apply for a partial sponsoring of the registration fee. This must be asked when applying to the school.

Please note that the language of this school is English.

CERN School of Computing
    • 3:00 PM 6:00 PM
      Arrival and registration of participants 3h
    • 6:00 PM 7:30 PM
      Self presentations 1h 30m
    • 7:30 PM 8:30 PM
      Welcome dinner RQ Hotel 1h RQ Hotel

      RQ Hotel

    • 9:00 AM 10:30 AM
      Welcome and Opening Ceremony 1h 30m
      • Opening remarks by Felipe Olivares León, Executive Director of SAPHIR and local organizer of the school (5 mins).

      • Remarks by Magdalena Vicuña, Dean of the Faculty of Architecture, Design, and Urban Studies (5 mins).

      • Remote Welcome from Enrica Porcari, CERN Chief information Officer (5 mins).

      • Short Presentation of CERN and its research by Alberto Pace, Director, CERN School of Computing (15 min).

      • Remarks by Ignacio Silva, Head of the Emerging Technologies Division of the Ministry of Science (5 mins).

      • Remarks by María Montt Strabucchi, Vice-Rector for International Affairs at the Pontifical Catholic University (5 mins).

      • SAPHIR presentation by Francisca Garay, Deputy Director of SAPHIR (20 min).

      • Presentation of the CERN School of Computing by Alberto Pace, School Director (20 min).

      Speakers: Alberto Pace (CERN), Enrica Maria Porcari (CERN), Felipe Andres Olivares Leon (Millennium Institute for Subatomic Physics at High Energy Frontier (CL)), Francisca Garay Walls (Pontifical Catholic University of Chile (CL))
    • 10:30 AM 10:45 AM
      Spare 15m
    • 10:45 AM 11:15 AM
      Break 30m
    • 11:15 AM 11:30 AM
      Announcements 15m
    • 11:30 AM 12:30 PM
      Introduction to Physics L1 1h
      Speaker: Helena Brandao Malbouisson (Universidade do Estado do Rio de Janeiro (BR))
    • 12:30 PM 1:30 PM
    • 1:30 PM 2:30 PM
      Lunch 1h
    • 2:30 PM 3:30 PM
      Tools and Techniques L2 1h
      Speaker: Arturo Sanchez Pineda
    • 3:30 PM 4:00 PM
      Coffee break 30m
    • 4:00 PM 5:00 PM
      Setup of exercise environments 1h
      Speaker: Andrzej Nowicki (CERN)
    • 5:00 PM 6:00 PM
      Tools and Techniques E1 1h
      Speaker: Arturo Sanchez Pineda
    • 6:00 PM 7:00 PM
      Tools and Techniques E2 1h
      Speaker: Arturo Sanchez Pineda
    • 7:00 PM 7:25 PM
      Transport by bus to Retaurant Mestizo from RQ Hotel 25m
    • 7:30 PM 10:30 PM
      Opening dinner at Restaurant Mestizo 3h

      https://mestizorestaurant.cl/

    • 8:45 AM 9:45 AM
      Introduction to Physics L2 1h
      Speaker: Helena Brandao Malbouisson (Universidade do Estado do Rio de Janeiro (BR))
    • 9:45 AM 10:45 AM
      Data Science L1 1h

      Data Science and Interactive data exploration

      Speaker: Leonid Serkin (ICN-UNAM (MX))
    • 10:45 AM 11:00 AM
      Announcements 15m
    • 11:00 AM 11:30 AM
      Coffee break 30m
    • 11:30 AM 12:30 PM
      Software design L1 1h

      Software design in the many core era (parallel and GPUs architectures)

      Speaker: Arturo Sanchez Pineda
    • 12:30 PM 1:30 PM
      Data Science L2 1h

      Data Science and Interactive data exploration

      Speaker: Leonid Serkin (ICN-UNAM (MX))
    • 1:30 PM 2:30 PM
      Lunch 1h
    • 2:30 PM 3:30 PM
      Data Science E1 1h

      Data Science and Interactive data exploration

      Speaker: Leonid Serkin (ICN-UNAM (MX))
    • 3:30 PM 4:30 PM
      Data Science E2 1h

      Data Science and Interactive data exploration

      Speaker: Leonid Serkin (ICN-UNAM (MX))
    • 4:30 PM 5:00 PM
      Photo and Coffee break 30m
    • 5:00 PM 7:30 PM
      Study time or sports 2h 30m
    • 7:30 PM 8:30 PM
      Dinner RQ Hotel 1h RQ Hotel

      RQ Hotel

    • 8:30 PM 10:00 PM
      Pub quiz 1h 30m

      An informal pub quiz, to be held after dinner at the RQ hotel.

