Build machine learning compiler performance solutions at Qualcomm, using C++ and Python to optimize algorithms for efficient Qualcomm hardware mappings. This Staff-level engineering role spans ML systems, compiler development, and workload analysis. You will connect PyTorch with Qualcomm ML/AI compiler flows while studying network design trends. Your work will inform future hardware and software features through performance, area, and power analysis. Key Responsibilities:

Develop and evolve production and exploratory ML/AI compilers Define algorithms for high-performance,



low-power Qualcomm hardware compilation Couple PyTorch efficiently to Qualcomm ML/AI compiler flows Create C++ and Python performance-driven simulation components Analyze simulations and recommend hardware and software enhancements Requirements:

Bachelor’s degree and four-plus years of related experience Master’s degree and three-plus years of related experience PhD and two-plus years of related experience Strong C++ and Python knowledge with algorithm development experience Performance analysis, debugging, and analytical problem-solving skills Apply compiler development, PyTorch integration, and performance analysis expertise to Qualcomm’s power-productive on-device AI systems.

#J-18808-Ljbffr

📌 Machine Learning Compiler Performance Engineer (Winnipeg)
🏢 QUALCOMM
📍 Winnipeg