tensor

American  

[ten-ser, -sawr] / ˈtɛn sər, -sɔr /

noun

1. Anatomy.  a muscle that stretches or tightens some part of the body.

2. Mathematics.  a mathematical entity with components that change in a particular way in a transformation from one coordinate system to another.

---

tensor British  

/ tɛnˈsɔːrɪəl, -sɔː, ˈtɛnsə /

noun

1. anatomy any muscle that can cause a part to become firm or tense

2. maths a set of components, functions of the coordinates of any point in space, that transform linearly between coordinate systems. For three-dimensional space there are 3 r components, where r is the rank. A tensor of zero rank is a scalar, of rank one, a vector

---

tensor Scientific  

/ tĕn′sər,-sôr′ /

1. A structure of quantities arranged by zero or more indices, such as a scalar (zero indices), a vector (one index), or a matrix (two indices), which is invariant under transformations of coordinates.

2. Any of various muscles that stretch or tighten a body part, as the muscle that acts to tense the soft palate, called the tensor palati.

---

Other Word Forms

• tensorial adjective

Etymology

Origin of tensor

1695–1705; < New Latin: stretcher, equivalent to Latin tend ( ere ) to stretch ( tend ¹ ) + -tor -tor, with dt > s

Example Sentences

For one, Alphabet’s tensor processing units, designed alongside Broadcom, were initially designed for internal use.

From MarketWatch

Then, as enthusiasm built further for Alphabet’s tensor processing units, which are an alternative to the graphics processing units sold by Nvidia, the company weighed in on that conversation as well.

From MarketWatch

Citing a person familiar with the matter, The Information External link reported Monday that Meta is in discussions to use Google’s tensor processing units, or TPUs, in its data centers starting in 2027.

From Barron's

It is still up in the air whether Meta would use the chips, known as tensor processing units or TPUs, to train its AI models or to do inference, one of the people said.

From The Wall Street Journal

The company said it trained its new Gemini 3 model on tensor processing units, its custom chips, and there seems to be growing outside interest in Google’s semiconductor work.

From MarketWatch