AI Application Support Specialist for MINERVA project

Context And Mission Supercomputers are essential for tackling the most challenging and complex scientific and technological problems. Beyond simulations, they now enable the use of neural networks and large language models to address engineering challenges, empowering companies to achieve their goals through AI technologies. This position will be part of the Operations Department within the User Support team. The team will focus on supporting AI applications for key projects utilizing BSC resources, including performance optimization, code porting, specialized AI tutoring, and enhancing the scalability and performance of models.BSC hosts one of the largest and most advanced supercomputers in Europe within the framework of EUROHPC-JU. The pre-exascale system MareNostrum 5 (MN5) delivers a sustained performance of 205 PFlops, with over 180 PFlops powered by NVIDIA H100 GPUs, making it one of the world’s leading AI and HPC infrastructures. MN5 represents a key technological pillar for Europe’s digital sovereignty, enabling groundbreaking research in artificial intelligence, data analytics, and scientific computing.A major innovation of the system is the MN5-AI partition, specifically designed to accelerate AI research and industrial innovation. This partition plays a central role in the AI Factories initiative, which aims to strengthen the European AI ecosystem by providing large-scale computational resources for training and deploying advanced AI models. Through MN5-AI, BSC contributes directly to building a federated network of AI supercomputing centers in Europe, fostering collaboration between academia, industry, and public institutions to advance trustworthy and high-performance AI technologies.Among the key initiatives supported by MN5 and MN5-AI is the MINERVA EuroHPC project, which aims to ensure that Europe’s AI communities can fully harness the power of the EuroHPC JU supercomputing infrastructure. MINERVA provides training, education, and technical assistance in collaboration with the EuroHPC Academy and the AI Factories initiative. It supports AI teams through a three-tier system that helps integrate HPC resources into their workflows, port AI applications to HPC systems, and pre-train or customize large-scale foundation models. MINERVA establishes a European “Community Hub” to advance large-scale AI research and development, offering access to datasets, software stacks, and best practices.The HPC User Support team is therefore crucial in optimizing the use of MN5 and MN5-AI, ensuring that users can fully exploit these resources for their AI and HPC workloads. The team assists key projects with daily operations, helping researchers and companies push the boundaries of what is computationally possible. Key Duties Optimize and adapt scientific application codes to new pre-exascale architectures and systems, focusing in AI applications and technologies. Improve the performance of existing AI environments, improving the serial efficiency and the scalability, changing, if necessary, the code or helping the developers with their required modifications. Choose and adapt algorithms and/or library routines to improve applications to specific computer architectures (accelerators, new programming models, etc.). Provide consultancy to scientists on AI solutions that can improve the evolution of their research. Generate performance analysis and benchmarks for selected applications and report the results to the applications developers. Collaborate with other functional groups at European and International level on technical matters related to supporting AI scientific application work.  Requirements Education A Bachelor's degree in Computer Science or a related discipline with a focus on Artificial Intelligence at a technical level is required. Master's degree in a related field is desirable. Essential Knowledge and Professional Experience Experience working with AI models and running them in parallel in HPC systems. Experience using performance analysis tools, and parallel debuggers for GPUs. Experience supporting and collaborating with external partners. Good understanding of Linux environment and Shell scripting. Experience working with Parallel programming codes (MPI and OpenMP) and batch systems like SLURM as a user. At least 2 years of experience in a similar position working with AI solutions. Additional Knowledge and Professional Experience Experience in managing big and collaborative projects and experience with git and SVN. Experience porting codes to accelerators specifically NVIDIA GPUS. A thorough understanding of high-performance computing architectures. Experience porting and optimizing applications on UNIX-based systems experience in Fortran, C, MPI, OpenMP, and parallel methods. Good level in English, both written and spoken Competences Excellent communication and interpersonal skills to be able to work within a team to complete tasks on schedule. Analytical problem-solving ability. Conditions The position will be located at BSC within the Operations Department We offer a full-time contract a good working environment, a highly stimulating environment with state-of-the-art infrastructure, flexible working hours, extensive training plan, restaurant tickets, private health insurance, support to the relocation procedures Duration: Open-ended contract due to technical and scientific activities linked to the project and budget duration Holidays: 22 days of holidays + 6 personal days + 24th and 31st of December per our collective agreement Salary: we offer a competitive salary commensurate with the qualifications and experience of the candidate and according to the cost of living in Barcelona Starting date: asap