Oxagon was supposed to be the world’s largest floating structure. An octagonal industrial platform in the Red Sea, hosting advanced manufacturing, logistics operations, research facilities, and an innovation campus. The rendering showed a vast geometric form sitting on the water’s surface — clean, symmetrical, and impossibly large. Bjarke Ingels Group, one of the world’s most celebrated architecture firms, designed the concept. NEOM’s promotional materials described it as the future of industrial development: zero-carbon, ocean-cooled, and connected to global shipping lanes.
As of the first quarter of 2026, no procurement activity has been recorded for the floating platform. No contracts have been awarded for floating components. No marine engineering has been commissioned. No floating structure of any kind has been built, tested, or prototyped at the Oxagon site. The floating city that was the defining concept of Oxagon — the element that distinguished it from every other industrial zone on every other coastline in the world — was quietly removed from the near-term programme without an announcement. It has been “pushed to the early 2030s” with no confirmed construction start date.
What exists at Oxagon is terrestrial. A port — 68 per cent complete — with more than four kilometres of quay wall and seven berths at depths of 10.5 to 18.5 metres, delivered by BESIX. Boskalis completed the port expansion dredging by March 2026. The container terminal, launching in 2026, will feature Saudi Arabia’s first fully automated Ship-to-Shore cranes — ZPMC is providing 10 STS gantry cranes, 30 electric rubber-tired gantry cranes, and 6 automated rail-mounted gantry cranes. The terminal stretches 900 metres with an 18.5-metre-deep marine basin. A second terminal is designed for up to 10 million TEUs annually. Total construction contracts awarded specifically for Oxagon reach $9.3 billion — making it the single largest recipient of NEOM contract spending after The Line. Full container operations at 1.5 million TEU annual capacity are scheduled for 2026. The NEOM Green Hydrogen plant — the $8.4 billion joint venture between NEOM, Air Products, and ACWA Power — is 80 to 90 per cent complete within the Oxagon zone. And in February 2026, NEOM announced a $5 billion partnership with DataVolt for an AI data centre campus at Oxagon.
The port works. The hydrogen plant works. The data centre is being built. None of these require a floating platform. None of them require an octagonal structure. None of them require the concept that gave Oxagon its name and its place in architectural discourse. The floating city has been replaced by a port, a chemical plant, and a server farm — useful infrastructure that could have been built on any stretch of Red Sea coastline without the prefix “floating” or the brand “Oxagon.”
The Engineering Question
The concept of a large-scale floating industrial platform is not inherently impossible. Floating structures exist at various scales: floating solar farms in Singapore and China, floating LNG terminals, semi-submersible oil platforms, and experimental floating architecture projects. The technology for floating structures is understood in principle and demonstrated at scales up to several hundred metres.
What has not been demonstrated is a floating structure at the scale Oxagon proposed. The octagonal platform would have needed to support manufacturing facilities, logistics operations, research laboratories, and potentially residential accommodation — all while maintaining structural integrity in the Red Sea’s tidal, current, and weather conditions. The engineering challenges include: mooring a structure of that scale in water deep enough to accommodate its draft; providing utilities (power, water, waste management) to a floating platform disconnected from terrestrial infrastructure; ensuring seismic and storm resilience; and managing the environmental impact of a massive floating industrial structure on marine ecosystems.
These challenges are not insurmountable. They are, however, extraordinarily expensive to solve, and the cost of solving them scales with the platform’s size in a non-linear relationship — each doubling of the platform’s area more than doubles the engineering cost. At the scale Oxagon proposed, the floating platform would have been the most complex marine engineering project ever attempted, exceeding the cost and complexity of any offshore oil platform, any floating LNG terminal, and any existing floating structure by orders of magnitude.
The decision not to procure floating components was, in engineering terms, the correct decision: the technology does not exist at the required scale, the cost of developing it would consume a significant fraction of NEOM’s remaining budget, and the functional requirements of the site — port operations, hydrogen production, data processing — are better served by terrestrial construction. The decision was correct. The announcement that preceded it — the announcement that created the brand, attracted the architecture firm, and generated the media coverage — was aspirational in a sense that the engineering profession would describe as fictional.
The BIG Design
Bjarke Ingels Group’s design for Oxagon was published widely in the architectural press. The renderings showed a vast octagonal form on the water, with modular units arranged in geometric patterns, connected by walkways and transport links, surrounded by clean ocean. The design was beautiful, geometrically precise, and — in the context of what has actually been built — an exercise in architectural imagination unconstrained by procurement, budget, or marine engineering reality.
BIG’s involvement in Oxagon has not been publicly revisited since the floating platform was deferred. The firm continues to be listed among NEOM’s architecture partners. Whether BIG’s contract covers only the floating platform concept or extends to the terrestrial development that is actually being built has not been disclosed. The firm has not commented publicly on the deferral.
