onsite
GPU Performance Modeling & Optimization Engineer - AMD
Software Engineer
GPU Performance Modeling & Optimization Engineer driving next‑generation AI and data center acceleration through advanced CUDA, C++ profiling, and machine learning‑based performance tuning.
About the role
Key Responsibilities
- Develop and refine GPU performance models for AI, data center, and gaming workloads.
- Implement CUDA and C++ optimizations to reduce latency and increase throughput.
- Leverage profiling tools (Nsight, VTune) to identify bottlenecks and validate performance gains.
- Collaborate with architecture, driver, and software teams to integrate model insights into product roadmaps.
- Document methodologies, benchmark results, and best‑practice guidelines for internal use.
Requirements
- Strong background in GPU programming with CUDA and C++.
- Proven experience in performance modeling, profiling, and optimization.
- Knowledge of machine learning workloads and their GPU acceleration patterns.
- Excellent analytical skills and ability to translate complex data into actionable insights.
- Effective communication skills for cross‑functional collaboration.