Traineeship

HPC User Support Traineeship at BSC-CNS

Traineeship details

About the hosting organisation

Host institute / company

Barcelona Supercomputing Centre

Country

Sector(s)

Academic / Higher Education, Non-profit / NGO, Public Sector / Government, Research Institute / National Laboratory

Duration and availability

Duration

6 months

Earliest start date

15/09/2026

About the traineeship

Traineeship name

HPC User Support Traineeship at BSC-CNS

Proposal ID

HPCTRAIN-OFFER-202601-015

Thematic area(s)

The traineeship will focus on supporting HPC users working on the MareNostrum5 supercomputer at the Barcelona Supercomputing Center (BSC-CNS). The activities will include Linux user support, parallel programming assistance, software compilation support, performance analysis, batch scheduling analysis, and post-mortem job analysis.

Name of the position

HPC User Support Trainee for Large-Scale HPC Clusters

About the position

This traineeship is focused on learning and understanding how user support is provided in a large-scale HPC environment such as MareNostrum5. MareNostrum5 is one of the largest supercomputers in the world in terms of node count, with more than 7,500 nodes. The system delivers over 40 PFLOPS sustained performance in CPU partitions and more than 175 PFLOPS sustained performance in GPU partitions. The trainee will gain hands-on experience by collaborating with expert support teams involved in application optimization, performance tuning, troubleshooting, and user assistance. The position will also provide insight into the technical and operational challenges of managing a complex HPC infrastructure used by more than 3,000 researchers and users from different scientific domains.

Minimum qualification level

Bachelor's degree completed, Studies in Computer Science, Engineering, Physics, Mathematics, or another technical discipline, Basic experience using Linux/Unix systems, Good programming and problem-solving skills, Previous exposure to HPC topics such as parallel programming, Linux system usage, or cluster computing will be considered an advantage, Familiarity with concepts such as MPI, OpenMP, or GPU programming is desirable but not mandatory, Good level of written and spoken English is required

Learning objectives

User support workflows in large-scale HPC infrastructures, Optimisation and fine-tuning of scientific applications, Performance analysis and debugging methodologies, Management of heterogeneous computing environments involving CPUs, GPUs, AI accelerators, and emerging technologies such as quantum computing, Best practices for efficient usage of HPC resources and scheduling systems

Tools and technologies to be used

C / C++, CUDA, Fortran, MPI, OpenACC, OpenMP, Python, PyTorch, Slurm, PARAVER, EXTRAE, NVIDIA NSYS, Linux/Unix

View the HPCTRAIN call for trainees call text

Description

The trainee will become part of the HPC User Support team at BSC-CNS and will participate in day-to-day support activities for users running scientific applications on MareNostrum5. The traineeship combines theoretical learning with practical experience in HPC operations, software optimisation, and performance engineering. Throughout the traineeship, the candidate will interact with researchers and support engineers, assisting in debugging execution problems, improving application performance, analyzing job execution behavior, and supporting users in the efficient use of HPC resources. The traineeship offers a unique opportunity to gain experience in one of Europe’s leading supercomputing facilities and to understand the challenges associated with operating cutting-edge HPC systems.

Submit your proposal

Please find more details on how to apply below.

Go to application portal