Noticias en español
- The Minister of Science, Aisén Etcheverry, emphasized that “collaboration at the Latin American level is essential for advancing development.”
- Mateo Valero, director of the Barcelona Supercomputing Center, was the event’s special guest, along with regional leaders from AMD, DELL Technologies, Lenovo, and NVIDIA.
After 10 years, the Latin American High Performance Computing Conference (CARLA) returned to Chile, bringing together global experts to enhance the exchange and dissemination of new ideas, techniques, and research in HPC (high-performance computing) and its areas of application.
This 2024 edition, held at the Faculty of Physical and Mathematical Sciences of the University of Chile, placed special emphasis on artificial intelligence, machine learning, quantum computing, simulation and modeling, and Exascale systems.
“Collaboration at the Latin American level is essential if we are to advance development, improve people’s well-being, address climate change challenges, health issues, and strengthen our democracies. We need technological capacities, and conferences like CARLA highlight the importance of computing infrastructure,” remarked Aisén Etcheverry, Chile’s Minister of Science, Technology, Knowledge, and Innovation, during the opening ceremony.
Philippe Navaux, president of the Advanced Computing System for Latin America and the Caribbean (SCALAC), highlighted that “we are at a pivotal moment where all countries, particularly in Latin America, are collaborating on HPC development thanks to AI. CARLA 2024 is catalyzing the participation of researchers, students, and companies interested in presenting their latest technologies.”
“It is a great pleasure for the University of Chile to host CARLA. This is a space to discuss scientific innovation. We are at a unique and historic moment for supercomputing. The demands are enormous, and the applications are limitless, from artificial intelligence to large language models (LLM) and deep neural networks. This makes supercomputing infrastructure one of the most relevant technologies today,” highlighted José Correa, Vice-Rector of Information Technologies at the University of Chile.
Ginés Guerrero, director of the National Laboratory for High Performance Computing (NLHPC) and host of the event alongside the Center for Mathematical Modeling (CMM), explained that “collaboration is key. Chile doesn’t need to start from scratch in supercomputing; it must strengthen these cooperation spaces with different countries.”
“The conference has been a success, with over 300 attendees from 28 countries. This year, we have support from 16 of the world’s leading high-performance computing companies. We hosted five thematic workshops, six advanced tutorials, five discussion panels, and featured more than 30 renowned speakers. We received 42 scientific papers, representing a 68% increase compared to the previous edition. A notable novelty was the inclusion of a school for HPC system administrators, a crucial space for sharing experiences and strengthening collaborations, as they are the backbone of our supercomputing centers. Special thanks to Antonio Russo and Nicolás Wolovick for leading the initiative and to all the professors for their efforts,” Guerrero detailed.
Mateo Valero in Chile
The 11th edition of CARLA, held between September 30 and October 4, featured the keynote talks: “Parallel Performance Analysis at Scale: From one node to a million HPC cores” by Bernd Mohr (Germany), ‘Confidence in HPC data science workflows’ by Marta Mattoso (Brazil), ‘HPC and data analytics for solving socio-environmental problems’ by José María Cecilia (Spain), ‘Accurate predictions for high performance processors’ by Alberto Ros (Spain), “Training AI models on massive datasets with DASF” by Edson Borin (Brazil), ‘When Sally met Harry or When AI met HPC’ by Ulises Cortés (Mexico), ‘Towards robust and large-scale cyber infrastructures for science and technology in Latin America and the Caribbean’ by Carlos Jaime Barrios (Colombia) and ‘Latest trends in high-performance computing architectures’ by Mateo Valero (Spain).
Precisely, the director of the Barcelona Supercomputing Center (BSC) stated that “my talk was about the design of accelerator chips for artificial intelligence, because it is the most important topic currently in the field of computer architecture. How do we design chips to run new and upcoming artificial intelligence applications with minimum energy cost and maximum speed”.
The Spanish scientist, who was awarded the Doctor Honoris Causa distinction by the University of Chile during his visit, presented the Mateo Valero Award during CARLA 2024 to Luis Alberto Nuñez de Villavicencio, PhD in Physics, full professor at the Universidad Industrial de Santander (UIS), in Colombia and associated with the Relativity and Gravitation Research Group of the UIS. In addition, this year there was special recognition for the outstanding career of SCALAC’s president, Philippe Navaux.
