This role pays more than
67%
of similar roles. Most pay
$177,087–$243,600
— the shaded band above.
At the midpoint, this role pays about
$233k
versus about
$210k
for comparable roles.
Based on 240 similar postings.
Employer
About Qualcomm
Qualcomm is a leading American semiconductor and telecommunications company based in San Diego, CA.
Qualcomm currently has
615 open roles
on FindRole.
Listed pay typically runs
$155,400–$232,600
across 593 roles with salary data.
As a Compute Architect and Power Engineer within the Low-Power AI Systems group, you will focus on performance optimization, data-path analysis, and architecture for embedded AI subsystems. You will drive power-efficient system designs by analyzing trade-offs across DSP and eNPU components to support XR and always-on AI workloads. Your daily responsibilities include performing memory-access analysis for cache, SRAM, and DDR, executing lab-based power measurements, and developing data reuse strategies. The role requires expertise in embedded processor architectures, RTOS, and Python for modeling and automation. You will also apply knowledge of fixed-point implementation and algorithm optimization to solve challenges in mobile, compute, and IoT platforms. Key technical competencies include power modeling, bandwidth analysis, and the use of specialized profiling tools to correlate silicon data with models for various on-device AI applications.
Optimize performance and power for DSP, eNPU, and memory subsystems in XR and always-on AI workloads.
Perform detailed data-path and memory-access analysis to identify bottlenecks in cache, SRAM, and DDR.
Drive workload partitioning, software optimizations, clock/bandwidth voting, and data reuse strategies.
Execute lab-based power measurements and correlate silicon data with system models.
Support the integration, benchmarking, and commercialization of LPAI solutions across mobile, XR, and IoT platforms.
Document performance findings, competitive analysis, and architectural recommendations for internal stakeholders.
What we're looking for
Bachelor's degree plus 6 years of experience, Master's plus 5 years, or PhD plus 4 years in Engineering, Information Systems, Computer Science, or a related field.
Experience with embedded processor architectures such as DSPs and NPUs including understanding of processor power behavior.
Experience working with embedded platforms, RTOS, and performance/power profiling tools.
Strong programming skills in Python for analysis, modeling, and automation.
Solid understanding of memory systems, data movement, bandwidth analysis, and Cache memory strategies.
Strong fundamentals in power modeling, power analysis, and system-level power optimization.
Hands-on experience with power measurement tools and data analysis techniques.
Knowledge of fixed-point implementation and algorithm optimization techniques.
Experience with Qualcomm DSP and LPAI architectures, SDKs, or internal power tools (preferred).
Background in audio or always-on AI use cases (preferred).
Exposure to ML inference workloads and their power-performance characteristics (preferred).
Electrical Systems Engineering
Digital Logic Design
ASIC
Power Modeling
High Level Synthesis
Machine Learning
Camera Systems
Waveform Analysis
Simulation Tools
Verification