AI EncyclopediaChips & deep tech
Chips and AI infrastructure
From chips, memory and manufacturing to data centers and electricity. See the technology and facilities that AI runs on, and how they fit together.
New here? Start with "First steps in chips and AI infrastructure"
Data centers & power2/3
The facilities packed with chips that run AI, and the power and money they take.
- AI data centerA large facility that gathers tens of thousands to hundreds of thousands of high-performance chips in one place to train and run AI.
- AI infrastructureA catch-all name for the chips, servers, data centers, electricity and networks needed to train and run AI.
- ServerA computer that takes requests from other computers and does work for them. AI services run on servers too.
- ComputeThe calculating power needed to train and run AI, meaning how many AI chips you can use and for how long.
- Cloud computingA way of renting big computers on the other side of the internet instead of using your own computer.
- Big TechA name for the handful of the world's very largest technology companies.
- Capital expenditure (CAPEX)Money spent to get things that will be used for a long time, such as buildings and equipment. In AI, it mostly means money spent on data centers and chips.
- SupercomputerA very large, very fast computer that can finish calculations that would take an ordinary computer years.
- Power gridThe whole route that carries electricity from power plants to homes, factories and data centers.
- Gigawatt (GW)A large unit for measuring electrical power. One gigawatt is 1 billion watts, about what one large nuclear reactor produces.
- Liquid coolingA way of cooling chips with a liquid such as water instead of air.
Other fields
More on Chips & deep tech
These concept articles were written by AI and checked again for accuracy before publishing. The English edition is translated from the Korean articles. They leave out things that change quickly, such as prices and rankings. Tell us if something is wrong. Report an errorLast updated