Elite HPC Companies
During five days, CARLA 2024 also brought together the highest authorities in Latin America and the Caribbean from manufacturers, integrators and channels of high performance computing solutions: DELL Technologies, Lenovo, AMD, NVIDIA, Intel, Versatus HPC, BeeGFS, EVIDEN, DNN, Sonda, IBM, Supermicro, Hewlett Packard Enterprise, Entel Digital, Emtec Group and ThinLinc.
Joaquim Merino, HPC & AI Solution Sales Director Lenovo, expressed that “the most relevant thing about CARLA 2024 is that here, in Santiago de Chile, are the people who are key from the point of view of how we think and what is going to be done in HPC. They are Trusted Advisors who make all the references for developing high performance computing and artificial intelligence.”
“For DELL it is fundamental to be able to participate in CARLA 2024, because it allows us to connect with the community, with the needs that researchers, students and developers have. This allows us to generate networking to understand the needs, take them inside our organization and transform them into solutions,” said the general manager of DELL Technologies for Chile and Bolivia, Mauricio Chacón.
AMD‘s general manager for Latin America, Nicolás Cánovas, emphasized that “participating in CARLA 2024 allows us to connect with leaders in supercomputing and strengthen this commitment. At the same time, we are focused on a more sustainable future, seeking to increase the energy efficiency of our processors by 30 times by 2025, and offering flexible solutions that fit the specific needs of the region. At AMD, we promote innovation through open ecosystems, collaborating with different organizations to make artificial intelligence more accessible in Latin America.”
Fabio Alves, Business Manager for Higher Education and Research at NVIDIA, said, “Here are the heads that think about our region and you can get to know NVIDIA technology a little better. NVIDIA is always remembered for the GPUs, but it is important that we show all the layers of software and all the layers of educational programs that we have for the region. It’s important to be at CARLA, because here are the people who are going to create high performance computing and artificial intelligence in our Latin American region.”
At the close of the event in Santiago de Chile, SCALAC authorities confirmed that CARLA 2025 will be held in the city of Kingston, Jamaica.
What is CARLA?
Building on the success of the previous editions of the High-Performance Computing Latin America Community (HPCLATAM) and the Latin American Conference on High Performance Computing (CLCAR), in 2014 the two main Latin American HPC workshops jointly developed the first Latin American High Performance Computing Conference, abbreviated as CARLA. The event was held in Valparaiso, specifically at the Chilean university Federico Santa Maria.
The main objective of the conference is to provide a regional forum to foster the growth of the HPC community in Latin America, through the exchange and dissemination of new ideas, techniques and research in high performance computing and its application areas.
Its 2024 edition was organized by the Faculty of Physical and Mathematical Sciences of the University of Chile, the Center for Mathematical Modeling (CMM), the National Laboratory for High Performance Computing (NLHPC), Advanced Computing System for Latin America and the Caribbean (SCALAC), the Center for Data Analysis and Supercomputing (CADS) of the University of Guadalajara, the Latin American Cooperation Network for Advanced Networks (RedCLARA), the National University Network (REUNA), CINECA, CSCS and EPFL Scitas.
Watch the best of CARLA 2024:
Center for Mathematical Modeling
The CMM is today the most active scientific research institution in mathematical modeling in Latin America. It is a center of excellence of the National Agency for Research and Development (ANID) of Chile, integrated by eight partner universities and located at the Faculty of Physical and Mathematical Sciences of the University of Chile. It is also the International Research Laboratory (IRL) #2807 of the French National Center for Scientific Research (CNRS).
Its mission is to create mathematics in response to problems in other sciences, industry and public policy. It seeks to develop science with the highest standards, excellence and rigor in areas such as data science, climate and biodiversity, education, resource management, mining and digital health.
Alonso Farías Ponce, journalist of the Center for Mathematical Modeling.
Posted on Oct 7, 2024 in News
