This article assumes you have intermediate or better programming skill with a C-family language (preferably but not necessarily Python) but doesn't assume you know anything about PyTorch tensors. The ...

What you see above is the shipping manifest of a Tensor G5 sample chip that can be found in publicly available databases. The description might seem really complicated, so let’s break it down.

Tensors represent multi-dimensional data points, for example as one-dimensional, two-dimensional, three-dimensional or more. In practical terms, tensors give neural network specialists a way to ...

Since every CPU on the market today can do both of these operations, it means that any desktop, laptop, or tablet can handle basic tensors. However, the above example contains 64 multiplications ...

The chip is code-named “NPT,” which we assume stands for “Newport Beach,” to fit in with the theme of beaches in California (for example Tensor G5 is “LGA” – Laguna Beach).

Google is not giving Tensor a generation signifier at launch, but the company will presumably append a number on the next version. (For example, the Titan M is succeeded by the Titan M2.) ...