Qualcomm is looking to optimize the power and constrained performance of next-generation cloud accelerator products by analyzing complex system-level AI use cases.
Requirements
- Experience in ARM assembly language, C, C++, Python
- Strong familiarity with ARM architecture
- Good understanding of System On Chip architectural concepts
- Experience in Power/Thermal/Performance optimizations to improve product level KPIs.
- Prior experience with AI use cases and/or products, etc., is a plus.
- Real-time operating systems, multi-threaded processor architectures
- More than 8 years of experience with programming (e.g., C, C++, Python), computer architecture, or embedded systems.
Responsibilities
- Own and drive the analysis of complex system level AI use cases involving Computer Vision Networks, Recommendation Systems and Gen AI workloads.
- Partner closely with various ecosystem stake holders in optimizing power and constrained performance of AI workloads on Qualcomm’s cutting edge inference accelerators.
- Contribute to development of algorithms and techniques to optimize power consumption and thermal performance of various use cases.
- Perform system level software optimizations using hardware and software power/thermal management techniques.
- Exploration of advanced machine learning techniques in solving system level power and limits management problems.
- Leverages advanced Software knowledge and experience to design, develop, create, modify, and validate critical embedded and cloud edge software, applications, and/or specialized utility programs.
- Designs and implements software sub-systems for complex, large-scale products and systems.
Other
- Works independently with minimal supervision.
- Provides supervision/guidance to other team members.
- Decision-making is critical in nature and highly impacts program, product, or project success.
- Requires verbal and written communication skills to convey complex information. May require negotiation, influence, tact, etc.
- Strong analytical and problem-solving skill
