HPC Lead System Administrator

About the role

The Lead System Administrator will oversee the operation of a large Linux‑based high‑performance computing (HPC) cluster comprising over 300 GPU/CPU nodes. This hybrid technical‑leadership position involves supervising a small team of experienced administrators while ensuring the reliability, performance, and security of compute resources, storage, and networking.

Key responsibilities

• Plan, execute, and monitor system upgrades to meet laboratory and customer requirements.
• Develop and enforce workload‑scheduler policies using SLURM, including GPU scheduling.
• Maintain parallel file systems (Weka IO, Lustre) and high‑performance networks (InfiniBand, TCP/IP).
• Automate node configuration and management through scripting (Bash, Python, Ruby) and configuration‑management tools (Puppet).
• Manage hardware failures, spare‑part inventories, and vendor relationships.
• Lead, mentor, and evaluate a team of 3‑4 HPC administrators.
• Coordinate multi‑project planning, scheduling, and reporting.
• Identify training needs, provide technical guidance, and respond to security or safety incidents.

Required profile

• Proven experience managing large Linux HPC clusters and leading technical teams.
• Strong analytical mindset with systematic problem‑solving abilities.
• Excellent written and oral communication skills in English.
• Ability to build effective relationships with researchers, faculty, and external suppliers.

Required skills

• SLURM workload manager, including GPU scheduling.
• Parallel file systems such as Weka IO and Lustre.
• High‑performance networking (InfiniBand, TCP/IP).
• Scripting languages: Bash, Python, Ruby.
• Configuration‑management tool: Puppet.
• Technical documentation proficiency.