Samsung Electronics Co. Ltd. has revealed that the next iteration of its flagship mobile chip will use Advanced Micro Devices Inc.’s RDNA 2 graphics processing unit architecture.

The South Korean electronics giant disclosed its collaboration with AMD in a Thursday tweet. Samsung also detailed that the new AMD-powered mobile processor is set to debut on Jan. 11.

Samsung manufactures the popular Galaxy S series of Android smartphones, as well as other mobile devices including a growing array of foldable handsets. The company’s device portfolio uses two separate lines of processors. Samsung ships smartphones that it sells in the U.S. with chips from Qualcomm Inc., while handsets offered by the company internationally use its own Exynos system-on-on-chip series.

The newest Exynos chip, the Exynos 2100, made its debut in January this year. It’s based on Samsung’s five-nanometer manufacturing process, which promises to provide 10% higher performance or up to 20% lower power consumption than earlier seven-nanometer silicon. Samsung uses extreme ultraviolet lithography technology to make its five-nanometer chips.

The Exynos 2100 combines a central processing unit with an onboard 5G modem and an artificial intelligence accelerator. Also included: a GPU based on Arm Ltd.’s Mali-G78 graphics card design. Samsung’s decision to equip its next-generation Exynos chip with AMD’s RDNA 2 architecture potentially suggests that the company has decided to reduce its use of Arm’s GPU designs in the mobile market. 

AMD’s RDNA 2 architecture powers its high-end GPUs for personal computers. One of the GPUs that uses RDNA 2, the Radeon RX 6700 XT, promises to deliver up to twice the performance of previous-generation chips in some situations. 

According to AMD, RDNA 2 provides 54% higher performance per watt than the original iteration of its RDNA architecture. One of the main contributors to the architecture’s increased efficiency is that AMD has reduced by 30% the power consumption of its compute units, the circuits that carry out calculations in a GPU. 

The compute units in RDNA 2 switch off some of their components when they’re not in use to reduce unnecessary electricity use. They can also more efficiently move data between different parts of a GPU. Data movement is one of the main factors that influence the efficiency of a chip, which means that optimizing the process reduces power consumption.

Samsung’s decision to use RDNA 2 in its next-generation chips could boost the performance of its 2022 mobile devices. International editions of the company’s flagship Galaxy S21 smartphone use the current-generation Exynos 2100. It’s possible that the 2022 successor to the Galaxy S21 will feature the new AMD-powered Exynos chip set to be announced by Samsung on Jan. 11.

In the U.S., Samsung ships its smartphones with Snapdragon chips from Qualcomm. If Samsung continues to use Qualcomm silicon this year, its flagship 2022 smartphones can likely be expected to feature the chipmaker’s newest Snapdragon 8 Gen 1 processor.  

Debuted earlier this month, the Snapdragon 8 Gen 1 features an array of improvements over its predecessor. One of the main highlights is an upgraded GPU that promises to provide 30% higher performance and a 25% improvement in power efficiency. The GPU is joined by a new onboard 5G modem chip that Qualcomm said supports download speeds of up to 10 gigabits per second. 

Image: AMD