Samsung Electronics plans to integrate its self-developed GPU into application processors for the first time in 2027, accelerating its edge-side AI layout

Samsung Electronics is promoting the complete autonomy of graphics processors, planning to equip its self-developed GPU in the Exynos chip set to launch in 2027, marking a key step for the world's largest memory chip manufacturer in building an end-to-end AI ecosystem.

According to a report by the Korea Economic Daily on the 25th, Samsung aims to use self-developed graphics IP in the application processor Exynos 2800, which is set to launch in 2027, replacing the existing collaborative solution.

This move will reduce Samsung's reliance on external suppliers and gain greater autonomy in the iteration of functions and features. Samsung's current mid-to-high-end mobile chips use the Xclipse GPU based on AMD's RDNA architecture.

Samsung plans to expand the application range of its self-developed GPU from smartphones to smart glasses, robots, automotive SoCs, and the AI-specific chip market.

Breaking Free from AMD Dependency and Diversifying Market Layout

Samsung Electronics currently uses the Xclipse GPU derived from AMD's RDNA series architecture in its mid-to-high-end mobile application processors. By shifting to a fully self-developed graphics architecture, the company will have greater control over product planning and technology roadmaps.

Reports indicate that Samsung plans to start with smartphones and gradually promote its self-developed GPU to a wider range of devices, including smart glasses, robots, and automotive chips, while also entering the AI-specific chip market. This strategy aims to establish a complete end-side AI product ecosystem