AlphaWire

ima_daily5min

Korean Technology Summer Seminar: Focusing on AI, memory and MLCC, Agentic AI will create huge demand for CPU and DRAM, and MLCC content is expected to surge (Morgan Stanley)

2026-07-15·ima-daily5min-0715-68-a961b42bb1
Street Signal | Korean Technology Summer Seminar: Focusing on AI, memory and MLCC, Agentic AI will create huge demand for CPU and DRAM, and MLCC content is expected to surge (Morgan Stanley)

Morgan Stanley's Korean technology team released an in-depth report on AI, memory and MLCC. There are three core points to watch: First, Agentic AI (agent artificial intelligence) will completely change the computing architecture.

Under the BS scenario, it may create up to US$238 billion in additional CPU (central processing unit) opportunities and 221EB of excess DRAM demand by 2030. Second, the supply and demand analysis of HBM and NAND shows that HBM will face a 17% supply shortage in 2026, and NAND will also have a 15% supply gap in 2026.

Third, as the motherboard power architecture (VR) evolves to a new solution, the MLCC (multilayer ceramic capacitor) content will increase by more than 180% compared to the previous generation solution.

These three major changes jointly point to the fact that AI is detonating demand from core computing to the entire industry chain such as storage and passive components. One sentence conclusion: AI demand is penetrating into deeper and wider fields.

AI Agent, HBM/DDR5 shortages and the surge in MLCC content are the three most certain investment main lines, forming the core narrative of Korean and global technology stocks in 2026-2027.

Positive/negative: Positive for Samsung Electronics (005930.KS), SK Hynix (000660.KS) (HBM and NAND), Samsung Electro-Mechanics, Taiyo Yuden and other MLCC manufacturers, as well as the AI CPU-related server industry chain.

The market has already recognized the driving force of AI and storage, but the revolutionary increase in MLCC content and the huge new demand for AI Agent (CPU) may not be adequately priced. Catalysts:

1) The finalization of HBM4 technical specifications and mass production schedule;

2) Deployment cases of AI Agent applications in major cloud vendors and the stimulation of server demand;

3) New product releases and production expansion plans of leading MLCC manufacturers (Murata, Samsung Electro-Mechanics).

Full text

Korean Technology Summer Seminar: Focusing on AI, memory and MLCC, Agentic AI will create huge demand for CPU and DRAM, and MLCC content is expected to surge (Morgan Stanley)

Morgan Stanley's Korean technology team released an in-depth report on AI, memory and MLCC.

Morgan Stanley's Korean technology team released an in-depth report on AI, memory and MLCC. There are three core points to watch: First, Agentic AI (agent artificial intelligence) will completely change the computing architecture. Under the BS scenario, it may create up to US$238 billion in additional CPU (central processing unit) opportunities and 221EB of excess DRAM demand by 2030. Second, the supply and demand analysis of HBM and NAND shows that HBM will face a 17% supply shortage in 2026, and NAND will also have a 15% supply gap in 2026. Third, as the motherboard power architecture (VR) evolves to a new solution, the MLCC (multilayer ceramic capacitor) content will increase by more than 180% compared to the previous generation solution. These three major changes jointly point to the fact that AI is detonating demand from core computing to the entire industry chain such as storage and passive components. One sentence conclusion: AI demand is penetrating into deeper and wider fields. AI Agent, HBM/DDR5 shortages and the surge in MLCC content are the three most certain investment main lines, forming the core narrative of Korean and global technology stocks in 2026-2027. Positive/negative: Positive for Samsung Electronics (005930.KS), SK Hynix (000660.KS) (HBM and NAND), Samsung Electro-Mechanics, Taiyo Yuden and other MLCC manufacturers, as well as the AI CPU-related server industry chain. The market has already recognized the driving force of AI and storage, but the revolutionary increase in MLCC content and the huge new demand for AI Agent (CPU) may not be adequately priced. Catalysts: 1) The finalization of HBM4 technical specifications and mass production schedule; 2) Deployment cases of AI Agent applications in major cloud vendors and the stimulation of server demand; 3) New product releases and production expansion plans of leading MLCC manufacturers (Murata, Samsung Electro-Mechanics).

← Back to archive