Head of AI Infrastructure & HPC

Role overview

Infrastructure leadership

• Design, implement, and expand GPU-accelerated compute infrastructure for AI and HPC applications.
• Define roadmaps covering NVIDIA GPU systems, high-speed interconnects, storage, and cluster management.
• Oversee deployments involving H100, H200, Blackwell, and HGX/DGX architectures.
• Direct engineering across InfiniBand, RoCE, RDMA, NVLink/NVSwitch, and distributed networking.
• Establish deployment methods for hardware integration, system configuration, qualification testing, and production readiness.
• Improve efficiency through GPU optimization, capacity planning, monitoring, and reliability engineering.
• Oversee cluster scheduling, workload management, and container orchestration.
• Introduce automation for provisioning, configuration management, and operational support.
• Partner with hardware manufacturers, datacentre providers, and technology vendors.
• Recruit, develop, and manage specialist engineering teams; shape infrastructure strategy, governance, and operations.

Technical expertise

• Relevant experience may include NVIDIA H100, H200, B200, HGX, and DGX systems; InfiniBand, Ethernet, RoCE, and RDMA.
• GPU tools and technologies include NCCL, CUDA, NVLink, NVSwitch, and DCGM; environments may include Linux HPC and distributed computing.
• Platforms may include Kubernetes, Slurm, Docker, and cluster scheduling; other areas include provisioning, hardware qualification, benchmarking, tuning, and diagnostics.
• Additional areas include parallel file systems, high-throughput storage, rack-level deployment, datacentre integration, and infrastructure for AI training and inference.
• Scripting and automation may use Python, Bash, Go, or Ansible.

Experience and qualifications

• Substantial experience across GPU infrastructure, HPC, or large-scale compute platforms.
• Significant involvement in building, scaling, or managing GPU-intensive infrastructure.
• Ability to translate complex infrastructure needs into reliable production systems and lead multidisciplinary engineering teams.

Workplace

• The posting describes the work arrangement as on-site/hybrid.

Compensation

• Competitive remuneration package reflecting the seniority and technical expertise required; no compensation cadence or amount is specified.