AI Software Engineer – NPU/Hexagon DSP Kernel Optimization
Core
Develop and optimize high-performance AI software components, neural network kernels, and runtime systems for Qualcomm NPU and Hexagon DSP processors.
Role type
Senior embedded software engineer (AI accelerator kernel optimization)
Builds
High-performance neural network operators and runtime software for NPU/DSP platforms
Domain
Embedded systems, AI accelerators, DSP architecture
Deliverable
production ML models
Required skills
C/C++, processor architecture, neural network operators, kernel optimization, fixed-point computation, quantization, memory hierarchy optimization, SIMD/vector processing, profiling, low-level optimization
Preferred skills
Assembly, hardware-aware software optimization, software-hardware co-optimization
Responsibilities
Develop and optimize AI software components for NPU and Hexagon DSP; Implement high-performance neural network kernels focusing on latency, power, and throughput; Profile runtime bottlenecks and apply hardware-aware optimizations; Partner with architecture teams to improve future NPU/DSP capabilities; Support customer issue resolution and platform bring-up
Seniority
Senior, hands-on IC
