Samsung Heavy Industries recruits Greek shipowner and Supermicro to bring 50MW floating AI data centers to market — can be powered by solid oxide fuel cells running on liquefied natural gas
Samsung Heavy Industries has partnered with Greek shipowner Capital Clean Energy Carriers, Lloyd's Register, and Supermicro to commercialize 50MW floating AI data centers powered by LNG fuel cells, sidestepping grid bottlenecks that stall land-based projects.
Samsung Heavy Industries Partners with Greek Shipowner and Supermicro for 50MW Floating AI Data Centers
Samsung Heavy Industries has signed a three-way agreement with Greek shipowner Capital Clean Energy Carriers and classification society Lloyd's Register to bring 50MW floating AI data centers to market, a development first reported by Chosun Biz and confirmed at the Posidonia shipping exhibition in Athens earlier this month.
The deal represents the latest push to solve one of the AI industry's most pressing infrastructure bottlenecks: power. Land-based data centers in the U.S. and Europe face multi-year grid connection queues, while hyperscalers' insatiable demand for compute continues to climb. Floating data centers, moored in ports or coastal waters, can sidestep those queues entirely — and Samsung Heavy believes its shipbuilding expertise gives it a decisive edge.
Under the memorandum of understanding, the three partners each bring distinct capabilities. Samsung Heavy Industries handles the technology design and construction, Capital Clean Energy Carriers leads project sourcing and investment, and Lloyd's Register covers classification, regulation, and certification. Samsung Heavy also signed a second memorandum with Lloyd's Register Advisory covering feasibility studies and a market assessment specifically for North America.
Power Architecture: LNG Fuel Cells and Subsea Cables
The 50 MW floating data center design, which received approval in principle from both the American Bureau of Shipping and Lloyd's Register in April, offers two power modes. When moored in ports or coastal waters, it can draw external power through subsea cables connected to the local grid. Alternatively, it generates its own electricity using solid oxide fuel cells running on liquefied natural gas (LNG), making it fully self-sufficient when grid access isn't available.
"Floating data centers offer a scalable and flexible solution, with the unique advantage of mobility," Jerry Kalogiratos, CEO of Capital Clean Energy Carriers, said in Lloyd's Register's announcement. That mobility means an AI data center could be built in a shipyard in South Korea, sailed to wherever compute demand is highest, and repositioned as market needs shift — something impossible with a land-based facility.
Supermicro Validates AI Servers for Life at Sea
The partnership with Supermicro addresses a critical engineering question: can precision AI hardware survive and perform reliably in a marine environment over a multi-year service life? Vibration from wave action, tilt from vessel motion, salt corrosion, and high humidity all threaten server reliability in ways that don't exist in conventional data centers.
Samsung Heavy Industries will develop offshore positioning control systems and engineering technologies to seal out salinity and moisture from the server compartments. Supermicro's role is to validate server operating conditions in simulated river and marine environments, testing whether its AI-optimized hardware can endure the rigors of an offshore deployment.
The shipbuilder is drawing heavily on its experience constructing floating LNG production facilities — arguably the closest analogous platform — to integrate power generation, seawater cooling, networking, and safety systems into a single hull. Seawater cooling is a particularly attractive feature for AI workloads: water temperatures in coastal and deep-sea environments are significantly colder than ambient air in most data-center-heavy regions, potentially reducing cooling energy costs by a meaningful margin.
Business Model: Shipowners as Data Center Operators
Under the proposed model, shipowners — like Capital Clean Energy Carriers — will purchase the floating data platforms and lease compute capacity to operators under long-term contracts, mirroring the way tankers and container ships are chartered in the maritime industry. This creates an entirely new asset class at the intersection of shipping, energy, and AI infrastructure.
The Competitive Landscape Heats Up
Samsung Heavy is far from alone in pursuing this vision. Japan's Mitsui O.S.K. Lines (MOL) is building a 73 MW floating data center with Karpowership targeting a 2027 deployment — significantly larger than Samsung's initial 50 MW design. China already claims the "world's first" offshore data center: a 24 MW subsea facility off Shanghai that entered full operation last month. Back in the United States, Nautilus Data Technologies operates a much smaller 6.5 MW data center barge at the Port of Stockton, California.
Samsung Heavy also has a potential anchor tenant. The shipbuilder signed a letter of intent with OpenAI in October of last year, covering joint development of floating data centers. This was part of a broader Samsung-OpenAI relationship that also included Samsung's 900,000-wafer-per-month Stargate memory commitment. However, it's worth noting that no signed customer contract for a named deployment location has been announced to date.
The floating data center concept is also attracting venture capital attention beyond traditional shipbuilding. Palantir co-founder Peter Thiel recently backed Panthalassa, a startup developing wave-powered AI data centers, to the tune of $140 million — underscoring the growing conviction that offshore infrastructure will play a meaningful role in meeting AI's insatiable demand for power and compute.


