GPU System Administrator

About the role

We are seeking a GPU System Administrator to manage large‑scale GPU platforms that power mission‑critical AI workloads. You will provision, maintain, validate, and troubleshoot GPU servers to ensure consistent, high‑performance infrastructure across racks and regions.

Key responsibilities

• Support GPU systems during the US night‑shift as part of the Singapore team.
• Perform node imaging, bring‑up, health checks, and lifecycle management across GPU systems.
• Run GPU validation, including NVLink checks, NCCL tests, and performance benchmarking.
• Monitor system health, troubleshoot issues, and escalate complex faults.
• Maintain accurate configuration, inventory, and fleet records.
• Assist with deployment of new racks, clusters, and regional expansions.
• Improve automation, runbooks, and operational processes.

Required profile

• Strong written and verbal English communication skills.
• Experience administering GPU‑based infrastructure.
• Robust Linux and HPC background.
• Proficiency with Slurm and familiarity with Kubernetes.
• Knowledge of networking, storage, and system performance tuning.
• Strong problem‑solving abilities and a collaborative approach to mission‑critical environments.

Required skills

• Linux
• HPC
• Slurm
• Kubernetes
• NVLink
• NCCL
• GPU servers
• Networking
• Storage
• Performance tuning
• Automation (scripting/runbooks)

What we offer

• Up to S$280,000 annual salary plus a 20% bonus.
• Remote work in Singapore (right to work without sponsorship required).
• A culture focused on innovative disruption within a fast‑growing AI infrastructure startup.