We are seeking a hands-on HPC / InfiniBand Engineer to deploy, validate and support rack-scale GPU compute environments. This is a full-time, 3-month contract, delivery-intensive role in a demanding project environment, requiring someone comfortable working hands-on through build, bring-up, testing and issue resolution.
Key responsibilities
- Deploy and validate GPU servers and rack-scale compute environments, including liquid-cooled systems.
- Configure, diagnose and troubleshoot InfiniBand fabrics, including subnet management, optics, cabling, switches and link-level issues.
- Bring up and validate compute/fabric environments, including connectivity, RDMA and performance testing.
- Run acceptance and stress testing, and troubleshoot hardware, connectivity, firmware, driver and performance issues.
- Administer Linux systems, including OS configuration, drivers, firmware updates and diagnostics.
- Produce clear documentation, test results and escalation notes.
- Work closely with infrastructure, networking, vendor and data-centre teams during delivery and support.
Required experience
- Hands-on experience with GPU server and rack-scale deployments, ideally including liquid-cooled systems.
- Strong InfiniBand knowledge, including subnet management, fabric diagnostics, optics, cabling and performance troubleshooting.
- Experience with firmware, drivers and Linux administration.
- Strong root-cause-analysis skills across hardware, networking, configuration and workload layers.
- Comfortable working in a fast-paced, delivery-focused project environment.
Highly desirable
- NVIDIA GB300, GB200, DGX, HGX or comparable high-end GPU platform experience.
- Experience with NVIDIA InfiniBand switches, ConnectX adapters, NVLink/NVSwitch and fabric-management tooling.
- Exposure to RDMA, NCCL, MPI, GPUDirect and distributed AI/HPC workloads.
Contract details
- 3 Month Contract for UAE-based candidates or ability to relocate at short notice for duration of project
- Focused on delivering clear milestones across key project phases
- Ideal for an experienced specialist comfortable working in a fast-paced, multi-stakeholder environment
- Visa and Private Healthcare provided
This role suits a pragmatic engineer who enjoys working directly with complex infrastructure and can take ownership from initial deployment through to a stable, validated production environment. Comparable NVIDIA roles emphasise deployment, validation, Linux administration, InfiniBand diagnostics and end-to-end troubleshooting across GPU and networking systems.
Your consultant
Michael Deane, Senior Recruitment Consultant specialising in C-Suite, VP and senior-level appointments, as well as technical and business development talent, across Enterprise technology in the EMEA region.