    • 8:45 AM 9:45 AM
      Data Analysis L1 1h
      Speaker: Toni Sculac (University of Split Faculty of Science (HR))
    • 9:45 AM 10:45 AM
      Software design L2 1h

      Software design in the many core era (parallel and GPUs architectures)

      Speaker: Arturo Sanchez Pineda
    • 10:45 AM 11:00 AM
      Announcements 15m
    • 11:00 AM 11:30 AM
      Coffee break 30m
    • 11:30 AM 12:30 PM
      Databases L1 1h
      Speaker: Andrzej Nowicki (CERN)
    • 12:30 PM 1:30 PM
      Tools and Techniques L3 1h
    • 1:30 PM 2:30 PM
      Lunch 1h
    • 2:30 PM 3:30 PM
      Software design E1 1h

      Software design in the many core era (parallel and GPUs architectures)

      Speaker: Arturo Sanchez Pineda
    • 3:30 PM 4:30 PM
      Software design E2 1h

      Software design in the many core era (parallel and GPUs architectures)

      Speaker: Arturo Sanchez Pineda
    • 4:30 PM 5:00 PM
      Coffee break 30m
    • 5:00 PM 7:30 PM
      Study time or sports 2h 30m
    • 7:30 PM 8:30 PM
      Dinner RQ Hotel 1h RQ Hotel

      RQ Hotel

    • 8:45 AM 9:45 AM
      Databases L2 1h
      Speaker: Andrzej Nowicki (CERN)
    • 9:45 AM 10:45 AM
      Data Analysis L2 1h
      Speaker: Toni Sculac (University of Split Faculty of Science (HR))
    • 10:45 AM 11:00 AM
      Announcements 15m
    • 11:00 AM 11:30 AM
      Coffee break 30m
    • 11:30 AM 12:30 PM
      Software design L3 1h

      Software design in the many core era (parallel and GPUs architectures)

      Speaker: Arturo Sanchez Pineda
    • 12:30 PM 1:30 PM
      Lunch 1h
    • 1:30 PM 7:30 PM
      Guided tour Santiago 6h

      During this tour, we will visit some famous places in Santiago such as Plaza de Armas and Plaza Ciudadania. The tour will end with a visit to the Museum of Human rights.

    • 7:30 PM 8:30 PM
      Dinner RQ Hotel 1h RQ Hotel

      RQ Hotel

    • 8:00 AM 10:00 AM
      Transport by bus from RQ Hotel 2h
    • 10:00 AM 5:00 PM
      Full day excursion to the Tricao Parc 7h

      https://www.tricao.cl/

    • 5:00 PM 7:00 PM
      Transport by bus to RQ Hotel 2h
    • 7:30 PM 8:30 PM
      Dinner RQ Hotel 1h RQ Hotel

      RQ Hotel

    • 8:45 AM 9:45 AM
      Montecarlo simulation L1 1h

      Advanced Montecarlo simulation

      Speaker: Jaime Alfonso Romero Barrientos (CCHEN Comision Chilena de Energia Nuclear (CL))
    • 9:45 AM 10:45 AM
      Machine Learning L2 1h
      Speaker: Andre Sznajder (Universidade do Estado do Rio de Janeiro (BR))
    • 10:45 AM 11:00 AM
      Announcements 15m
    • 11:00 AM 11:30 AM
      Coffee break 30m
    • 11:30 AM 12:30 PM
      Data Technologies L3 1h
      Speaker: Alberto Pace (CERN)
    • 12:30 PM 1:30 PM
      Montecarlo simulation L2 1h

      Advanced Montecarlo simulation

      Speaker: Jaime Alfonso Romero Barrientos (CCHEN Comision Chilena de Energia Nuclear (CL))
    • 1:30 PM 2:30 PM
      Lunch 1h
    • 2:30 PM 3:30 PM
      Montecarlo simulation E1 1h

