onsite
Principal AI Performance Engineer - Arm
Software Engineer
Lead AI performance optimization for edge devices, delivering best‑in‑class inference on Arm hardware. Drive customer‑centric tuning, profiling, and code‑level enhancements to meet real‑world performance and power targets.
About the role
Key Responsibilities
- Collaborate with customers to profile and optimize AI workloads on Arm platforms, ensuring top‑tier inference performance and power efficiency.
- Lead end‑to‑end performance engineering, from model conversion to kernel‑level tuning, using industry‑standard profiling tools.
- Develop and maintain reusable performance libraries, scripts, and automation pipelines in Python and C++.
- Provide technical guidance and best‑practice documentation to customers and internal teams.
- Analyze performance bottlenecks, propose architectural changes, and validate improvements through rigorous testing.
Requirements
- 10+ years of experience in high‑performance computing, with a focus on AI inference on embedded or edge devices.
- Deep expertise in Arm architecture, compiler toolchains, and low‑level optimization techniques.
- Proficiency in Python and C++ for performance tooling and automation.
- Strong analytical skills, with experience using profiling and tracing tools (e.g., Arm DS-5, gprof, perf).
- Excellent communication skills and a proven track record of delivering customer‑centric performance solutions.