Stage 1: From manifesto to concept
The Spark by Snohetta
The building is a hybrid data centre and community sports facility located in Norway, integrating public and private uses while leveraging the heat generated by the data centre to warm the surrounding programs like a pool, spa, and housing. What caught my attention wasn't just the sustainability model, but the architectural expression of the data centre itself - a stark, black monolithic volume embedded at the core of an otherwise transparent and open structure. This centrepiece felt powerful, almost sacred. There’s something almost religious about the way it is framed: symmetrical, elevated, and slightly withdrawn, as if the entire building is organised around this opaque black box. I was drawn to this sense of reverence - a technological altar - and experimented with this concept in some of my early design sketches. Whilst this form didn't carry through, it still influenced how I thought about hierarchy, opacity, and symbolism in architectural form especially the notion of a monument and how it related to my dissertation. This is an idea that I circle back to naturally in my thought process to end up with such a monumental and striking end building.
At this point in the project, I spent time fully rounding out the narrative - something that proved incredibly useful last year and has again helped me clarify the direction of the design. After discussions around AI, ethical ambiguity, and the idea of mediating machines, I began to see the building less as a static monument and more as a force in tension with the landscape. I started to imagine the project as something fragmented, latent - building toward a moment of drama. A presence that pierces through the terrain, not simply placed on the land but emerging from it. There’s an inevitability to it, as if the architecture was always meant to rise from this place. The landscape itself becomes a character in the story - referencing my dissertation’s discussion of Actor-Network Theory (ANT), where site, infrastructure, and technology are all in dialogue.
I realised my dissatisfaction with earlier symmetrical forms came from their overly resolved, monumental feel. The “two towers” idea was too explicit — it risked becoming a symbol without depth. Instead, I started to explore forms that feel like they’re breaking through the surface. I’m still questioning whether this rupture happens on the higher ridgeline or the flatter plain, but either way, I see booleans and subtractions playing a central role — cutting, slicing, exposing.
Stage 2: The idea of "Slumber"
At this stage, I was stuck. I had the narrative, I had the site - but I just couldn’t find a form that felt right. I’d tried trenching, carving, embedding, even referencing past work where I built into a dam - but nothing clicked. The mountain felt too dominant, and any object I placed at its base felt swallowed. When I explored the flatter terrain, it offered more presence, but the forms felt arbitrary or disconnected from the drama of the site.
One of my tutors described the blasted road cut through the 50m-high ridge as the site’s “big reveal.” That stuck with me. I kept circling back to the idea that this landscape had already been violently opened, and perhaps the architecture could echo that gesture - not just sit in the terrain, but slice it open further.
I had a concept for an AI research centre that was already exploring themes of containment, disruption, and visibility. I wanted the architecture to reflect those ideas - not just functionally, but tectonically. A trench made sense. A fracture made sense. But my earlier trench-based schemes spanned 500m or more - far too large, and not human in scale.
It all clicked when I looked at a natural dip in the ridgeline, roughly 50m wide. I realised I could insert a tall, thin volume into that dip - a piercing form - something narrow, deliberate, and almost ritualistic. It wouldn’t sit on the mountain. It wouldn’t be hidden inside. It would cut through it.
That became the turning point. I didn’t need to choose between building on the flat land or inside the mountain. I could bisect it — insert a structure that felt like a scar or a horn, rising from the ground like an AI obelisk. This move gave me the drama I was missing and became a literal and conceptual threshold - a bridge between landscape and machine, between past and future.
Stage 3: Slumber is now a building that bisects the ridgeline.
From this stage the site and overall concept has been determined. From now I am refining and making the concept a building.
Following a discussion with the structural engineer, it was determined that my project should have 4 cores not 3. Here is my design process for this change. I also thickened the walls up to a new build up I had designed.
I also spoke with an environmental engineer who gave me insight into my strategies and how effective they were.
On each, the circled numbered points are the points I went into the discussion wanting to talk about.
Progressing and looking at 1 bed accomodation units with private outdoor spaces, using space over the corridor and double heights to maximise open spaces and tuck away sleeping and private functions
I then decided it was more feasible to flip the underground portion. This allowed me to curve light into the space, allowed for more south light to then control and allowed for more height in the terrain to expand certain portions.
Then experimented with modelling my windows between fins idea. This started to work nicely as a concept after some tweaking with fin length and width whereby the windows are only need at very direct head on angles, anything else they become nearly invisible. Here are the results:
At this stage, I began to think more critically about the building in section. I didn’t want a simple, stacked ‘car park’ configuration where each level felt monotonous and disconnected. Instead, I started carving into the form, introducing variation between floors to create moments of spatial and visual connection. The narrow, vertical nature of the design became an opportunity to explore dynamic floor relationships - encouraging interaction, permeability, and a sense of vertical continuity throughout the building.
Following discussions with the environmental engineer, I became more confident in the building’s ability to harness natural light. Its slender footprint allows daylight to penetrate deep into the plan, meaning that I needed to consider how light wells and other features could be altered and directed. With AI this is simple, install mechanical systems which react such as louvres. Narrow glazing helps minimise glare while maintaining sufficient illumination - supporting both the building’s environmental performance and its internal atmosphere.
I also began to explore the integration of wellbeing-focused spaces - specifically, a “light spa” designed to combat Seasonal Affective Disorder, which is particularly relevant in Iceland’s extreme lighting conditions. These communal areas would be interspersed among the lab and accommodation floors, offering moments of relief and reflection. I was inspired by James Turrell’s immersive installations, where curved, seamless surfaces gently reflect and diffuse light- removing hard shadows and creating a soft, almost surreal spatial quality. I also really like the ramped section out of the building to the top. It felt like a waste not to allow people up this high and I love spaces that looked "carved out" of masses. Sinking this area down 1m means the buildings mass itself becomes the railing and this space looks seamless, ramping up on top of everything.
