Lead AI GPU Compute Cluster Architect

Radiant; Greater London; On-site; GBP 120,000 - 190,000; Full time

Benefits offered by this job

25 days leave

Medical insurance

Cycle to Work

Gympass

Share program

Parental leave

Job summary

Radiant in the United Kingdom seeks a Cluster Architect to lead low-level design of AI HPC deployments and translate high-level requirements into production-grade infrastructure. You will collaborate with the Infrastructure Solutions and Engineering teams to define reference architectures, drive RFx evaluations, and ensure scalable, reliable GPU-based systems for demanding workloads.

The role emphasizes documentation, vendor engagement, and hands-on delivery to turn designs into deployable

Qualifications

• BSc in Computer Science, Electrical/Computer Engineering, Physics, Mathematics, or other related Engineering fields.
• 5+ years of solution architecture experience.
• Deep expertise in datacenter engineering (mechanical, electrical and plumbing), GPU systems, high performance networking (InfiniBand), including a proven understanding of network topologies and storage architecture.
• Proficiency in system-level aspects, encompassing Operating Systems, Linux kernel drivers, GPUs, NICs, and infrastructure software solutions.
• Demonstrated experience producing detailed infrastructure designs (including network topologies, rack layouts, BoMs, power/cooling planning…).
• Deep understanding in datacenter fit‑out, and power/cooling constraints
• Excellent interpersonal and communication skills, with a track record interfacing with various types of stakeholders and writing clear deliverables.
• Comfortable working in fast-moving, high-complexity technical environments.

Responsibilities

• Translate inputs into Radiant’s reference architecture and low-level designs blueprints.
• Remain abreast of technology advancements and roadmaps with key vendors in the space.
• Lead the low-level design of high-density AI supercomputer systems (compute, interconnect, storage, management).
• Participate in the Architecture Review Board to ensure collaboration and alignment behind designs.
• Evaluate tradeoffs between performance, power, cooling, serviceability, and cost throughout the procurement process.
• Drive creation of detailed low-level designs, including rack elevations, data center floor plans, power distribution and BoM.
• Drive vendor evaluations during RFx processes to ensure designs meet capabilities and long-lead item constraints.
• Hands‑on role in manufacturing, burn‑in, cabling, configuration, benchmarking, and implementation of new deployments.
• Define acceptance criteria with quality, reliability, safety, and performance goals in mind.
• Work with Software/solutions engineering teams to ensure infrastructure supports AI workloads.
• Produce architecture diagrams, runbooks, and operational guidance for handover to Infrastructure Operations.
• Establish delivery best practices and share lessons learned.

Skills

Datacenter engineering

GPU systems

High performance networking

System-level engineering

C/C++ programming

CUDA

Education

BSc in Computer Science, Electrical/Computer Engineering, Physics, Mathematics, or other related Engineering fields

Tools

Kubernetes

Docker

Slurm

Job description

Radiant in the United Kingdom seeks a Cluster Architect to lead low-level design of AI HPC deployments and translate high-level requirements into production-grade infrastructure. You will collaborate with the Infrastructure Solutions and Engineering teams to define reference architectures, drive RFx evaluations, and ensure scalable, reliable GPU-based systems for demanding workloads.

The role emphasizes documentation, vendor engagement, and hands-on delivery to turn designs into deployable