AI/HPC Consultant

An exciting opportunity has arisen to join Computacenter's consultancy practice, helping enterprise customers design and deliver cutting-edge NVIDIA-based AI and HPC infrastructure solutions. This role focuses on high-performance networking, AI cluster architecture, InfiniBand, Spectrum-X, RoCEv2, and physical infrastructure design across GPU-accelerated environments. Working closely with customers, vendors and internal stakeholders, you will provide technical leadership across architecture, design assurance and infrastructure optimisation., * Design High-Level and Low-Level Designs (HLDs/LLDs) for NVIDIA AI/HPC infrastructure, including GB300, B300, DGX, HGX and NVL72 environments.

• Design and validate InfiniBand backend fabrics, Spectrum-X front-end networks and Spectrum-X Ethernet backends.
• Provide technical expertise across RoCEv2, InfiniBand, VXLAN EVPN, BGP, leaf-spine architectures and high-performance networking.
• Create, review and optimise Bills of Material (BoMs) covering switches, optics, fibre assemblies, DAC/ACC/AEC cabling, adapters, DPUs and licensing.
• Run customer discovery workshops, requirements gathering sessions, architecture governance reviews and technical design discussions.
• Produce diagrams, port maps, cable matrices, design documentation and implementation guidance.
• Act as a technical lead, collaborating with customers, vendors, consultants, pre-sales teams and delivery stakeholders.

• Strong Network Architecture or Data Centre Networking background with experience in HPC, AI Infrastructure or low-latency network fabrics.

• Hands-on experience with InfiniBand, RoCEv2, Spectrum-X Ethernet, congestion management and lossless Ethernet technologies.

• Strong understanding of VXLAN EVPN, BGP, leaf-spine design, routing, resiliency and modern data centre networking principles.

• Practical knowledge of optical and copper connectivity including SMF/MMF, MPO/MTP, DAC, AEC/ACC, QSFP, QSFP-DD, OSFP and related technologies.

• Experience interpreting NVIDIA reference architectures and translating them into customer-ready designs and commercially accurate Bills of Material. *Strong stakeholder engagement, consultancy and technical documentation skills. Additional Information:

• Experience with NVIDIA DGX, HGX, GB300, B300, NVL72, ConnectX and BlueField/SuperNIC technologies is highly desirable.

• Knowledge of physical infrastructure design including rack layouts, switch placement, cabling strategy, cooling and power considerations is preferred.

• Exposure to liquid cooling, storage integration, management fabrics and out-of-band networking is beneficial.

• Relevant certifications such as Cisco CCNP/CCIE, NVIDIA Networking, Arista, Juniper, HPE or equivalent are desirable.

• Experience within Consultancy, Professional Services, Technical Architecture or Pre-Sales environments would be advantageous.