Concept
What is AI compute?
The basic resource behind training and running AI models.
Learn
Learn the market behind AI
AI runs on a physical stack: chips, power, data centers, cloud capacity, and the prices buyers pay to access them. Start here to learn what AI compute is, why it matters, and how it is becoming a market of its own.
What is AI compute?Basics
The capacity behind training and inference — chips, time, and access.
Why it mattersEconomics
Compute shapes how fast AI scales, what it costs, and who can compete.
How the market worksMarket
Chips, cloud, power, rentals, spot, and forward pricing all move supply and price.
Start here
New to the market? Begin with these.
Four lessons that set up everything else on ComputeTape.
Concept
Why compute matters
Why chips, power, and capacity are becoming economic constraints.
Compare
H100 vs H200 vs B200
How accelerator generations affect performance, supply, and cost.
Unit
What is a GPU-hour?
The basic unit behind compute pricing.
Learning paths
Explore by topic
Once you know the basics, explore the infrastructure, market structure, and emerging topics that shape compute supply and price.
Compute basics
Concept
What is AI compute?
The basic resource behind training and running AI models.
Unit
What is a GPU-hour?
The basic unit behind compute pricing.
Compare
H100 vs H200 vs B200
How accelerator generations affect performance, supply, and cost.
Infrastructure
Power
Why power matters
Why electricity and site capacity shape AI compute markets.
Data centers
What is a data center?
The physical site where chips, power, cooling, networking, and operations come together.
Cooling
Why cooling matters
Why heat limits how densely AI chips can be deployed and operated.
Networking
Why networking matters
Why fast interconnects turn individual chips into useful AI clusters.
Memory
Why memory matters
Why high-bandwidth memory can constrain accelerator supply and model performance.
Market structure
Operator
What is a neocloud?
Compute-first cloud operators and why they matter.
Rentals
What are GPU rentals?
How buyers rent accelerator capacity and what rental signals can reveal.
Spot
What are spot prices?
How short-term, interruptible capacity can become a pricing signal.
Futures
What are compute futures?
Forward-looking pricing for compute capacity.
Curve
How to read a forward curve
How curve shape becomes a market signal.
Emerging topics
Project
What is Terafab?
A proposed chip and compute-capacity project.
Project
What is Stargate?
A mega-scale AI infrastructure buildout and what it says about future compute supply.
Project
What is Colossus?
xAI’s large-scale compute buildout and why power can become the bottleneck after GPUs arrive.
Project
What is Project Rainier?
AWS’s custom-silicon AI cluster and why proprietary chips matter to compute supply.
Project
What are Prometheus and Hyperion?
Meta’s multi-gigawatt AI campuses and what industrial-scale compute really means.
Why now
Why compute matters now
- AI demand is turning chips, power, and data-center capacity into scarce economic resources.
- The cost of compute affects model training, inference, cloud margins, and who can scale.
- As compute becomes more measurable, tradable, and constrained, it starts behaving more like a market.