Role Overview
Qualcomm AI team develops hardware and software for ML solutions across data center, edge, automotive markets. Seeking engineers with experience in ML frameworks, compiler technology, vectorization, optimization, programming in memory, and ML toolchains. Fast-paced, cross-functional environment requiring strong communication and execution.
Responsibilities
- Research, design, develop ML framework components and compilers for performance and code-size needs
- Analyze software requirements, consult with architecture/HW engineers, implement solutions for Qualcomm SoCs
- Identify system integration issues, interface with development, integration, and test teams
Requirements
- Bachelor's in Engineering/CS + 4+ years systems engineering (or Master's + 1 year, or PhD)
- Preferred: Internal knowledge of PyTorch, TensorFlow; model optimization (torch compile)
- LLVM or industrial compiler development; compiler internals knowledge
- SIMD/multi-threaded high-performance code; DSP kernels
- C/C++, Python (3+ years); OOP, TDD, GoogleTest (3+ years)
- Git; low-level implementation with HW/simulators
- NN architectures, ML compiler workload synthesis; programming in memory
- HW accelerators, HW/SW co-design; C++14/17 advanced features
- RTOS and Linux on SoC; profiling/optimization techniques