N環球 
Samsung Electro-Mechanics Partners with NVIDIA to Expand ABF Substrate Capacity for Next-Gen AI Chips
Samsung Electro-Mechanics announced on Oct 21 a strategic partnership with NVIDIA to jointly invest in dedicated ABF package substrate lines in Sejong and Busan, South Korea, supplying NVIDIA's next-generation AI processors and high-speed networking switches under a multi-year agreement starting in 2026.
Why it matters
AI data-center buildouts are outpacing the supply chain, and package substrates have become a hidden bottleneck for AI chip production. NVIDIA locking in multi-year capacity with a key supplier signals the AI chip leader expects demand to keep growing for years; for Samsung Electro-Mechanics, it is a ticket into the high-value AI substrate market, competing head-on with Taiwanese substrate makers.
Full report
Samsung Electro-Mechanics announced on October 21 that it has entered into a strategic business partnership with NVIDIA, with the two companies jointly investing to expand dedicated manufacturing facilities for ABF package substrates used in AI semiconductors.
Under the agreement, dedicated production lines will be established at key sites in Sejong and Busan, South Korea. Deliveries to NVIDIA begin in 2026 under a multi-year supply agreement covering advanced package substrates for NVIDIA's compute chipsets, as well as substrates for high-speed networking switches that enable ultra-fast chip-to-chip connectivity inside data centers. Samsung Electro-Mechanics expects supply volumes to grow at a double-digit annual rate, and by 2028 the capacity allocated to NVIDIA is projected to account for a significant portion of its total annual package substrate production.
Duck-Hyun Chang, President and CEO of Samsung Electro-Mechanics, said it is "a great honor" to join NVIDIA as a key infrastructure partner, combining advanced FC-BGA manufacturing with high-reliability MLCC expertise to deliver NVIDIA's AI platforms on time. Debora Shoquist, NVIDIA's executive vice president of operations, said the partnership broadens NVIDIA's substrate partner network and adds manufacturing capacity to scale production.
Next-generation AI chips demand ever more from substrates: larger sizes, embedded passive components such as multilayer ceramic capacitors (MLCCs) and silicon capacitors, ultra-fine circuit patterning and warpage control — all critical to yields and long-term reliability. That high barrier to entry is precisely why NVIDIA is locking in capacity early.
💬 Comments
No comments yet. Be the first!