AI Infrastructure Expert
Description
About the Role
Drive the technology development of AI compute and intelligent computing infrastructure, and formulate the technology strategy and roadmap for intelligent computing. Design and develop high-performance, scalable AI infrastructure architectures, including networking, storage, Kubernetes container clusters, and GPU compute clusters. Develop scenario-specific AI infrastructure solutions for training and inference based on business needs. Optimize and operate the AI infrastructure and resource platform to improve AI resource utilization. Collaborate with algorithm and engineering teams to optimize the alignment between model architecture and compute resources, providing underlying technical support and performance tuning for business teams.
Responsibilities
- Drive the technology development of AI compute and intelligent computing infrastructure, and formulate the technology strategy and roadmap for intelligent computing;
- Design and develop high-performance, scalable AI infrastructure architectures, including networking, storage, Kubernetes container clusters, and GPU compute clusters;
- Develop scenario-specific AI infrastructure solutions for training and inference based on business needs;
- Optimize and operate the AI infrastructure and resource platform to improve AI resource utilization;
- Collaborate with algorithm and engineering teams to optimize the alignment between model architecture and compute resources, providing underlying technical support and performance tuning for business teams.
Qualifications
- Bachelor's degree or above, computer science or electronics related majors preferred;
- 8+ years of work experience, with at least 5 years in AI and large model related fields;
Required Skills
- Familiarity with mainstream or China-made (domestic) GPU computing systems and technical architectures, with hands-on experience in large-scale distributed systems, cloud computing platforms, or intelligent computing center construction preferred;
- Deep understanding of computer networking (such as TCP/IP, BGP, VxLAN), distributed storage systems, and GPU hardware principles;
- Familiarity with container technologies (such as Docker, Kubernetes), proficient in at least one programming language (such as Python, Go, C++), with experience in Kubernetes cluster deployment, management, and optimization;
- Familiarity with the GPU computing ecosystem (such as NVIDIA GPU Operator, CUDA), with GPU performance tuning experience;
- Solid understanding of large model technologies, with end-to-end LLM/multimodal model deployment experience and familiarity with multi-node, multi-GPU cluster deployment and tuning;
- Strong problem analysis and solving abilities, able to quickly locate and solve complex system problems, good at collaborating with team members and across departments to drive efficient project implementation.
- Fluent in spoken Chinese for business contexts.
Skills
Sign in to read and join the discussion.