Description
Microsoft is hiring a Principal Software Engineer - Performance Tooling for its Artificial Intelligence Frameworks team. The role focuses on benchmarking, debugging, profiling, and optimizing large-scale model training and inference workloads across CPUs and GPUs, including Microsoft hardware, while reducing deployment time and hardware footprint. The position requires a bachelor's degree in computer science or a related technical field and at least six years of technical engineering experience with C++ or Python; preferred qualifications include advanced degrees, extensive high-performance application experience, deep LLM and GPU profiling expertise, and familiarity with CUDA, ROCm, and Triton. Compensation ranges from USD 142,800 to USD 274,800 annually across the United States, with higher ranges for the San Francisco Bay Area and New York City metropolitan areas.