      Advanced Montecarlo simulation

      Speaker: Jaime Alfonso Romero Barrientos (CCHEN Comision Chilena de Energia Nuclear (CL))
    • 3:30 PM 4:30 PM
      Montecarlo simulation E2 1h

      Advanced Montecarlo simulation

      Speaker: Jaime Alfonso Romero Barrientos (CCHEN Comision Chilena de Energia Nuclear (CL))
    • 4:30 PM 5:00 PM
      Coffee break 30m
    • 5:00 PM 7:30 PM
      Study time or sports 2h 30m
    • 7:30 PM 8:30 PM
      Dinner RQ Hotel 1h RQ Hotel

      RQ Hotel

    • 8:45 AM 9:45 AM
      Lightning talks 1h
      • HAL: A local HPC analysis facility environment for LHC data 6m

        To support the computational demands to analyse large datasets, in this work we detail the assembly and configuration of a local high-performance computing (HPC) infrastructure. We also describe the adaptation of data processing frameworks for compatibility with the Worldwide LHC Computing Grid (WLCG), enabling distributed training and inference workflows in customized cluster hosted at the São Paulo Research and Analysis Center (SPRACE). Strategies for managing and analyzing datasets of tens of terabytes are presented, including parallel I/O, memory-efficient data formats, and scalable preprocessing pipelines. A significant portion of the study is dedicated to the role of Python as a primary programming language in high-energy physics, particularly through the use of columnar data processing libraries like Awkward Array and Uproot wrapped in the Pepper framework developed by Compact Muon Solenoid (CMS) collaboration. These tools enable fast, vectorized analysis of ROOT files, facilitating physics insights from LHC Run 3 data. Overall, we highlight the synergy between modern ML, scalable computing infrastructure, and efficient data analysis tools, providing a path forward for the next generation of discoveries in particle physics.

        Speakers: Dr Marcio De Sousa Mateus Junior (SPRACE - UNESP), Márcio de Sousa Mateus Júnior (SPRACE - UNESP)
      • Feasibility Study of Measurement of $\eta_c \rightarrow p\bar{p}$ at SPD Experiment 6m

        This preliminary study examines the possibility of obtaining measurements $\eta_c \rightarrow p\bar{p}$, from proton-proton collisions at energy $\sqrt{s}= 27$ GeV in the context of Spin Physics Detector, SPD, at its final stage. By employing Monte Carlo simulations alongside the spdroot framework we propose the analysis between background and signal decay channels for $\eta_c \rightarrow p\bar{p}$ studying kinematic parameters; invariant mass, transverse momentum, polar angle distributions, momentum distribution, Feynman-x spectra, and cosine of the angle $\alpha$ between $p$ and $\bar{p}$. Previous studies, with only Pythia8 event generator, suggest signal/background ratio of $10^{-3}$. In this preliminary study, we conducted a realistic simulation considering both, the event generation and detector response of the SPD. Future work will expand upon these findings using the lxui cluster from JINR-MLIT.

        Speaker: Jesus Andres Duran Gonzales (Federal University of Rio de Janeiro (BR))
      • The SHiP experiment at the ECN3 high-intensity beam facility at the CERN SPS 6m

        The BDF/SHiP experiment is a general purpose intensity-frontier experiment for the search of feebly interacting GeV-scale particles and to perform neutrino physics measurements at the HI-ECN3 (high-intensity) beam facility at the CERN SPS, operated in beam-dump mode, taking full advantage of the available 4x10^19 protons per year at 400 GeV. CERN recently decided in favour of BDF/SHiP for the future programme of this facility. The setup consists of two complementary detector systems downstream an active muon shield. The forrmer, the scattering and neutrino detector (SND), consists of a light dark matter (LDM) / neutrino target with vertexing capability. The latter, the hidden sector decay spectrometer (HSDS), consists of a 50 m long decay volume followed by a spectrometer, timing detector, and a PID system. BDF/SHiP offers an unprecedented sensitivity to decay and scattering signatures of various new physics models and tau neutrino physics.

