Samsung building facility with 50,000 Nvidia GPUs to automate chip manufacturing

In a bold move to enhance its chip manufacturing capabilities, Samsung announced on Thursday its intention to acquire and deploy an impressive array of 50,000 Nvidia graphics processing units (GPUs). This initiative aims to bolster the production of chips for mobile devices and robotics, leading to the establishment of what the company is dubbing an "AI Megafactory." While specific timelines for the facility's construction remain undisclosed, this partnership marks another significant collaboration for Nvidia, which continues to play a pivotal role in the development and implementation of cutting-edge artificial intelligence technology. This announcement follows Nvidia CEO Jensen Huang's recent address in Washington, D.C., where he detailed partnerships with various companies, including Palantir, Eli Lilly, CrowdStrike, and Uber. Huang's visit to South Korea, where he was seen socializing with Samsung Chairman Lee Jae-yong and other prominent business figures, highlights the strengthening ties between Nvidia and Samsung. Furthermore, other notable Korean firms, such as SK Group and Hyundai, are reportedly making similar investments in GPU technology, as stated by Nvidia. Raymond Teh, Nvidia's senior vice president for Asia-Pacific, emphasized the company’s collaboration with the Korean government to advance its ambitious AI leadership objectives. This partnership complements Huang's assertion that Nvidia's current-generation GPU, known as Blackwell, alongside their upcoming Rubin GPU, has created a business portfolio valued at approximately $500 billion. This optimistic forecast has significantly boosted Nvidia's stock, making it the first company to achieve a staggering $5 trillion market capitalization. In light of this collaboration, Nvidia representatives have revealed plans to work closely with Samsung to refine the Korean tech giant's chipmaking lithography platform to seamlessly integrate with Nvidia's GPUs. This adaptation is expected to yield a remarkable 20-fold increase in performance for Samsung. Additionally, Samsung intends to utilize Nvidia's simulation software, Omniverse, to further enhance its operations. Known primarily for its mobile phone production, Samsung has also expressed intentions to leverage Nvidia's GPUs to power its own AI models within its devices. This relationship is mutually beneficial, as Samsung not only partners with Nvidia but is also a crucial supplier of high-performance memory that Nvidia relies on for its AI chips, particularly its high bandwidth memory. The two companies are set to collaborate on optimizing Samsung's fourth-generation HBM memory for AI applications.