Samsung and Arm Launch 2nm Edge AI Chip Collaboration in Late August, with OpenAI Seen as Potential Client

According to South Korean media reports, Samsung Electronics and chip architecture designer Arm officially initiated a joint development project for next-generation edge AI system-on-chip (SoC) solutions in late August, with plans to utilize Samsung's advanced 2-nanometer process for mass production. Arm reportedly approved the initial non-recurring engineering expenses for the project by the end of August, marking a substantive step forward in the collaboration. Under the分工 arrangement, Arm is responsible for providing AI accelerator architecture and core design technologies, Samsung's System LSI division handles SoC design, and the foundry division manages production using the 2nm advanced process. Industry sources have indicated that OpenAI is a potential end customer for these custom chips, as the AI company seeks to reduce its reliance on Nvidia's GPUs for inference workloads. The collaboration operates under a limited-use licensing framework that restricts the technology to specific customers and products. For Samsung, this project represents an opportunity to build an integrated "design-plus-manufacturing" AI chip solution; if successful, it could help the company accumulate mass-production experience serving AI chip clients on advanced process nodes, strengthening its foundry competitiveness against TSMC.