Systems Administrator, GPU/AI Infrastructure

Network: deploy, support, and administer lab networking, in coordination with the dedicated Network Engineer role for switch and fabric configuration.

Compute: deploy, support, and administer physical and virtual compute infrastructure across the lab.

Storage: deploy, support, and administer lab storage systems across block and file tiers.

Operating Systems: install, patch, and administer operating systems across lab infrastructure.

Docker: deploy and administer Docker container runtime environments.

Kubernetes: own full production Kubernetes administration, including cluster operations, role-based access control (RBAC), cluster networking, and version upgrades.

NVIDIA Technology Stack: administer NVIDIA drivers, the CUDA toolkit, InfiniBand fabric, Run:AI orchestration, and ClearML across the lab’s GPU infrastructure.

Configuration Validation: validate lab configurations against NVIDIA reference architecture (Lenovo Validated Design) certification requirements.

Day-to-Day Support: provide day-to-day administration, support, and engineering functions for the network, compute, and storage layers in the physical lab.

Rack and Stack: perform rack-and-stack, structured cabling, and smart-hands support for lab hardware alongside higher-level administration duties.

Bangalore Secondary Support: provide secondary technical support for NVIDIA technology stack capital equipment hosted in Bangalore.

Lab Coordination: coordinate with the adjacent ISG AI COE lab (Tech Marketing, customer proof-of-concepts) and the Morrisville Executive Briefing Center given shared physical proximity.

NVIDIA Relationship: maintain a direct working relationship with NVIDIA field contacts based in Morrisville.

2+ years of systems administration experience with hands-on breadth across network, compute, storage, and operating system layers, plus a GPU/AI infrastructure specialization.

Production Kubernetes Administration: demonstrated experience with cluster operations, RBAC, cluster networking, and version upgrades in a production environment.

Docker: hands-on experience administering Docker container runtime environments.

Hands-on NVIDIA administration: demonstrated experience administering NVIDIA drivers and the CUDA toolkit.

InfiniBand Fabric: working familiarity with InfiniBand fabric in a multi-node GPU environment.

GPU Orchestration: experience with Run:AI, ClearML, or a comparable GPU orchestration and MLOps toolset.

Physical Infrastructure: comfortable with hands-on hardware work, including rack-and-stack, structured cabling, and smart hands, in addition to higher-level administration.