        Speaker: Vasilisa Guliaeva (EPFL - Ecole Polytechnique Federale Lausanne (CH))
      • Machine Learning studies to search for Dark Matter with CMS 6m
        Speaker: Thiago Henrique De Sousa (Universidade do Estado do Rio de Janeiro (BR))
      • Cosmic-Ray Neutron Spectroscopy across Chile: A Data-Driven Analysis Workflow 6m

        Cosmic-ray–induced neutrons at ground level provide valuable information about atmospheric and environmental processes. This lightning talk presents a data-driven workflow developed to measure neutron energy spectra and correlate them with local environmental variables across different sites in Chile. The analysis combines experimental data acquired with a custom high energy response neutron spectrometer and environmental measurements from a portable weather station. Both systems operate concurrently during field measurements, and their datasets are temporally aligned during post-processing to enable joint analysis of heterogeneous data streams with different sampling rates and structures. Pulse shape–based filtering techniques are applied to digitized detector signals to identify neutron-like events. Neutron energy spectra are reconstructed using unfolding algorithms that integrate experimental measurements with detector response functions computed through Monte Carlo simulations using the Geant4 toolkit. The full analysis pipeline is implemented in ROOT, leveraging the columnar and declarative model of RDataFrame to efficiently process, filter, and correlate multi-source data.

        Speaker: Franco López-Usquiano (Chilean Nuclear Energy Commission (CCHEN), Millenium Institute for Subatomic Physics at High-Energy Frontier (SAPHIR), Andrés Bello University (UNAB))
      • EOSHPM Internals: Managing Multi-PB Storage for CERNBox 6m

        As a Storage System Administrator at CERN, I manage large-scale storage infrastructure that supports the CERN community and the data-intensive workflows of the LHC experiments. Central to this ecosystem is EOS, CERN’s open-source, multi-petabyte storage system powering CERNBox, the organization’s cloud collaboration platform providing 1 TB to every user and up to 100 TB on demand. Today, CERNBox manages more than 4.1 billion files and 12 PB of data for over 27,000 users worldwide. In this presentation, I will introduce EOSHPM—the underlying EOS instances that serve CERNBox’s home, project, and media spaces. I will walk through their architecture, performance characteristics, operational experience, and how they enable scalable, reliable storage for global scientific collaboration.

        Speaker: Pablo Medina Ramos
      • Mütüä, An Easy Access LLM and Retrieval Augmented Generation Pipeline using HPC platforms 6m

        The use of large language models (LLMs) has increased in the last years. These models have shown great capabilities in natural language processing (NLP) tasks useful for researchers, like text summarization, code generation and translation. Nonetheless, these models can struggle due to the lack of context for generation. To go over this limitation, new techniques like retrieval augmented generation (RAG) have been used. Another tool greatly used by researchers are High Performance Computing (HPC) platforms. HPC platforms hold great computing capabilities, enough to run LLM applications. In order to fully utilize these resources, we introduce Mütüä, the RAG LLM pipeline for HPC platforms. Mütüä provides an easy-access platform through the HPC web-based interface Open OnDemand, to provide accessibility to the RAG technique even for non-technical HPC platform users. The Mütüä pipeline also manages resources through the Slurm software, to make use of the HPC accelerators in a resource-efficient and user transparent way.

        Speaker: Luis Carlos Navarro Todd
      • The Pierre Auger Observatory 6m

        The Pierre Auger Observatory in Malargüe, Argentina, is the world's largest experiment studying ultra-high-energy cosmic rays through the extensive lightning showers they generate. It uses a hybrid design: a 3,000 km² array of water Cherenkov detectors on the surface, complemented by fluorescence telescopes (and denser infill arrays) to measure the shower particles on the ground and their development in the atmosphere. This combination allows for accurate reconstruction and key results on the energy spectrum and arrival directions at the highest energies, thus constraining the sources and propagation of cosmic rays.

        Speaker: Jose Ochoa
      • Chronicles: An Open-Source, Decentralized Phonebook for Meaningful Relationships 6m

        Chronicles is an open-source project I developed out of a long-standing interest in geography, people, and how relationships evolve over time. Rather than acting as a traditional contact list, Chronicles is a personal, decentralized “people database” designed to help maintain meaningful connections with friends, family, and acquaintances. The core idea is simple: keep lightweight, intentional records of people’s interests, experiences, travels, and achievements—information that often fuels the best conversations but is easy to forget over time. In Chronicles, the responsibility for nurturing relationships stays with the user, not an algorithm. For example, if I am planning a trip to Greece, instead of searching the web, I can easily recall which friends have lived there, visited recently, or share related interests—and start a more personal and relevant conversation. In this talk, I will briefly introduce the motivation behind Chronicles, its design philosophy, and how open-source, decentralized tools can support more human-centered ways of staying connected.

