Hardware that supports software and code running across the Internet and devices remains just as important to search as Google Ads and the Marketing and Ad Manager suite of products, formerly known as DoubleClick.

It’s important to marketers because on Thursday Google revealed its work on an application-specific integrated circuit — a semiconductor commonly referred to as ASIC, that will carry out machine learning tasks in IoT devices.  

The company calls the ASIC Edge TPU, a tiny accelerator focused on artificial intelligence.

Edge TPU is designed to carry out the task that machine learning teaches it to do.

And while these chips are designed for enterprise jobs in the cloud, not consumer smartphones, it’s not out of the realm of possibilities to use them in handsets and smart speakers or hubs.

Edge TPU will automate any type of task that requires cloud computing such as manufacturing or the application used by Smart Parking, a sensor-based system for shopping centers, airports, and universities.

Smart Parking alerts the system when parking spaces are available or when they are empty. It also lets operators know when vehicles have overstayed their time.

Controlling the hardware that software and code run on gives the inventor of the device an added benefit, such as the ability to view data it might collect, so it can do a better job of matching advertising and services with consumers.

Google also on Thursday announced with Lenovo that the manufacturer will begin shipping its first smart display, which will give users a way to visually interact with Google Assistant. The display also will allow users to stream videos on YouTube, such as how-to or cooking videos.

It’s really just a matter of time until we see ASICs in Google smartphones.