Senior GPU Architect - Performance and Yield Optimization
Core
Shape GPU architectures to improve product yield while maintaining performance and simplicity by analyzing manufacturing defects and developing isolation/disabling methods.
Role type
Senior IC GPU Architect (Performance and Yield Optimization)
Builds
GPU architectures, software tools, and optimization frameworks for NVIDIA GPUs
Domain
Semiconductor / GPU Architecture / Yield Optimization
Deliverable
production ML models
Required skills
GPU architecture, computer architecture fundamentals, C++, Python, constraint-based reasoning, combinatorial optimization, silicon yield analysis, multi-die architecture design
Preferred skills
Hands-on GPU architecture, large-scale SoC architecture, defect tolerance/harvesting/redundancy experience, chiplet/modular processor architectures, silicon/manufacturing data analysis
Responsibilities
Analyze manufacturing defects to isolate or disable affected regions; build software tools and models capturing architectural rules; develop optimization techniques for large configuration spaces; evaluate yield, performance, area, and cost tradeoffs; move proposals into production across architecture, design, and software teams
