GPU CFD Solver Engineer (LBM, HPC)

Siemens; Vlaams-Brabant; Hybride; EUR 65 000 - 90 000; Plein temps

Résumé du poste

Siemens Software is seeking a CFD Software Engineer to join the solver development team for Simcenter Ultrafluid. You will contribute to GPU-accelerated solver development, implementing high-quality features and collaborating across HPC, pre/post-processing, and meshing teams.

The role requires an MSc/PhD in mechanical engineering or related field, at least two years of CFD solver experience, and strong C++ and parallel programming skills. Knowledge of Python or Julia, LBM, and CUDA is preferred.

Qualifications

• MSc or PhD in mechanical engineering, computational mechanics, computer science, or related field.
• At least two years of hands-on CFD solver development experience.
• Strong C++ and parallel programming skills.

Responsabilités

• Contribute to the code development of Simcenter Ultrafluid in the agile team.
• Specify, implement, test, and document new high-quality features of the solver in collaboration with teammates.
• Contribute to performance benchmarking and optimization in a massively parallel GPU environment.
• Collaborate with product management, application specialists, and QA engineers to meet industrial needs.
• Attend daily stand-ups and regular reviews, retrospectives, and planning meetings.

Connaissances

CFD solver dev

C++ parallel

English fluency

Agile teamwork

Formation

MSc/PhD in mechanical engineering or related

Outils

CUDA toolkit

LBM knowledge

Python/Julia

Description du poste

Siemens Software is seeking a CFD Software Engineer to join the solver development team for Simcenter Ultrafluid. You will contribute to GPU-accelerated solver development, implementing high-quality features and collaborating across HPC, pre/post-processing, and meshing teams.

The role requires an MSc/PhD in mechanical engineering or related field, at least two years of CFD solver experience, and strong C++ and parallel programming skills. Knowledge of Python or Julia, LBM, and CUDA is preferred.