The architectural press published BIG’s Oxagon renderings alongside analyses of the design’s formal qualities, its modular logic, and its potential to redefine industrial architecture. The renderings circulated in the same publications that, months later, would publish Dezeen’s investigation of architecture firms’ complicity in NEOM’s human rights conditions. The two stories — the beautiful floating city and the documented abuses of the workforce building it — coexist in the same media ecosystem, in the same publications, sometimes in the same issue. The contradiction is the profession’s, not the publications'.
The Pivot
The contracts awarded at Oxagon in 2025 and 2026 describe a site that has pivoted from architectural spectacle to industrial utility.
The $5 billion DataVolt partnership, announced in February 2026, is for a hyperscale AI data centre campus — a 1.5 GW facility using Red Sea seawater for cooling, designed for net-zero operations. The partnership represents the largest single new investment at NEOM since the construction suspension of The Line. It is a significant commitment. It is also a data centre — the most generic form of industrial infrastructure available, built on land, using proven technology, with a revenue model based on cloud computing demand rather than floating city tourism.
Hexagon, the Swedish-American geospatial company, secured a $2.7 billion contract for smart-city data infrastructure. The contract covers sensors, connectivity, and data management systems — the digital infrastructure that any modern industrial zone requires.
The pivot is from spectacle to substance. The data centre and the geospatial infrastructure have customers — cloud computing companies, AI workloads, and industrial data users — who exist today and whose demand does not depend on NEOM becoming a city. The floating platform had no customers because the customers were supposed to be created by the city that the platform was supposed to serve. The circular dependency — the same circular dependency that killed The Line, Trojena, and the Mukaab — was present at Oxagon from inception. The pivot resolves the dependency by removing the spectacle and keeping the substance.
The Hydrogen Plant
The NEOM Green Hydrogen plant is the jewel in Oxagon’s crown — and the clearest demonstration that the useful components of NEOM were never dependent on the floating city concept.
The $8.4 billion joint venture — financed with $6.1 billion in non-recourse financing from 23 international banks — is on track to generate four gigawatts of solar and wind power and produce up to 600 tonnes of green hydrogen daily. Green ammonia exports are expected to begin in 2027. The project is the world’s largest green hydrogen facility and represents the most technologically and commercially significant component of the entire NEOM programme.
The hydrogen plant does not float. It does not require a floating platform. It does not require an octagonal geometry. It requires sun (abundant in Tabuk), wind (reliable on the Red Sea coast), water (available from the sea for electrolysis), and a port (the one being built at Oxagon). Every input is terrestrial. Every output serves a global market — green ammonia for shipping fuel, industrial feedstock, and energy storage — that operates independently of whether NEOM is a city, a port, or an empty stretch of coastline.
The plant’s success is NEOM’s most important achievement and NEOM’s most devastating critique. The most valuable component of the programme is the component least connected to the programme’s defining vision. The floating city, the linear city, the mountain resort, and the giant cube were supposed to justify NEOM’s $500 billion price tag. The hydrogen plant justifies its own $8.4 billion price tag without reference to any of them. NEOM needed the hydrogen plant. The hydrogen plant did not need NEOM.
What Oxagon Teaches
Oxagon’s trajectory — from floating industrial octagon to terrestrial port-plus-data-centre — is the cleanest illustration of the principle that governs the entire giga-project portfolio: the useful survives; the spectacular does not.
The port is being completed because ports have revenue. The hydrogen plant is being completed because hydrogen has buyers. The data centre is being built because computing has demand. The floating platform was deferred because floating platforms have neither revenue, buyers, nor demand — they have renderings, press coverage, and the admiration of an architecture profession that evaluates proposals on formal qualities rather than engineering feasibility.
The octagonal brand — the name “Oxagon” itself — was derived from the floating concept. With the floating concept deferred indefinitely, the brand describes a shape that has no physical referent at the site. The port is not octagonal. The hydrogen plant is not octagonal. The data centre campus is not octagonal. The site is a stretch of Red Sea coastline with industrial infrastructure under construction. It is useful, functional, and economically rational. It is not floating, not octagonal, and not the future of anything that architectural renderings promised.
The rendering was beautiful. The reality is a port. And the port is more valuable than the rendering ever was.
This analysis draws on NEOM’s Oxagon specifications and promotional materials; BIG’s published design documentation; BESIX’s port construction reports; the NEOM-DataVolt partnership announcement (February 2026); Hexagon’s smart-city contract; ACWA Power and Air Products reporting on the Green Hydrogen plant; reporting by Newsweek, House of Saud, Computer Weekly, and ESG Today; marine engineering literature on floating structure feasibility; and NEOM’s quarterly programme updates. Vision2030.AI is editorially independent and is not affiliated with NEOM, PIF, or any official Vision 2030 entity.