Google Pixel 10 Series' Tensor G5 Chip to Arrive With Several Changes, Including Custom Camera ISP: Report

Google's Pixel 10 series of smartphones is expected to arrive latest this year, along with its next-generation Tensor G5 chip. While the company's its previous smartphone processors were produced by Samsung, Google is expected to switch to Taiwan-based TSMC this year. Ahead of the launch of the Google Pixel 10 series, details of several replacements for Samsung-designed components on the Tensor G5 chip have surfaced online, giving us an idea of what to expect from the company's upcoming processor.

Tensor G5 Will Arrive With Google's Audio Processor, DSP, and TPU

Citing a source at Google, Android Authority reports that the company plans to carry over some of its own components from the current generation Tensor G4 to its upcoming chipset. These include the firm's always-on compute (AoC) audio processor, the Google Emerald Hill memory coprocessor, the Google GXP (DSP), and the Google EdgeTPU.

The Tensor G5 will be equipped with Arm Cortex CPU cores, just like its predecessor. Unlike the Tensor G4 and older models, the upcoming chip will feature a GPU from Imagination Technologies DXT, according to the report.

This year, the Tensor G5 chip will reportedly drop support for Google's BigWave (AV1) and Samsung MFC video codecs, in favour of Chips&Media's WAVE677DV video IP. Instead of an Exynos display controller, Google will use a VeriSilicon DC9000 core, while the smartphone will have access to a custom ISP created by Google, instead of a customised version of Samsung's ISP.

Previous Tensor chips used Samsung controllers for interfaces and controllers (UFS, USB3, and PWM), and Google is expected to replace these with alternatives from third party providers such as Synopsys, SmartDV, and Faraday technologies.

If the claims in the report are accurate, Google's upcoming Tensor G5 chip will arrive with several hardware changes later this year. The chip is expected to be produced using TSMC's 3nm process technology, and could deliver performance and efficiency improvements over last year's chipset, the Tensor G4.