Linux HPC Architect

Innovate Recruitment Ltd; Stevenage; Hybrid; GBP 90,000 - 120,000; Full time

Benefits offered by this job

Hybrid work model

Personal development programme

International exposure

Job summary

Innovate Recruitment Ltd is seeking an HPC Architect to define the long-term roadmap for scalable HPC environments, including compute, storage, and interconnects. You will collaborate with multiple teams across national and international boundaries to ensure services meet evolving business needs.

DV clearance is required; BPSS checks after offer stage will be conducted. The role is hybrid with 2–3 days on-site and involves leading edge workloads in a dynamic, multinational setting.

Qualifications

• Experience designing HPC architectures and scalable storage.
• Strong understanding of high-throughput compute environments.
• Familiarity with performance benchmarking and capacity planning.

Responsibilities

• Define the long-term HPC architecture roadmap aligned with business goals.
• Architect large-scale HPC clusters including CPU, GPU, and accelerators.
• Design high-performance networking and storage architectures for HPC.
• Provide technical leadership to HPC engineering and operations teams.
• Lead performance benchmarking and workload modelling.

Skills

HPC architecture

Storage design

InfiniBand networking

SLURM orchestration

MPI/CUDA

Cloud HPC

Tools

Kubernetes

RDMA

CUDA

MPI

Job description

Linux HPC Specialist / HPC Architect

An exciting opportunity for a passionate HPC Architect to join a unique, multi-national Information Management function. Ideal candidates should be committed to the future of software-defined data centres and open source technologies, with the dedication to support our engineering community in a dynamic, fast-paced development environment.

Security Clearance | This role will require DV Clearance; due to this we can only consider sole British citizenship holders. As a minimum, and after offer stage, all successful candidates will need to undergo HMG Basic Personnel Security Standard (BPSS) checks.

Dynamic (hybrid) working: 2–3 days per week on-site due to workload classification

The opportunity:

We are seeking a forward-thinking HPC Architect with expertise in storage and infrastructure to define the long-term future roadmap of our HPC environments. You will ensure alignment with business objectives in collaboration with our business areas, and be responsible for building scalable, high-throughput infrastructure.

You will manage and maintain HPC environments including compute nodes, storage and interconnects, working closely with your team and wider stakeholders across the organisation. You will also engage with multi-national, multi-cultural teams to drive capability development cross-functionally.

• Become part of a well-respected team recognised for its professionalism, dedication and flexibility
• Work both nationally and internationally, broadening your horizons within a culturally diverse organisation
• Partner with engineers and HPC vendors to ensure HPC services meet evolving business needs
• Join a values-led organisation committed to investing in its workforce through a personal development programme

What we're looking for from you:

Architecture & Strategy

• Define the long-term HPC architecture roadmap aligned with organisational goals
• Evaluate and select technologies across compute, storage, networking, and cloud HPC
• Establish standards, reference architectures, and best practices

System Design & Engineering Oversight

• Architect large-scale HPC clusters, including CPU, GPU, and accelerator‑based systems
• Design high-performance networking (InfiniBand, RDMA, high-speed Ethernet)
• Define storage architecture (parallel file systems, object storage, tiered storage)
• Provide technical leadership and oversight to HPC engineering and operations teams

Performance & Scalability

• Ensure systems are designed for optimal throughput, latency, and scalability
• Lead performance benchmarking, capacity planning, and workload modelling
• Identify and eliminate architectural bottlenecks

Workload & Software Ecosystem

• Define strategies for workload orchestration (SLURM, Kubernetes for HPC, etc.)
• Guide software stack design (MPI, CUDA, compilers, libraries, containers)
• Ensure reproducibility, portability, and performance of workloads