Hardware Architect

The Opportunity

Fossefall is building multi-datacenter AI infrastructure, and the decisions made about compute and storage hardware today will shape how effectively our AI factories can scale and operate in the future.

We are looking for a Hardware Architect to take technical ownership of everything that goes into the rack, from GPU platforms and servers to the high-performance storage systems that support them. You will help define the hardware architecture for new sites, lead the technical relationship with key vendors and take delivered infrastructure through installation, validation and acceptance into production.

This is a founding role within the hardware function. You will establish the technical standards for the domain and, over time, help shape the hardware and storage capabilities around you. Working closely with our Network Architect, Platform team, project organisation and operations colleagues, you will make sure our infrastructure is designed as one coherent system rather than a collection of individual components.

What you'll do

- Own the architecture for GPU systems, servers and high-performance storage across Fossefall's current and future datacenters.
- Define rack-scale compute architecture, initially including NVIDIA GB300 NVL72-class systems, and establish how the platform should evolve as capacity and the number of sites increase.
- Lead the requirements, benchmarking and selection process for high-performance storage, balancing performance, capacity, parallel file systems and object storage.
- Act as the principal technical contact for hardware vendors, covering configuration, design reviews, technical terms, support arrangements, delivery and hardware replacement processes.
- Represent hardware requirements in datacenter planning, including power density, liquid cooling, floor loading and rack layout.
- Define repeatable processes for hardware delivery, installation, burn-in, acceptance testing and handover to operations.
- Partner with the Network Architect to define NIC and DPU strategy, storage networking, cabling architecture, rack elevations, and the end-to-end relationship between compute, storage, and network platforms.
- Provide clear technical recommendations on GPU, server and storage technologies, including emerging roadmaps and their practical implications for Fossefall.
- Establish technical standards, documentation and ways of working for the hardware domain.
- Help define the future hardware and storage functions and participate in building the teams as Fossefall grows.

About you

You bring significant experience designing, deploying or operating server and storage infrastructure at datacenter scale. You have taken a substantial hardware deployment from vendor evaluation and selection through delivery, validation and production use.

You combine deep technical understanding with practical judgement. You can move comfortably between architecture discussions, vendor design reviews, performance benchmarks and hands-on work in the datacenter. You understand that the strongest design is not simply the most technically ambitious one, but the one that can be operated reliably, supported effectively and repeated as the organisation scales.

We believe you will bring:

- Hands-on knowledge of modern server hardware, including CPU and GPU platforms, PCIe and NVMe topologies, memory, firmware lifecycle and hardware management interfaces.
- Strong experience with high-performance storage at scale, including real-world benchmarking, performance analysis and tuning.
- Understanding of the physical requirements associated with high-density infrastructure, including power, liquid cooling and rack density.
- Experience working directly with established hardware vendors and contributing to the technical side of significant hardware purchases.
- The ability to explain complex designs and trade-offs clearly to project managers, facilities teams, executives and commercial stakeholders.
- A practical, structured approach to turning technical decisions into standards, documentation and repeatable implementation.
- The desire to build a technical function from the ground up and the confidence to work effectively while some structures and responsibilities are still developing.
- Clear communication skills in English. Nordic language skills are a bonus.

Experience in one or more of the following areas would be valuable:

- GPU platforms at scale.
- High-performance storage technologies such as parallel file systems, NVMe over Fabrics and object storage.
- AI storage data paths, including GPUDirect Storage, checkpointing and data loading for training or inference workloads.
- Liquid-cooled, high-density rack deployments.
- AI or HPC cluster management and validation tooling.
- InfiniBand, RoCE and storage networking.
- Automation for hardware provisioning, firmware management, inventory or datacenter infrastructure management.
- Hyperscale, neocloud or large-scale GPU infrastructure environments.
- Working as an early technical hire and helping build a new engineering function.

We do not expect one person to have worked with every platform or technology listed. What matters most is a strong foundation in datacenter-scale compute and storage, credible hands-on experience and the ability to evaluate unfamiliar technologies with sound technical judgement.

What success looks like

Within the first few months, Fossefall has a clear and accountable technical owner for compute and storage hardware across its datacenter programme. The GPU platform and high-performance storage architecture have been evaluated and defined, with the resulting requirements reflected in network, power, cooling and rack-level decisions.

Vendor configurations and technical commitments are supported by structured design reviews, testing and clear acceptance criteria. Hardware delivery, installation, burn-in and production handover follow a documented and repeatable process that can be reused as new halls and sites come online.

Colleagues across Platform, network, project delivery, facilities, operations and procurement understand when and how to involve the role. The initial technical standards for the hardware and storage domains are in place, and Fossefall has a practical foundation on which to build the future function and team.

Why Fossefall?

Fossefall is building for long-term growth, which means many important ways of working are still being shaped. Joining now means having the opportunity to influence how the organization grows, how teams collaborate and how high-quality work gets done. You will work close to capable colleagues, contribute to meaningful priorities and help build something that can last.

About Fossefall

Fossefall is a Nordic AI infrastructure company founded in 2025. We develop and operate AI factories powered by renewable energy, turning clean Nordic power into scalable, sovereign computing capacity for Europe, built in partnership with local communities, energy producers and technology leaders across the world. As a team, we are still small enough to know each other as people, and our culture is shaped by ownership, curiosity, warmth and collaboration.

Interested?

Make sure your CV is up to date and tell us briefly why this opportunity is interesting to you and how your experience connects to what we are looking for. As part of your application, we would also like you to briefly describe a GPU or storage deployment that you helped design or bring online, including what you would do differently today.

We review applications as they come in and keep the role open until we find the right colleague. Due to the nature of our business, background checks and/or security clearance may be required as part of the employment process.