System Performance Engineer (NSP/NPU, AI/ML)
Core
Profile and optimize system performance on Snapdragon chipsets, focusing on AI/ML workloads using NPU cores, CPU, and DDR.
Role type
Senior IC system performance engineer (SoC/AI)
Builds
Optimized silicon performance for mobile, automotive, and data center chipsets
Domain
Semiconductor hardware engineering, AI/ML inference, SoC architecture
Deliverable
production ML models
Required skills
NPU micro-architecture, deep learning fundamentals (CNNs, transformers), model optimization (quantization, pruning), CPU architecture (caches, pipelines), DDR specifications (LPDDR4/5/6), performance profiling tools (Windows Performance Toolkit, Snapdragon Profiler, Intel VTune, Linux perf), parallel processing architectures, OS internals (kernel scheduler, MMU), benchmarking (GeekBench, SPECInt, Dhrystone)
Preferred skills
NoC profiling, Linux/QNX driver development, thermal/power domain knowledge, competitive analysis
Technologies
TensorFlow Lite, PyTorch Mobile, ONNX Runtime, QNN, Vela, OpenVINO, Windows Performance Toolkit, Snapdragon Profiler, Intel VTune, Linux perf
Responsibilities
Drive performance analysis on silicon using benchmarks; analyze load patterns across IPs to identify bottlenecks; correlate Perf KPIs of SoC subsystems with projections; evaluate performance at various junction temperatures; optimize System performance parameters of SoC infrastructure; collaborate with cross-functional teams on chipset performance activities
Seniority
Mid-level, hands-on IC