        Speaker: Jonathan Samuel (CERN - IT-CD-DPP)
      • Symbolic Neural Nets in Dynamic Control System 6m

        Within the framework of Symbolic Neural Nets in Dynamic Control Systems, this work presents the mathematical modeling of a cryostat aimed at obtaining a state-space representation of its dynamic behavior. A physics-based model is first developed using fundamental principles to identify the main variables and interactions of the system. Subsequently, symbolic neural network techniques are applied to approximate the nonlinear dynamics and derive the corresponding state equations. This approach integrates classical modeling with symbolic learning methods, enabling an accurate representation of complex dynamic systems. The results highlight the suitability of symbolic neural networks for system identification, simulation, and future control-oriented applications. design choices vary across fields and use cases.

        Speaker: Fernanda Javiera Zaata Bascuñán Zaata Bascuñán
    • 9:45 AM 10:45 AM
      Machine Learning L3 1h
      Speaker: Andre Sznajder (Universidade do Estado do Rio de Janeiro (BR))
    • 10:45 AM 11:00 AM
      Announcements 15m
    • 11:00 AM 11:30 AM
      Coffee break 30m
    • 11:30 AM 12:30 PM
      Data Technologies L4 1h
      Speaker: Alberto Pace (CERN)
    • 12:30 PM 1:30 PM
      Machine Learning L4 1h
      Speaker: Andre Sznajder (Universidade do Estado do Rio de Janeiro (BR))
    • 1:30 PM 2:30 PM
      Lunch 1h
    • 2:30 PM 3:30 PM
      Machine Learning E1 1h
      Speaker: Andre Sznajder (Universidade do Estado do Rio de Janeiro (BR))
    • 3:30 PM 4:30 PM
      Machine Learning E2 1h
      Speaker: Andre Sznajder (Universidade do Estado do Rio de Janeiro (BR))
    • 4:30 PM 5:00 PM
      Coffee break 30m
    • 5:00 PM 7:30 PM
      Study time or sports 2h 30m
    • 7:30 PM 8:30 PM
      Dinner RQ Hotel 1h RQ Hotel

      RQ Hotel

    • 9:00 AM 9:45 AM
      CENIA: AI at the service of the public. 45m

      National center of AI in Chile presentation

      Speaker: Carlos Sing
    • 9:45 AM 10:45 AM
      Machine Learning L5 1h
      Speaker: Andre Sznajder (Universidade do Estado do Rio de Janeiro (BR))
    • 10:45 AM 11:00 AM
      Announcements 15m
    • 11:00 AM 11:30 AM
      Coffee break 30m
    • 11:30 AM 12:30 PM
      Data Technologies E3 1h
      Speaker: Alberto Pace (CERN)
    • 12:30 PM 1:30 PM
      Data Technologies E4 1h
      Speaker: Alberto Pace (CERN)
    • 1:30 PM 2:30 PM
      Lunch 1h
    • 2:30 PM 3:30 PM
      Machine Learning E3 1h
      Speaker: Andre Sznajder (Universidade do Estado do Rio de Janeiro (BR))
    • 3:30 PM 4:30 PM
      Machine Learning E4 1h
      Speaker: Andre Sznajder (Universidade do Estado do Rio de Janeiro (BR))
    • 4:30 PM 5:15 PM
      Transport to football field 45m
    • 5:15 PM 6:45 PM
      Traditional football match at Canchas Futbol, PUC San Joaquín 1h 30m Canchas Futbol, PUC San Joaquín

      Canchas Futbol, PUC San Joaquín

    • 6:45 PM 7:30 PM
      Transport back to the Hotel 45m
    • 7:30 PM 8:30 PM
      Dinner RQ Hotel 1h RQ Hotel

      RQ Hotel

    • 9:00 AM 12:00 PM
      Departure of participants 3h