[Edaily Reporter SOYEON KIM ] SamsungElectroMechanics is strengthening its collaboration with NVIDIA, the world’s largest artificial intelligence (AI) semiconductor company, in the field of next-generation semiconductor package substrates. NVIDIA will directly invest in the expansion of SamsungElectroMechanics’ domestic production facilities, and SamsungElectroMechanics has agreed to provide a long-term supply of substrates for AI semiconductors based on this investment. By establishing a cooperative framework that goes beyond their existing supply relationship to encompass investment and production, SamsungElectroMechanics’ high-value-added package substrate business is expected to accelerate its growth.

On the 21st, SamsungElectroMechanics and NVIDIA announced a strategic business partnership aimed at strengthening the global AI semiconductor supply chain and expanding next-generation AI infrastructure. SamsungElectroMechanics has been supplying semiconductor package substrates to NVIDIA, and with this announcement, the two companies have formalized their partnership and agreed to significantly expand the scale of supply.

In particular, NVIDIA will directly invest funds to expand SamsungElectroMechanics’ production capacity for semiconductor packaging substrates. Based on this joint investment, SamsungElectroMechanics and NVIDIA plan to establish dedicated production lines for NVIDIA at their Sejong and Busan facilities. The products to be supplied include substrates for NVIDIA’s AI computing chipsets and packaging substrates for network switches responsible for ultra-high-speed data transmission within data centers. SamsungElectroMechanics will begin a multi-year, long-term supply agreement starting this year.

Through this partnership, NVIDIA has secured a stable supply chain for the advanced package substrates essential to the mass production of AI semiconductors. As securing production capacity in the semiconductor packaging sector becomes increasingly important amid expanding global investment in AI infrastructure, SamsungElectroMechanics is also solidifying its global market dominance by aggressively expanding its high-value-added AI semiconductor package substrate business, which features high technological barriers to entry.

In particular, supply volumes to NVIDIA are expected to grow rapidly. SamsungElectroMechanics forecasts that, starting this year, the volume of package substrates supplied to NVIDIA will increase by an average of more than 10 percent annually. By 2028, supplies to NVIDIA are projected to account for a significant portion of SamsungElectroMechanics’ total package substrate production capacity.

As AI semiconductor performance continues to advance, the importance of package substrates—which connect high-performance chips and reliably transmit power and signals—is also growing. In particular, as the collaboration with NVIDIA expands to encompass not only AI computing chipsets but also substrates for data center networks, SamsungElectroMechanics’ related business is expected to gain momentum.

Semiconductor package substrates are core components that electrically connect semiconductor chips to the motherboard and transmit signals and power. As the computational performance and data processing volume of AI semiconductors increase, the size of the substrates grows and their circuit structures become more complex. Consequently, technologies for implementing ultra-fine circuits and suppressing warpage caused by thermal deformation are considered key competitive advantages that determine product yield and quality.

SamsungElectroMechanics is targeting the market for next-generation AI semiconductor substrates by combining its advanced flip-chip ball grid array (FC-BGA) manufacturing technology with multilayer ceramic capacitor (MLCC) and silicon capacitor technologies. The company has secured technology that embeds passive components—such as MLCCs and silicon capacitors—within the substrate to reduce power noise generated during high-speed computation and improve power efficiency.

Jang Deok-hyun, President of SamsungElectroMechanics, emphasized, “We are honored to join NVIDIA as a key infrastructure partner driving global AI innovation,” adding, “By combining our accumulated advanced FC-BGA manufacturing capabilities with our unrivaled MLCC technology, we will fully fulfill our role as a total solutions provider.”

Debora Shoquist, Vice President of Operations at NVIDIA, said, “To keep pace with the rapid expansion of AI infrastructure, we need to scale up production capacity on a global scale.” She added, “Through our collaboration with SamsungElectroMechanics, we have expanded our network of PCB supply partners and secured additional production capacity. This will enable us to increase production scale and supply the accelerated computing platforms our customers need.”