דרושים Production Factory Planner ב-אנבידיה ב-South Korea, Seoul

What you will be doing:

• Maintain an up-to-date understanding of the philosophy, architecture, and deployment methods of various evolving NVIDIA Reference Architectures—e.g., NVIDIA DGX SuperPOD Reference Architecture, NVIDIA Cloud Partner Reference Architecture, and NVIDIA Enterprise Reference Architecture.

• Analyze and understand the requirements of customer-initiated AI training or inference clusters.

• Identify the NVIDIA Reference Architecture that best matches customer needs and effectively communicate its value proposition to collaborators.

• Facilitate seamless communication between NVIDIA's internal deployment teams and customers during the implementation of AI clusters based on Reference Architectures.

• Provide hands-on technical support to developers after the AI Factory has been deployed, ensuring that AI training and inference workloads run effectively on the infrastructure.

What we need to see:

• Bachelor’s degree or higher in Computer Science, Computer Engineering, or a related technical field.

• Solid understanding of basic principles behind cluster orchestration, such as compute resource provisioning and dynamic prioritization based on user demand.

• Minimum of 3 years of hands-on experience operating AI training or inference clusters that leverage Kubernetes with NVIDIA GPUs.

• Proficiency in key technologies including: Container Runtime Interface (CRI), Container Network Interface (CNI), Calico, NVIDIA GPU Operator, NVIDIA Network Operator, and Kubeflow Training Operator.

Ways to stand out from the crowd:

• Foundational knowledge and experience with network technologies—such as InfiniBand and Ethernet—in AI cluster environments, including compute fabric interconnects between GPU servers, storage fabric integration, and in-band networks for system administration.

• Familiarity with the role of storage in AI training/inference clusters, including hands-on experience with vector databases and leading commercial storage solutions.

• Experience integrating MLOps platforms into Kubernetes environments, such as deploying Airflow for orchestrating distributed training workloads.