INDEPENDENT INTELLIGENCE / AI INFRASTRUCTUREFROM TOKENS TO POWER
01 / Interface
The economics
of AI.
Layer by layer.
Every interaction creates demand for compute. Start with what people use, then follow the infrastructure underneath.
Explore GPU marketsPrompts / Applications / Revenue
COMPUTE REFERENCEUSD / GPU-hSnapshot · 30 September 2026
NVIDIA H100 SXM$3.85On-demand reference ↗NVIDIA H200$4.68On-demand reference ↗NVIDIA B200$7.48On-demand reference ↗THE MARKET / GPU RENTAL
One GPU.
Many price points.
Compare posted offers, then inspect the conditions behind each price.
Explore the marketON-DEMAND REFERENCEUSD / GPU-HOUR
NVIDIA
$3.85H100 SXM
NVIDIA
$4.68H200
NVIDIA
$7.48B200
Snapshot · 30 September 2026
Posted-price references, not transaction prices or available inventory. See methodology ↗
THE ECONOMICS / INVESTIGATE
From a GPU-hour
to a return on capital.
A quoted price is one part of the equation. Change the assumptions and follow hardware, electricity and idle capacity through to the return.
Open the Giganomica modelHardware+Power+Utilisation→Economics
Who builds the AI economy?
Explore the layers from applications to generation.
↗DOCUMENTED RELATIONSHIPSConnect the pieces.
Trace the suppliers, customers and projects with evidence.
↗