性能优化工程师（AI计算方向）-基础库与工具链
Core
Analyze online/offline inference and training scenarios to identify software hotspots and hardware bottlenecks, then propose and implement optimization solutions.
Role type
Senior IC performance optimization engineer (AI computing)
Builds
Optimized inference and training performance for LLM/NLP/recommendation/visual/audio business scenarios
Domain
AI computing, high-performance computing, heterogeneous hardware
Deliverable
production ML models
Required skills
Computer architecture, operating systems, compilation principles, network basics, LLM/NLP/recommendation/visual/audio training and inference optimization, AI model training/inference frameworks (PyTorch or TensorFlow), AI operator compilation optimization (Xla/Torch.Compile/Triton), performance evaluation and estimation, AI computing hardware-software co-design
Preferred skills
Experience with heterogeneous card construction and compilation, performance evaluation and estimation, operator performance optimization, AI computing hardware-software co-design
