Key Responsibilities :
Analyze and optimize HPC application performance (MPI, OpenMP, CUDA).
Run and evaluate benchmarks (HPL, STREAM, OSU, SPEC MPI, etc.).
Use profiling tools (VTune, HPCToolkit, ARM MAP, Perf) for tuning.
Compare CPU / GPU / accelerator architectures and interconnects.
Support performance reports and recommendations for HPC environments.
Requirements : Minimum 2
years of experience
in HPC performance tuning and benchmarking ( no upper limit ).
Proficiency in at least one programming language : C, C++, or Python
(C or C++ preferred for low-level performance tuning).
Experience with
MPI, OpenMP , and
GPU programming (CUDA or OpenCL) .
Familiarity with job schedulers such as
SLURM
or
PBS .
Good understanding of
HPC hardware , including memory hierarchy, interconnects, and processor architecture (CPU / GPU / Accelerators).
Performance Engineer • India