计算系统解决方案专家
Core
Design and lead end-to-end solutions for domestic AI computing infrastructure, optimizing model training/inference performance and building high-performance distributed systems.
Role type
Senior IC AI Infrastructure Solutions Architect
Builds
High-performance distributed computing systems, optimized AI inference/training pipelines, and custom high-performance operators (Gemm, FAttention, All2All).
Domain
AI Infrastructure / Domestic Computing Hardware
Deliverable
production ML models
Required skills
Python, C++, system design, AI model optimization, cluster inference optimization, high-performance operator development, chip architecture knowledge, distributed computing
Preferred skills
Large model precision tuning, cluster inference experience, GPU/Ascend operator development, chip manufacturing/packaging knowledge
Technologies
Python, C++, GPU, Ascend, Gemm, FAttention, All2All
Responsibilities
Guide medium-to-long-term network planning and budget; lead cross-product solution design; optimize inference performance and system throughput on domestic hardware; assist in migrating self-developed models; develop high-performance custom operators; research next-gen chip design and architecture; build high-reliability distributed computing systems.
