AMD signed to use Samsung Foundry 4nm

Over the past few years, Samsung has increasingly lagged behind TSMC in the semiconductor foundry domain. Encountering numerous challenges in chip production, such as stable yields, prominent clients like NVIDIA and Qualcomm have shifted their orders for next-generation high-end chips to competitors. However, Samsung’s recent advancements in 4nm technology have rekindled interest from chip design companies in its cutting-edge processes, enhancing the likelihood of collaboration.

According to Wccftech, AMD has signed a contract with Samsung, transferring a portion of its 4nm chip orders from TSMC. It is rumored that AMD is adjusting the design of its Ryzen 7040 series, codenamed Phoenix, to accommodate Samsung’s 4nm process. As a result of missing the original timetable, the release of these chips has been delayed from April to May, prompting TSMC’s 4nm production line to operate at maximum capacity in order to fulfill unallocated orders as efficiently as possible.

Shifting towards Samsung has granted AMD greater possibilities and scalability. In the past, the choice of TSMC’s 4nm process by major clients such as Apple, Qualcomm, and NVIDIA had limited AMD’s capacity allocation and bargaining power with TSMC. Recent news suggests that Qualcomm’s future Snapdragon 8 platform may adopt a dual-foundry strategy, utilizing both TSMC’s N3E process and Samsung’s 3nm GAA process, a tactic AMD could potentially emulate.

In addition to Qualcomm, Google is commissioning Samsung to manufacture its next-generation Tensor G3, employing the 4LPP process. For the upcoming Exynos 2400, a more advanced and efficient 4LPP+ process is anticipated. Rumors have circulated that AMD may also opt for the 4LPP+ process.