Panthalassa's stacked technical risks
Panthalassa
The core issue is sequencing, because Panthalassa is trying to commercialize four failure sensitive systems at once instead of plugging compute into someone else’s proven offshore platform. It needs a wave power system that works for long stretches at sea, autonomous marine operations that keep the node positioned and serviceable, data center hardware that stays online in a harsh saltwater environment, and a manufacturing process that can repeat this economically at fleet scale.
-
Aikido strips out one of the hardest layers by using floating offshore wind and existing steel fabrication supply chains. Its pitch is not better ocean physics, it is faster deployment using distressed wind sites, modular steel structures, and standard offshore suppliers.
-
HiCloud and Samsung Heavy show how rivals can enter the market by borrowing mature pieces. HiCloud put a 24 MW offshore wind linked underwater data center into commercial operation near Shanghai in May 2026, while Samsung Heavy won approval in principle for a 50 MW floating data center concept built on shipbuilding know how.
-
That matters because every extra technical dependency multiplies go to market risk. A customer can underwrite a floating hull, a cable connection, or a cooling design separately. It is much harder to underwrite a single vendor that must prove energy generation, marine autonomy, uptime, and factory scale all together.
The likely direction is a split market. Integrated ocean nodes can win where fuel logistics, remote siting, and sovereign placement matter most. Larger mainstream AI infrastructure demand is more likely to flow first to offshore designs that keep one novel element and import the rest from wind, shipbuilding, and conventional data center supply chains.