AI Glossary
The complete dictionary of Artificial Intelligence
ASIC (Application-Specific Integrated Circuit)
An integrated circuit customized for a particular use, rather than intended for general-purpose use.
TPU Interconnect System
A high-speed network system that connects multiple TPUs to work together as a single unit.
Systolic Architecture
A network of processing units where data flows in a rhythmic pattern, similar to blood flow in the circulatory system.
Bfloat16 (Brain Floating Point)
A 16-bit floating-point format designed for machine learning applications, maintaining the same range as 32-bit float but with reduced precision.
Inference Accelerator
Hardware specifically designed to speed up the inference phase of machine learning models.
FPGA (Field-Programmable Gate Array)
An integrated circuit that can be configured by the customer or designer after manufacturing.
XLA (Accelerated Linear Algebra)
A domain-specific compiler for linear algebra that optimizes TensorFlow computations.
TPU Pod
A large-scale configuration of multiple TPUs connected together to form a supercomputer-scale machine learning accelerator.