The Tensor AP for Pixel 8 Phones
Google has reportedly decided to produce the mobile application processors (APs) for its next-generation smartphone Pixel 8 at Samsung Electronics' 3-nm foundry lines. The U.S. tech giant is developing the chip in cooperation with the Korean smartphone giant.
Google is developing the third-generation Tensor with Samsung Electronics' System LSI Division. The mobile AP will be used for the Pixel 8 smartphone, scheduled to come out in the second half of next year.
The Tensor is a mobile chip developed by Google last year and was first applied to the Google smartphone Pixel 6 in 2021. Google had used Qualcomm’s mobile chips until 2020 but succeeded in developing the Tensor with support from Samsung Electronics.
The Tensor is expected to be produced only for smartphones for the time being. However, Google has a plan to expand the application of its own chip to tablets, laptops and servers through its own chip diffusion strategy. This is because an improvement in semiconductor performance contributes to boosting core services and product performances.
Google’s smartphone market share stood at a mere 3 percent as of the end of last year. In the saturated smartphone market, Google smartphones have a meager presence. On the other hand, Google's Android smartphone operating system (OS) accounted for 72 percent of the global mobile OS market as of the end of last year. Google believes that if it makes its own chip and combines it with an optimized operating system, it can enhance the quality of its products and services in an integrated ecosystem.
Google chose Samsung Electronics as its partner for this strategy. Cooperation with Google also helps Samsung sharpen its competitiveness in the overall semiconductor business.
Some experts say that Google is unlikely to use Samsung Electronics' 3-nm lines to produce its mobile chips because Google's Pixel phone is not an expensive premium product that requires cutting-edge technology. They say that Google is expected to continue to use a 4-nm process instead of adopting a costly 3-nm process.