European supercomputers: buying versus building 

In 2017, Europe created the EuroHPC initiative and its associated legal funding structure, the “EuroHPC JU” Joint Undertaking with two main objectives. The first objective is to acquire, build and deploy world-class high performance computing (HPC) infrastructure across Europe. The second objective is to conduct research and development to build HPC hardware manufactured in Europe, as well as the applications (software) that would run on future locally developed European supercomputers.This talk will cover both objectives in detail. On the one hand, Europe has recently committed a substantial amount of money to the first goal. For example, in the June 2024 Top-500 list, 9 of the Top-20 supercomputers are from Europe. We will go deeper and describe the two main components of the heterogeneous MareNostrum 5 supercomputer, listed separately in positions 8 and 22 of the June 2024 Top-500. Installed at our Barcelona site, MareNostrum 5 represents a good illustration of the challenges of building a contemporary supercomputer; for example, space requirements dictated that BSC could no longer implement it within our Church. Therefore, the MareNostrum 5 had to be installed in a larger space; while the Church will be used to install our first Quantum Computer, thus fulfilling the prophecy made by Dan Brown in his book "Origin".

On the other hand, and as the second part of my talk, I will describe the European approach to design general Made-in-Europe processors and accelerators leveraging the RISC V Open Instruction Set Architecture (ISA).  Currently, this approach is embodied in a couple of large-scale European research projects, namely the European Processor Initiative, EUPilot, Eprocessor, as well as some nationally funded projects. I will briefly describe these projects, including the proof-of-concept chips that successfully boot Linux. I will briefly hint at the future and describe the initiatives that Europe and the BSC are pursuing with the main goal of developing software and hardware for the MareNostrum 6 supercomputer that should be a reality in 2027-2028.

AuroraGPT: Rationale, Challenges and Development of an AI Research Assistant

Innovative methods, new instruments, disruptive techniques, and groundbreaking technologies have led to significant leaps in scientific progress. The increasingly powerful Large Language Models (LLMs) released each month have already sped up research activities such as concept explanation, literature search, and summarization. The transformative potential of AI in research activities, in particular, foundation models, raises important questions about their performance in science activities, their potential application in different contexts, and their ethics.  In this talk, I will first explore the notion of AI research assistants and then discuss the gap between an ideal AI research assistant and the current LLMs, focusing on HPC and parallel computing research problems. The gap motivates the development of research-oriented LLMs. AuroraGPT is developed as an open foundation model trained specifically with scientific data to explore solutions toward the realization of effective AI research assistants. I will describe the activity, challenges, and progress of the different groups developing the key aspects of AuroraGPT. I will particularly focus on the critical and hard task of LLMs' scientific skills,  safety, and trustworthiness evaluation.

Did you know that Madrid was a relatively small town, practically unknown outside of Spain before the year 1561?  The city’s fortunes changed that year when it burst upon the scene of European politics by becoming the permanent capital of Spain.  The dynasty at the head of this change was known as the Habsburgs, a family that ruled the country and much of the known world from the 16th to the 18th century, and who were referred to in Spain as the House of Austria.

The historic center of Madrid was built up predominantly during the reign of that same dynasty, and this fascinating walking tour of Madrid de los Austrias, takes you through that area, giving you the best introduction to the Spanish capital.

Bridging the Data Gaps to Democratize AI in Science, Education and Society

The democratization of Artificial Intelligence (AI) necessitates an ecosystem where data and research infrastructure are seamlessly integrated and universally accessible. This talk overviews the imperative of bridging the gaps between these components through robust services, facilitating an inclusive AI landscape that empowers diverse research communities and domains. The National Data Platform (NDP) aims to lower the barriers to entry for AI research and applications through an integrated services approach to streamline AI workflows, from data acquisition to model deployment. This approach underscores the importance of open, extensible, and equitable systems in driving forward the capabilities of AI, ultimately contributing to the resolution of grand scientific and societal challenges. Through examining real case studies leveraging open data platforms and scalable research infrastructure, the talk will highlight the role of composable systems and services in NDP to catalyze a platform to empower users from all backgrounds to engage in meaningful research, learning, and discovery.