On Monday, Google premiered its new Pixel phones, which feature an AI-optimized Tensor chip. The search giant says that the internally-developed chip has an embedded machine learning model, adding that the Pixel 6 and Pixel 6 Pro will be available in the fall.
At present, the Alphabet unit is only willing to share the device’s full specifications when it launches but has released a few details early as part of the preview announcement.
The Pixel 6 will have a 6.4-inch FHD+ display while the Pixel 6 Pro will have a 6.7-inch QHD+ display.
Apart from having different resolutions (FHD+ equals a resolution of 2,200 pixels by 1,080 pixels and QHD+ describes a resolution of 3,200 by 1,800 pixels), the Pixel 6 Pro will have three rear cameras, while the Pixel 6 will ship with just two.
Google is introducing the Tensor chip, which was developed internally and references the search giant’s Tensor Processing Units it deploys in its data centres to run machine learning algorithms.
The Tensor chip inside the Pixel 6 and Pixel 6 Pro is fitted with a dedicated module, specifically to run machine learning code.
As part of the announcement, Google shared an overview of the chip. It is described as a system-on-chip with several processors, each with a dedicated computing task. Most smartphones use SOCs, which have a central processing unit, a graphics processing unit, a machine learning module, and other pertinent components.
The Pixel 6 and Pixel 6 Pro can be expected to feature a processing components layout similar to that. The devices will also have a new version of Titan M, a custom-design chip Google developed and first used in the Pixel smartphones released in 2018.