HPC Cluster Architect

TL;DR

HPC Cluster Architect (NVIDIA GPU/HPC): Own end-to-end cluster architecture for large-scale dedicated GPU deployments from customer requirements to production handover with an accent on compute, networking, storage, and physical design. Focus on designing high-performance network fabrics, engaging with vendors, and providing technical oversight during deployment and validation.

Location: UK (Remote)

Company

hirify.global is behind Hyperstack, a full-stack AI cloud delivering on-demand and private GPU infrastructure to AI researchers and enterprises running compute-intensive workloads.

What you will do

• Own end-to-end cluster architecture for large-scale NVIDIA GPU deployments, including rack layouts, BOM, power, cooling, and production handover
• Design high-performance network fabrics across compute (InfiniBand, RDMA, NVLink/NVSwitch), storage, and WAN with topology and scaling strategies
• Engage directly with OEMs and vendors to validate hardware, review quotes, and optimize designs commercially
• Provide technical oversight during deployment, hardware validation, performance testing, and issue escalation
• Act as senior technical leader across Solutions Architecture, Cloud Engineering, and data centre partners to build standardized reference designs

Requirements

• Proven experience designing and delivering GPU-based HPC or AI clusters at scale across full lifecycle
• Deep hands-on knowledge of NVIDIA GPU platforms (H100/H200/B-series) and reference architectures
• Strong InfiniBand/RDMA design experience including topology and performance tuning
• Solid grounding in Linux systems, PCIe topology, NUMA alignment, and server performance
• Background from OEM, hyperscaler, neo-cloud, or HPC environment with full design-to-deployment exposure
• Confident engaging with customers, vendors, and teams as technical authority

Nice to have

• Experience with Spectrum-X or next-generation Ethernet fabrics
• Prior involvement in 1,000+ GPU cluster deployments and benchmarking (NCCL, MLPerf)
• Exposure to air-cooled/liquid-cooled HPC and infrastructure-as-code

Culture & Benefits

• Competitive salary and annual discretionary bonus
• Employee wellbeing benefits and 25 days holiday plus public holidays
• Flexible working (remote or hybrid depending on role and location)
• Real ownership, autonomy, and career progression in a fast-growing company
• Collaborative international culture with trust, transparency, and mission-driven colleagues