By Switch Datacenters | June 2026
The pressure on data center operators has never been greater. AI-driven workloads are multiplying. Power grids are strained. Customers demand capacity faster than traditional construction methods can deliver it. And regulators expect sustainability to be built in from day one — not bolted on afterwards.
At Switch Datacenters, we don’t wait for the industry to catch up. We build ahead of it.
That’s why, when we set out to expand our Amsterdam colocation capacity, we made a deliberate choice: go modular. Not modular as a buzzword, but modular as a fully engineered, factory-built, repeatable platform — one designed from the outset to scale across multiple sites, not just solve a single project.
On 25 June 2026, our Chief Construction & Engineering Officer Geurt Woudenberg and Chief Sustainability Officer Cara Mascini joined our partners Fredrik Nordling (CEO, NordicEPOD) and Konstantin Santikos (Head of Energy Systems Programs, Eaton) for a broadcasted session on Data Center Dynamics’ Building at Scale Europe broadcast to discuss exactly how we did it — and where it’s taking us next.
Here’s what we shared.
Why Modular — And Why Now
The honest answer is that traditional builds can no longer keep pace with demand. A conventional data center project often runs to 24 months or more. It requires complex on-site engineering, a large physical workforce, and layers of sequential testing before you can bring a single watt of capacity online. In a market where power availability is already constrained and customers need certainty, that timeline is simply too slow.
Modular infrastructure changes the equation. By moving assembly, integration, and testing into the factory, you compress the on-site delivery window dramatically, reduce the number of variables on the day of commissioning, and — crucially — create a repeatable blueprint that improves with every deployment.
For this project, we worked with NordicEPOD and Eaton to develop fully integrated, factory-built power modules for our Amsterdam colocation facility. NordicEPOD delivered the pods in approximately 18 weeks from design sign-off to site arrival. That pace isn’t possible with conventional methods. It is possible when you’ve done the hard engineering work upfront, in the factory, where conditions are controlled and quality is consistent.
What ‘Repeatable Without Losing Flexibility’ Actually Means
One of the questions we hear most often when we talk about modular design is: doesn’t standardisation mean compromise? Don’t you sacrifice the ability to meet specific customer requirements in exchange for speed?
The answer is no — if you design the platform correctly.
Our approach was to define the non-negotiable requirements first: power density, redundancy levels, cooling architecture, sustainability performance, and end-of-life circularity. Once those were locked in as design constants, we could build flexibility into everything else — the configuration of pods, the sequencing of deployments, the adaptability to site-specific constraints.
The result is a design that started as a client-specific solution and has evolved into a standardised, repeatable framework that can be deployed across multiple sites without starting from scratch each time. That’s not a loss of flexibility. That’s what genuine scalability looks like in practice.
The Role of Partners: Eaton and NordicEPOD
A modular platform is only as strong as the ecosystem behind it. That’s why the partnership with Eaton and NordicEPOD was central to how this project came together — not as a vendor relationship, but as a genuine design collaboration.
For Eaton, this deployment marked the first time their full power ecosystem was integrated into a single pod: UPS, power distribution, switchgear, and monitoring, all factory-assembled and tested as a unified system rather than as individual components installed sequentially on-site. The significance of that integration is hard to overstate. It means the module arrives at the facility already proven. The risk that traditionally lives on-site — integration failures, commissioning delays, unexpected incompatibilities — is addressed before the module ever leaves the factory.
NordicEPOD brought the pod architecture itself: the structural design, the engineering of transport and deployment logistics, and the discipline to move from concept to delivered hardware in a compressed timeline. Their challenge — and they met it — was to design pods that could be built quickly without compromising the quality and consistency that Switch Datacenters’ customers expect.
The feedback loop between our team and our partners during the design phase was intensive and necessary. Geurt’s operational requirements influenced the final architecture. Cara’s sustainability criteria shaped material choices and end-of-life planning. That kind of collaborative iteration is what turns a one-off pilot into a replicable platform.
Sustainability: Designed In, Not Added On
Modularity and sustainability are not separate conversations. They are the same conversation.
Factory-built construction reduces on-site activity, which directly lowers the environmental impact of the build phase — relevant in the Dutch regulatory environment, where constraints around nitrogen emissions and local impact assessments are real operational challenges. Fewer on-site trades, shorter construction periods, and tighter logistics control all contribute to a cleaner build footprint.
Beyond the construction phase, our modular design incorporates circular economy principles from the outset. The question of what happens to a module at the end of its operational life is not an afterthought — it is a design constraint. Components are specified with refurbishment and reuse in mind, aligned with EU sustainability directives and the expectations of customers and investors who are increasingly holding us accountable to those standards.
Cara put it directly during the session: sustainability is no longer a differentiator in isolation. It is becoming a baseline expectation. Operators who build it in from day one — into the design, the supply chain, the partner relationships — will be better positioned than those who treat it as a compliance exercise.
We treat it as a design requirement. That shows up in our PUE figures, our heat reuse programmes, and now in our modular deployment approach.
From Pilot to Platform: What Comes Next
The Amsterdam deployment was the proof of concept. It demonstrated that a fully integrated, factory-built power module can be delivered at pace, commissioned cleanly, and operated reliably within a live colocation environment.
But the point was never just to prove it once. The point was to build something we could do again — faster, with fewer adjustments, across more sites.
That is what a platform means. The lessons from this deployment — on design, on factory integration, on logistics, on site commissioning — are already informing how we approach future projects, including our expansion into Poland and the broader pipeline of capacity we are developing across Europe.
As AI workloads continue to drive demand for higher power density and faster deployment timelines, the operators who have built a repeatable delivery model will have a significant structural advantage. We are building that model now — deliberately, with the right partners, and with sustainability embedded at every stage.
Watch the Full Session
The complete 45-minute broadcast — From Pilot to Platform: How Switch Datacenters Is Shaping Repeatable Modular Data Center Delivery — is available on demand via Data Center Dynamics’ Building at Scale Europe 2026 series.
Watch the broadcast on DataCenterDynamics.com →
Want to Know More?
If you’d like to understand how our modular platform could support your infrastructure requirements — whether you’re planning a new deployment, scaling existing capacity, or exploring what AI-ready colocation looks like in practice — we’d be glad to talk.
