Manager Systems Engineering (Ai Hpc)

Iren; Sydney; Hybrid; AUD 200,000 - 260,000; Full time

Job summary

IREN is seeking an engineering leader for the Deployment team, guiding ~12 HPC systems engineers who design and build compute, storage and network infrastructure for large-scale GPU clusters. You will own end-to-end delivery, review designs, and balance reliability, cost, and time to capacity while staying hands-on enough to mentor and review incidents.

Reporting to the Director of Cloud and Network Engineering, you will shape the roadmap for schedulers (Slurm/Kubernetes), file systems,

Qualifications

• 3+ years managing engineers, including teams of 8+
• 5+ years in HPC, large-scale Linux infrastructure
• Hands-on depth in schedulers, file systems, fabrics, GPU node operations
• Experience planning hardware capacity with data center, facilities and procurement
• Clear written and verbal communication with engineers, executives, vendors, customers
• Ability to make decisions with incomplete information and explain reasoning

Responsibilities

• Lead the Team: Manage, coach, and grow 12 engineers with 1:1s and career plans
• Set technical direction: shape roadmap for scheduling, file systems, fabrics, image pipelines
• Take ownership of project delivery: end-to-end delivery, realistic commitments, escalate blockers
• Represent team in architecture forums and translate needs into priorities
• Collaborate with data center ops, networking, platform software, security, and customers

Skills

People management

HPC expertise

Scheduling systems

Vendor management

Tools

Slurm

Lustre

GPFS

NVIDIA drivers

NCCL

DCGM

Kubernetes

Job description

Job Type: Full-time|Location: Sydney, Australia|Department: Information Technologies (IT) | Reporting to: Director, Cloud and Network Engineering |Work Location: Hybrid

IREN is a leading AI Cloud Service Provider, delivering large-scale GPU clusters for AI training and inference. IREN’s vertically integrated platform is underpinned by its expansive portfolio of grid-connected land and data centers in renewable-rich regions across the U.S, Canada (existing) and Australia and Spain (in the pipeline).

With 100% renewable energy, we build, own and operate our data centers and take pride in being at the forefront of sustainable solutions for the ever-evolving applications of high-performance compute. We believe that human progress is invaluable, but it should be done in the right way – responsibly, sustainably and having a positive impact on the communities we operate in.

About the role

The Deployment team takes GPU clusters from racked hardware to a validated, handed-over system that a customer can run production AI workloads on. Our current estate is NVIDIA GB300 NVL72, B300 and B200 on NDR and XDR InfiniBand and Spectrum-X fabrics, at the scale of thousands of GPUs per site, across multiple sites, with new sites and takeovers from vendor-managed builds arriving every quarter.

You will lead a team of approx. 12 HPC systems engineers who design and build the compute, storage, and network infrastructure behind our large-scale GPU and HPC clusters. The team owns the delivery of the platform: cluster bring-up, scheduler and job management, parallel storage, high-speed interconnects, OS and firmware baselines.

This is a hands-on engineering management role. You will spend most of your time on people, priorities, and delivery, but you need enough technical depth to review a design, challenge an incident root-cause, and make trade-offs between reliability, cost, and time to capacity. You report to the Director of Cloud and Network Engineer and partner closely with data center operations, networking, platform software, security, and customer-facing teams.

What you’ll do

Lead the Team

• Manage, coach, and grow 12 systems engineers across junior to principal levels, including regular 1:1s, performance reviews, and career development plans.
• Hire and onboard engineers as the platform grow and build a bench of technical leads who can own major subsystems without you.

Set technical direction

• Shape the roadmap for scheduling (Slurm or Kubernetes), parallel file systems, InfiniBand or RoCE fabrics, image and configuration management.
• Drive automation so that cluster builds, node lifecycle, and firmware updates are repeatable and tested rather than hand-run.
• Review designs and major changes, and arbitrate trade-offs between reliability, cost, security, and speed of delivery.
• Represent the team in architecture forums and translate customer and research-user needs into engineering priorities.

Take ownership of IREN Project Delivery

Delivering projects on time is the single most important measure of this role. Capacity that lands late is revenue IREN cannot recognise and customers it cannot serve, so this is where you will be judged first.

• Own end-to-end delivery of the team's projects
• Set realistic commitments, then hold them. When a date is at risk, speak up with options and a recommendation, not after the deadline passes.
• Clear blockers fast, whether that means escalating a vendor, re-sequencing work across the team, or pulling in help from other IREN teams.

What you bring

• 3+ years managing engineers, including at least one team of 8 or more, with a track record of hiring, developing, and retaining strong people.
• 5+ years in HPC, supercomputing, or large-scale Linux infrastructure, with hands-on time before moving into management.
• Working depth in at least three of: Slurm or similar schedulers, parallel file systems (Lustre, GPFS, Weka, VAST), InfiniBand or RoCE fabrics, GPU node operations (NVIDIA drivers, NCCL, DCGM), configuration management and image pipelines, and observability at cluster scale.
• Experience planning hardware capacity and working with data center, facilities, and procurement teams on delivery timelines.
• Clear written and verbal communication with engineers, executives, vendors, and customers, including the ability to deliver bad news early.
• Comfortable making decisions with incomplete information and explaining the reasoning afterwards.

Nice to have

• Bringing up GPU clusters of 1,000+ accelerators, or multi-site HPC environments.
• Kubernetes for AI workloads alongside traditional batch scheduling.
• Multi-tenant or customer-facing HPC, including SLAs, chargeback and security boundaries between tenants.
• Performance engineering: MPI, NCCL tuning, storage I/O profiling or fabric topology design.
• Exposure to SOC 2 or ISO 27001 as they apply to infrastructure teams.
• Growing a team through a period of rapid expansion.

Why IREN

• Scale that few teams see. Thousands of the latest NVIDIA GPUs per site, across multiple sites, with new capacity landing every quarter.
• Real ownership. Your team owns the platform end to end, from bring-up to hand-over, and your delivery is directly tied to revenue and customer outcomes.
• Sustainable by design. 100% renewable energy across data centres we build, own and operate. We believe human progress should be done responsibly and with a positive impact on the communities we operate in.
• Room to grow. A fast-expanding platform means a fast-expanding team, with the chance to build your own bench of technical leads.