Google develops mobile processor with tensor cores for machine learning

Google announced the Pixel 6 and Pixel 6 Pro smartphones, powered by its own Tensor SoC, a one-chip system with tensor cores.

So excited to share our new custom Google Tensor chip, which has been 4 yrs in the making (📎 for scale)! Tensor builds off of our 2 decades of computing experience and it’s our biggest innovation in Pixel to date. Will be on Pixel 6 + Pixel 6 Pro in fall. https://t.co/N95X6gFxLf pic.twitter.com/wHiEJRHJwy

— Sundar Pichai (@sundarpichai) August 2, 2021

Representatives said the chipset can handle the most powerful AI and ML models. They noted that since the launch of the first-generation smartphones developers have built AI-powered features such as HDR+ and Night Sight to achieve high-quality photos.

“You will see new camera features, speech recognition, and many other Pixel 6 capabilities,” said Rick Osterloh, Google’s Senior Vice President for Devices and Services.

He added that the new chipset includes the Titan M2 security core, which will make Pixel 6 “the most secure smartphone at the hardware level”.

Both devices feature a metal frame and a glass back panel.

Pixel 6 has a 6.4-inch display with Full HD+ resolution, while the higher-end model features a 6.7-inch display with curved edges and QHD+ resolution.

Further details, release dates and pricing will be announced at a dedicated event in autumn 2021.

In May, at Google’s I/O developer conference, Google unveiled the Tensor Processing Unit TPUv4, designed for cloud computing.

In June, researchers from Google Research said they used reinforcement learning to accelerate chip design from months to six hours.

Subscribe to ForkLog News on Telegram: ForkLog AI — all the news from the world of AI!