AI infrastructure has its own language.
This glossary explains the terms you will come across in the AI sector.
A
Accelerator
A specialized chip designed to speed up demanding computing workloads. GPUs are the most common accelerators used for AI.
ASIC
Application-Specific Integrated Circuit. A chip designed for one particular type of workload rather than general-purpose computing.
C
Capex
Short for capital expenditure. Money spent on long-term assets such as data centers, servers, power equipment, and chip factories.
Colocation
A data center model where companies rent space, power, and connectivity instead of building and operating their own facility.
D
Data Center
A facility containing servers, networking equipment, cooling systems, and power infrastructure used to run digital services and AI workloads.
G
GPU
Graphics Processing Unit. Originally designed for graphics, GPUs are now the dominant chips used to train and run many AI models.
H
HBM
High Bandwidth Memory. Very fast memory placed close to AI processors so huge amounts of data can move quickly between memory and compute.
Hyperscaler
A very large cloud or technology company operating enormous computing infrastructure, such as Microsoft, Amazon, Google, or Meta.
I
Inference
The process of using a trained AI model to generate an answer, prediction, image, or other output.
Interconnect
The high-speed links that allow chips, servers, and data centers to exchange data.
L
Liquid Cooling
Cooling technology that uses liquid instead of—or alongside—air to remove heat from high-density computing equipment.
P
PPA
Power Purchase Agreement. A long-term contract in which a company agrees to buy electricity from a power producer.
PUE
Power Usage Effectiveness. A measure of data-center energy efficiency. The closer the number is to 1.0, the less energy is being spent on things other than computing.
R
Rack Density
The amount of computing power—and usually electrical power—packed into a single server rack.
T
Training
The computational process used to teach an AI model by exposing it to large amounts of data.
Transformer
In power infrastructure, a device that changes electricity from one voltage level to another. Transformers are critical equipment for connecting large data centers to the grid.