Develop, parallelize, and optimize scientific application kernels for a custom SPU architecture. Work from mathematical formulations to C++/Python reference implementations, profile and benchmark performance, collaborate with architecture teams, and write technical documentation to achieve HPC performance targets.

Develop, test, and optimize scientific computing applications and numerical methods for CPUs and GPUs. Measure and improve computational performance, collaborate across research and engineering, and contribute to software testing, deployment, and evolving technical needs.