High‑Performance Computing (HPC) Security Engineer
* Job Title: High Performance Computing (HPC) Security Engineer
* Employment Type: Contract (Freelance)
* Clearance Required: SC-Cleared OR DV-Cleared (preferred)
* Duration: 12 months initial → 2-3-year extensions already budgeted
* Start Date: ASAP
* Location: United Kingdom
Role Summary
This is a "Tier 1" National Security High-Performance Computing (HPC) Contract. Build, optimise and performance-tune exascale-class GPU clusters that run the countries most sensitive workloads (think cryptanalysis, nuclear stockpile modelling, real-time ISR processing, and sovereign AI training). Lock those same systems down to the strictest UK government security standards so they survive and remain auditable in a full zero-trust, high-side environment, doing it all on hardened Linux at extreme scale.
IMPORTANT
This is not a "Linux admin with some CUDA experience" role.
This is not a "Cyber analyst who once logged into a cluster" role.
This role is for UK-Cleared SMEs who have had experiences at the intersection of raw High-Performance Computing (HPC), deep Linux systems mastery, and genuine National-Security-grade Cyber Defence.
Day‑to‑day you will:
* Profile and refactor classified codes to hit 85-95 % of roofline on H100/H200/B100 and MI300X/MI350 GPUs.
* Design and enforce security controls that survive NCSC penetration tests at supercomputing scale.
* Write or approve the actual SELinux policies, audit rules, and secure boot configurations that get deployed across thousands of nodes.
* Lead incident response and forensic analysis if (when) something anomalous appears on a high-side cluster.
* Automate secure DevSecOps pipelines that work behind cross-domain guards and data diodes.
* Brief 1-4-star officers and SCS/SES civilians on performance gains, risk posture, and future roadmap.
* Influence multi-billion-pound future procurements (you will see the requirements documents before most of industry does)
* Explain (in plain English) why your latest kernel patch or CUDA kernel optimisation just delivered a 38 % real-world speedup on a workload that directly affects national security outcomes.
* Prove that the system is still fully compliant with NCSC CPF and JSP 604.
Essential Requirements
* Active UK SC or DV Clearance (DV-Cleared Preferred).
* Proven overlap of High-Performance Computing + Cyber Security + Linux.
* Ability to work onsite 3-4 days/week in secure facilities.
* High-Performance Computing (HPC):
o Large-scale GPU/CPU cluster optimisation (SLURM, PBS Pro, Bright Cluster Manager, NVIDIA Base Command).
o Code tuning with CUDA, HIP/ROCm, SYCL, Kokkos, OpenMP 5.x+ offload, MPI.
o Delivering >80-90 % of theoretical peak on classified workloads
* Cyber Security:
o Hardening HPC environments to NCSC CPF, JSP 604, HMG IA standards.
o Zero-trust implementation, STIGs, CIS benchmarks, secure boot, audit at supercomputing scale.
o Vulnerability management and secure patching strategies in air-gapped networks.
* Linux Systems Mastery:
o Expert-level RHEL 8/9, AlmaLinux, Rocky or hardened Ubuntu in restricted environments.
o Kernel tuning, filesystems (Lustre, GPFS, BeeGFS), security-enhanced Linux (SELinux mandatory profiles).
o Secure containerisation with Apptainer/Singularity on high-side networks.
Highly Desirable (fast-tracks you to interview)
* Active DV Strap or Enhanced Developed Vetting.
* Recent/relevant experience at Raytheon, DSTL, GCHQ, NCSC, Porton Down, Boscombe Down or relevant programmes.
* Certifications: CISSP, CISM, CCP (any level), RHCE, CompTIA Security+.
* Background in cryptanalysis, hydrocodes, SAR processing or large-scale secure AI training.
Next Steps
If interested, please send your most recent CV to my email so that we can speak in more detail.
I will put a 15-minute meeting into our diaries as a reminder. Feel free to suggest a different time/day if the one I have selected does not suit you.
#HighPerformanceComputing #HPC #Linux #Security #Cyber #CyberSecurity #Contract #Freelance
#J-18808-Ljbffr