Samsung Electronics announced on Thursday its plan to purchase 50,000 graphics processing units (GPUs) from Nvidia to establish a facility aimed at automating its semiconductor manufacturing. The project, which Samsung has termed an “AI Megafactory,” is designed to improve the company’s production of chips for mobile devices and robotics.
According to the announcement, the facility will integrate a large-scale cluster of Nvidia’s GPUs to power artificial intelligence systems. Samsung did not provide a timeline for the construction or completion of the factory.
The collaboration involves multiple facets of both companies’ operations. Nvidia representatives stated on Thursday that they will work with Samsung to adapt the Korean company’s chipmaking lithography platform for use with Nvidia GPUs. They projected this integration would result in a 20-fold performance improvement for Samsung. Additionally, Samsung will utilize Nvidia’s simulation software, Omniverse, as part of the initiative.
Beyond manufacturing, Samsung confirmed it will use the new GPU cluster to run its own proprietary AI models for its consumer devices.
The agreement highlights the dual relationship between the two technology companies, as Samsung is both a customer and a key supplier for Nvidia. Samsung produces high bandwidth memory (HBM), a component used in large quantities alongside Nvidia’s AI chips. As part of the partnership, Samsung stated it will collaborate with Nvidia to optimize its fourth-generation HBM for AI-specific applications.
The announcement followed a series of related events involving Nvidia’s leadership. On Tuesday in Washington, D.C., Nvidia CEO Jensen Huang announced collaborations with several other companies, including Palantir, Eli Lilly, CrowdStrike, and Uber. During that event, Huang also forecasted a $500 billion book of business from its current and next-generation GPUs, named Blackwell and Rubin. Following the forecast, Nvidia’s market capitalization reached $5 trillion.
Shortly after his U.S. appearance, Huang was in South Korea, where local media reported he met with Samsung Chairman Lee Jae-yong and other business leaders.
In a call with reporters on Wednesday, Raymond Teh, Nvidia’s senior vice president of Asia-Pacific, commented on the company’s regional strategy. “We’re working closely with the Korean government to support its ambitious leadership plans in AI,” Teh said. Nvidia also noted that other major South Korean corporations, such as SK Group and Hyundai, are deploying similar quantities of its GPUs.
