Qualcomm
Low power AI/ML System Architect
$155K–$233K
San Diego, California · On-site · Full-time · Posted Oct 22, 2026
- Experience
- Bachelor's degree: 4+ years; Master's degree: 3+ years; PhD: 2+ years of Systems Engineering or related work experience.
- Education
- Bachelor's, Master's, or PhD in Engineering, Information Systems, Computer Science, or a related field.
About the job
Architects and optimizes embedded AI subsystems for low power and high performance, with emphasis on always-on AI, XR, and ambient use cases.
What you'll do
- Analyze and optimize AI system performance and power for always-on workloads on Snapdragon SoCs.
- Support integration, benchmarking, and commercialization across Mobile, XR, Compute, and IoT platforms.
- Analyze data paths and memory access to identify performance-efficiency bottlenecks.
What they ask for
- Experience with embedded ML accelerators, NPUs, or processors, and understanding of low-power design principles.
- Experience with embedded platforms, RTOS, and performance/power profiling tools.
- Strong programming skills in Python, pytorch, and executorch.
- Understanding of memory systems, data movement, bandwidth analysis, cache strategies, power modeling, and power optimization.
Nice to have
- Experience with Qualcomm NPU and LPAI architectures or SDKs.
- Exposure to ML training and inference workloads and their power-performance characteristics.
Pay
$155,400.00 - $233,200.00 USD base salary
Tools
Snapdragon SoCs, RTOS, Python, pytorch, executorch, NPU, LPAI, DSP
AI focus
Optimizes power and performance for embedded, on-device AI systems, including always-on workloads; analyzes AI subsystem data paths, memory access, and ML workload trade-offs.
Model-written summary · the original posting is authoritative