AI-Driven Next-Generation Substrates… Competition to Secure a Lead in Glass Substrates and FO-PLP Heats Up
Competition in Advanced Packaging Technologies Accelerates as AI Semiconductors Grow Larger
FO-PLP and Glass Substrate Market Projected to Reach 12.6 Trillion Won by 2030
Following TSMC and Intel, SamsungElectroMechanics and LG Innotek Expand Investments
[Edaily Reporter JAEMIN SONG ] The importance of advanced packaging is also growing in the competition for artificial intelligence (AI) semiconductors. This is because, as ultra-large chips combining AI accelerators and high-bandwidth memory (HBM) are being introduced one after another, limitations are becoming apparent in boosting performance and power efficiency using existing packaging technologies alone. Consequently, fan-out panel-level packaging (FO-PLP) and glass substrates are emerging as key next-generation technologies, and the competition for investment among global semiconductor companies is intensifying.
A panoramic view of SamsungElectroMechanics’ Suwon campus. (Photo: SamsungElectroMechanics) According to market research firm Counterpoint Research on the 3rd, the global FO-PLP and glass substrate market is projected to grow more than 12-fold over six years, from $650 million (approximately 1 trillion won) in 2024 to $8.1 billion (approximately 12.6 trillion won) in 2030. Products for AI and high-performance computing (HPC) are expected to account for 45.6% of total FO-PLP market revenue and drive market expansion.
This trend is driven by the increasing size of AI semiconductors. As chip sizes grow and HBM stacking increases to boost AI computing performance, traditional round wafer-based packaging faces limitations in terms of production efficiency and thermal management. FO-PLP utilizes panels instead of round wafers, allowing for higher area utilization and making it well-suited for implementing ultra-large packages. Glass substrates are also considered a suitable technology for next-generation AI semiconductors, as they exhibit lower thermal expansion and higher flatness compared to conventional organic substrates.
Global companies are also accelerating their efforts to secure a leading position in the market. TSMC is developing “CoPoS (Chip on Panel on Substrate),” a next-generation panel-based packaging platform, and is pushing forward with the adoption of glass substrates. Intel is also developing technology to introduce glass substrates across its entire range of advanced chips by 2030. As global back-end (OSAT) companies such as ASE and PTI move to expand their production capacity, competition in advanced packaging appears to be spreading throughout the entire supply chain.
Domestic companies are also accelerating their proactive investments. SamsungElectroMechanics recently signed an agreement with Dongwoo HANCHEM CO.,LTD to establish a joint venture worth approximately 4800억 won, aiming to secure a production base for glass core—a key material for glass substrates. The strategy is to accelerate commercialization by internalizing the material supply chain, following the pilot line at its Sejong plant. LG Innotek is fostering FC-BGA for AI as a future growth engine while simultaneously pursuing the development of next-generation glass substrate technology, and Absolix, a subsidiary of SKC, is focusing on securing global customers based on its U.S. production hub.
The long-term growth outlook for the AI memory market is also driving these investments. Recently, Micron projected that memory supply shortages driven by expanding AI demand will persist beyond 2027. The industry expects demand to continue growing not only for memory but also for components and packaging that support AI semiconductors, such as MLCCs, flip-chip ball grid arrays (FC-BGA), and glass substrates.
An industry official stated, “As competition over AI semiconductor performance intensifies, the importance of packaging technology will inevitably grow,” adding, “Although the markets for glass substrates and FO-PLP are still in their early stages, the companies that secure technological capabilities first and secure big tech clients early on will ultimately take the lead in the market.”
The biggest bottleneck for AI servers is memory. As the KV cache—where large language models (LLMs) store past computations—accumulates, the required memory capacity increases exponentially. This prob…
Concerns about “tax risks” have been raised in some quarters of the financial investment, pharmaceutical, and biotech markets regarding Genosco, a subsidiary of OSCOTEC Inc.(039200)specializing in new…
As July began (June 29–July 3), the pharmaceutical and biotech industries turned their attention to Celltrion Pharm Inc.’s large-scale investment in production facilities and AriBio’s successful fundr…