Senior HPC / CUDA Engineer

About the role

EPAM is seeking a Senior HPC / CUDA Engineer to drive performance improvements for scientific computing applications in the geoscience domain. The role focuses on seismic imaging and Full Waveform Inversion workloads on modern NVIDIA GPU platforms.

Key responsibilities

• Design, develop and optimise CUDA kernels and GPU‑accelerated code for computationally intensive workloads.
• Analyse performance across consumer and data‑centre NVIDIA GPUs, profiling bottlenecks at kernel, PTX, memory and multi‑GPU communication levels.
• Develop distributed multi‑GPU solutions using MPI, NCCL and NVSHMEM and maintain high‑quality C++/Python scientific software components.
• Debug complex runtime issues and contribute to tools for seismic data preprocessing, visualisation and analysis.

Required profile

• Strong experience in high‑performance computing and GPU‑accelerated software development.
• Deep knowledge of NVIDIA GPU architecture, execution models, memory hierarchy and advanced CUDA optimisation techniques.
• Proven ability to work with large‑scale, performance‑critical scientific workloads.
• Excellent analytical and problem‑solving skills for investigating runtime and scalability issues.

Required skills

• CUDA programming and GPU performance optimisation.
• C++ development for performance‑critical applications.
• Python for scientific software.
• MPI, NCCL, NVSHMEM for distributed multi‑GPU computing.
• GPU profiling tools and PTX analysis.
• HDF5 and scientific data formats.

What we offer

• Fully remote work from anywhere in Armenia.
• Opportunities to work abroad for up to two months per year.
• Access to internal learning resources, certifications and English classes.
• Comprehensive medical and family care package, stock purchase plan and referral bonuses.