Tensor Processing Unit (TPU)
Definition
TPUs are specialized computing chips from Google for machine learning.
An accelerator developed by Google for tensor-based machine learning computations, especially matrix operations.
Simply explained
A TPU is a specialized computer for certain AI computing tasks. Like a suitable tool, it is tailored to these tasks.
Why it matters in practice
The term contributes to secure, maintainable, and high-performance web applications as well as resilient technical decisions.
Example from business
A company compares TPU and GPU for a large training task based on cost, runtime, and portability.
Limits and pitfalls
- TPU is a product family name; performance depends on generation, data type, compiler, and workload.
Origin & history
Google deployed the first TPU in production starting in 2015 and published architecture and performance results in 2017.
Use in mid-sized companies
- Possible example: A company compares TPU and GPU for a large training task based on cost, runtime, and portability.
Frequently asked questions
What does Tensor Processing Unit (TPU) mean?
An accelerator developed by Google for tensor-based machine learning computations, especially matrix operations. What should one pay attention to? TPU is a product family name; performance depends on generation, data type, compiler, and workload.
Sources
- In-Datacenter Performance Analysis of a TPU · Jouppi et al., In-Datacenter Performance Analysis of a TPU
Source checked on 10.10.2026. Editorial application examples are hypothetical; no evidence of real customer results.
Editorially reviewed on: