Lead GPU HPC Architect for Scalable AI Infra

Hamilton Barnes; United Kingdom; Remote; GBP 140,000 - 160,000; Full time

Benefits offered by this job

4-day working week

Remote-first working

Uncapped holidays

Remote flexibility

Job summary

Hamilton Barnes is seeking a Lead Technical Architect to design large-scale GPU clusters for AI and HPC engagements. You will craft enterprise-grade designs from workshops to PoCs, guiding both engineering teams and sales through RFP responses.

The role offers a 4-day workweek, remote flexibility, and a highly autonomous environment focusing on cutting-edge GPU, InfiniBand, and storage technologies.

Qualifications

• 5+ years in Solutions Architecture, Systems Engineering or Technical Pre-Sales in HPC or AI infra.
• Deep hands-on knowledge of NVIDIA HGX/DGX and Blackwell platforms.
• Expert InfiniBand and RoCE networking knowledge; GPUDirect RDMA and storage.
• Kubernetes and bare-metal orchestration incl MPI Operator.
• UK based requirement.

Responsibilities

• Author enterprise-grade High-Level and Low-Level Design docs for large-scale GPU clusters.
• Produce detailed Bills of Materials for compute, NVLink, fabrics, cabling, cooling, power and storage.
• Architect scale-up and scale-out topologies across InfiniBand and RoCEv2 with tuning.
• Deliver multi-tenant architectures, integrated with GPUDirect Storage and HPC file systems.
• Lead technically alongside sales; run architectures workshops and respond to RFPs/RFIs.
• Design and oversee PoCs, benchmarking with NCCL tests, GPUDirect RDMA and MLPerf/Megatron-LM.

Skills

Solutions Architecture

Systems Engineering

Technical Pre-Sales

Kubernetes

MPI Operator

Executive communication

Education

Bachelor's degree in Computer Science / Electrical / Systems Engineering

Tools

Slurm

Ansible

Terraform

PyTorch NCCL

GPUDirect RDMA

Job description

Hamilton Barnes is seeking a Lead Technical Architect to design large-scale GPU clusters for AI and HPC engagements. You will craft enterprise-grade designs from workshops to PoCs, guiding both engineering teams and sales through RFP responses.

The role offers a 4-day workweek, remote flexibility, and a highly autonomous environment focusing on cutting-edge GPU, InfiniBand, and storage technologies.