AJP Asia Intelligence Agency

Samsung Electro-Mechanics Strengthens AI Infrastructure Alliance with NVIDIA


Samsung Electro-Mechanics and NVIDIA announced on November 18 that they will strengthen their collaboration to support the expansion of NVIDIA's accelerated computing platform production in response to the surging global demand for AI infrastructure.

The two companies had previously announced a strategic partnership for semiconductor package substrate supply in October. Through this enhanced collaboration, Samsung Electro-Mechanics will provide advanced package substrates for NVIDIA's computing and high-speed networking products, thereby supporting the expansion of NVIDIA's accelerated computing platform production.

Samsung Electro-Mechanics will continue to expand its supply collaboration for advanced package substrates used in NVIDIA's next-generation graphics processing units (GPUs). To this end, the company will enhance the operational capabilities of its existing factories in Sejong and Busan, which are key manufacturing hubs in South Korea. This includes improving production line efficiency, maintaining top-level quality control systems, and operating a flexible production system to respond to demand fluctuations.

Based on its advanced packaging process technology capabilities, which meet the ultra-fine circuit implementation, flatness maintenance, and passive component embedding required for next-generation AI semiconductors, Samsung Electro-Mechanics plans to continue supplying package substrates for NVIDIA's high-performance computing (HPC) and ultra-fast data transmission in data centers.

Along with the partnership announcement in October, NVIDIA committed to providing substantial funding for the construction of dedicated factories for AI semiconductor package substrate production at the Sejong and Busan facilities.

To further strengthen this collaboration, NVIDIA is considering additional funding support for expanding production infrastructure at the Sejong and Busan facilities. The additional funding will be utilized for expanding production infrastructure and introducing cutting-edge production equipment, allowing Samsung Electro-Mechanics to establish a full-scale mass production system starting in 2029.

Through this collaboration, NVIDIA will proactively secure a stable supply chain of advanced substrates essential for AI semiconductor mass production, alleviating global supply bottlenecks in AI infrastructure. Meanwhile, Samsung Electro-Mechanics will solidify its position in the high-value AI semiconductor package substrate market, which has the highest technological entry barriers, thereby strengthening its dominance in the global market.

Additionally, the scope of collaboration between the two companies will extend to multilayer ceramic capacitors (MLCCs). A stable power supply is a key requirement for AI accelerated computing platforms operating in high-density and high-output environments, and Samsung Electro-Mechanics will also collaborate on MLCCs needed for NVIDIA's computing platform based on its world-class high-reliability MLCC technology capabilities.

Through this strengthened collaboration, Samsung Electro-Mechanics is expected to further solidify its position as a total solution provider encompassing substrates and MLCCs for AI semiconductors.

Jang Deok-hyun, CEO of Samsung Electro-Mechanics, stated, "By combining world-class substrate micro-processing technology with high-reliability MLCC technology, we will fully fulfill our role as a total solution provider to ensure that NVIDIA's AI platform is supplied to the global market with the best performance and quality in a timely manner."




* This article has been translated by AI.