Platform Architect - Nvidia AI/GB300

The key focus is NVIDIA HGX GB300 / NVL72 and RTX 6000 series servers, so proven experience designing and deploying NVIDIA GPU infrastructure is essential

You’ll translate complex technical requirements into HLDs, LLDs and platform architecture, working across infrastructure, networking, storage, security, applications and data-centre teams. Key Requirements

• Strong Platform / Infrastructure Architecture experience across compute, storage, networking and Linux.
• Expert-level Kubernetes architecture experience.
• Strong Slurm and Run experience - essential.
• Proven experience with GPU/HPC environments and large-scale AI platforms.
• Hands-on experience with NVIDIA HGX GB300 / NVL72, including NVLink, NVSwitch and Grace Blackwell architecture.
• Experience with NVIDIA RTX 6000 series GPU servers.
• Strong understanding of GPU workload scheduling, partitioning and sharing, including MIG, vGPU and time-slicing
• Strong understanding of InfiniBand, RoCE, Spectrum-X, GPUDirect RDMA/Storage and high-performance AI fabrics.
• Experience with Terraform, Ansible, Python/shell, Git and CI/CD

Highly Desirable

• Experience with NVIDIA AI Factory / GB300 OR GB200 reference architectures.
• Understanding of rack-scale liquid cooling, 100kW+ rack power densities, CDUs and data centre infrastructure.
• Experience working with colocation/facilities teams on high density GPU deployments.
• Knowledge of DGX BasePOD / SuperPOD architectures.