LPAI Architecture and Systems Engineer

As part of Qualcomm's Audio and Low-Power AI (LPAI) Architecture group, your role involves architecting, analyzing, and optimizing DSP and embedded NPU (eNPU) performance across Snapdragon value-tier platforms. Your primary focus will be on architectural analysis, optimization, and deployment of low-power AI solutions, enabling efficient on-device intelligence for audio, sensors, and always-on use cases. You will work on tasks such as eNPU scheduling, memory hierarchy, compression/quantization strategies, and clock/BW voting to facilitate efficient on-device AI. Your responsibilities include: - Analyzing, designing, and optimizing LPAI (Hexagon DSP, eNPU, TCM/UBUF/LLC/DDR) for performance, power, and area efficiency in value-tier chipsets. - Conducting architectural analysis and benchmarking of LPAI subsystems to identify bottlenecks and propose solutions for improved throughput and efficiency. - Collaborating with hardware and software teams to define and implement enhancements in DSP/eNPU microarchitecture, memory hierarchy, and dataflow. - Developing and validating performance models for AI workloads, including signal processing and ML inference, on embedded platforms. - Prototyping and evaluating new architectural features for LPAI, such as quantization, compression, and hardware acceleration techniques. - Supporting system-level integration, performance testing, and demo prototyping for the commercialization of optimized DSP/eNPU solutions. - Working closely with cross-functional teams to ensure the successful deployment and commercialization of value-tier chipset features. - Documenting architectural analysis, optimization strategies, and performance results for internal and external stakeholders. Qualifications required for this role include a solid background in DSP architecture, embedded NPU design, and low-power AI systems. You should have proven experience in performance analysis, benchmarking, and optimization of embedded processors, strong programming skills in Embedded C/C++, Python, and Matlab, as well as expertise in fixed-point and floating-point implementation focusing on ML/AI workloads. Additionally, excellent communication, presentation, and teamwork skills are essential for this position. The educational qualifications needed are a Masters or PhD degree in Engineering, Electronics and Communication, Electrical, Computer Science, or a related field. Preferred qualifications include familiarity with Qualcomm DSP/eNPU architectures, SDKs, and tools, experience with LPAI frameworks and design, and knowledge of audio or sensor signal processing frameworks. Qualcomm is an equal opportunity employer and is committed to providing accommodations for individuals with disabilities during the application/hiring process. For more information about this role, please contact Qualcomm Careers.,