Supercomputers are indispensable tools for solving the most challenging and complex scientific and technological problems through simulations. BSC will host one of the Pre-exascale machines from EuroHPC JU. This new machine, named MareNostrum5, will have more than 200PFlops sustained, making it one of the most powerful supercomputers in the world. It will feature accelerated nodes with GPUs and general-purpose nodes with modern architecture to maximize code performance.
The open job position will be integrated into the Operations department within the High-level support team. The team will work on applications support for key European projects, focusing on improving the scalability and performance of scientific applications and tools towards Exascale, Data Analytics, Machine Learning, and code refactoring.
Key Duties
1. Optimize and adapt scientific application codes to new pre-exascale architectures and systems.
2. Improve the performance of existing parallel codes, enhancing serial efficiency and scalability, including code modifications or collaborating with developers.
3. Adapt algorithms and mathematical libraries to specific architectures, such as accelerators and new programming models.
4. Provide consultancy to scientists on new architectures and programming models.
5. Generate performance analyses and benchmarks, reporting results to application developers.
6. Collaborate with European and international teams on technical aspects of scientific application support.
Requirements
Education
* PhD in computer science or a related scientific field aligned with HPC.
* Bachelor's in Computer Science or related discipline with at least 3 years of experience in HPC code development.
Essential Knowledge and Experience
* Experience porting and optimizing applications on UNIX-based systems using Fortran, C, MPI, OpenMP, and parallel methods.
* Proficiency with performance analysis tools and parallel debuggers.
* Experience collaborating with external partners.
* Strong understanding of Linux environments and Shell scripting.
* Experience with parallel programming codes (MPI and OpenMP).
Additional Knowledge and Experience
* Experience managing large, collaborative projects, and familiarity with git and SVN.
* Experience porting codes to GPGPU accelerators.
* Deep understanding of HPC architectures.
Competencies
* Excellent communication and interpersonal skills for teamwork.
* Strong analytical and problem-solving abilities.
Conditions
* Position based at BSC within the Operations Department.
* Full-time contract with a stimulating environment, state-of-the-art infrastructure, flexible hours, training, and benefits.
* Open-ended contract linked to project and budget duration.
* 23 paid vacation days plus holidays on December 24th and 31st.
* Competitive salary based on qualifications and experience, aligned with Barcelona's cost of living.
#J-18808-Ljbffr