Articles topic
AI Inference
Essays on inference latency, persistent regional capacity, and placing AI workloads near demand.
← All articles · Home · Infrastructure
-
Complete Guide
What Is a Subsea Data Center? (Complete Guide)
A complete guide to subsea and underwater data centers: history, benefits, challenges, power, networking, cooling, maintenance, environmental concerns, floating vs subsea, and AI inference.
· 22 min read
-
Industry Statistics
AI Infrastructure Statistics: Power, Cooling, Latency, and Networking
A living reference of AI infrastructure statistics: rack power density, PUE ranges, cooling methods, metro latency bounds, and cable landing context. Updated over time for researchers and operators.
· 12 min read
-
State of the Industry
State of Subsea AI Infrastructure 2026
A living quarterly brief on subsea and underwater data centers: deployments, GPU and rack-power trends, cooling, AI inference demand, announcements, regulation, and cable landing growth.
· 16 min read
-
AI Inference
Why AI Inference Changes Infrastructure Geography
Why AI inference and interactive workloads change where infrastructure must live, how training differs, and why coastal and regional capacity matter more as agents and real-time apps scale.
· 10 min read
-
GEOGRAPHY
The Latency Tax: Why AI Infrastructure Geography Still Matters
Why network geography imposes a measurable cost on AI inference, how LEO satellite networks and coastal compute compare as solutions, and why raw accelerator performance is not the same as delivered throughput.
· 16 min read
-
Coastal Demand
The Durable Geography of AI Demand: Why Coastal Metros Will Keep Mattering
The world’s people, enterprises, trade, and networks remain concentrated around major coastal metros. Seabase is building AI infrastructure for that durable geography.
· 12 min read
-
AI Placement and Capacity
Nori: Connecting AI Workloads to the Right Compute
Nori is Seabase's placement, reservation, monitoring, and infrastructure-routing layer for connecting AI workloads with qualified regional compute capacity.
· 10 min read
-
INFRASTRUCTURE
Beyond Chatbots: The Case for Persistent Regional AI
Why the next generation of AI workloads requires reserved, geographically distributed compute rather than on-demand cloud access: from enterprise agents to real-time perception, robotics, and sovereign systems.
· 12 min read
Other topics: Geography · Infrastructure · Cooling · Energy · Environmental