Key Takeaways
- Samsung’s 2026 investment plan exceeds 110 trillion won ($73.24 billion), marking a 22% increase from 2025
- Capital allocation outpaces TSMC’s annual expenditure, focusing on AI memory, advanced storage, and cutting-edge semiconductor technology
- Company aims to catch up with SK Hynix in high-bandwidth memory (HBM) production for Nvidia’s AI processors
- Transition toward AI-focused chips creates supply constraints for conventional memory chips used in automotive and mobile sectors
- Industry experts predict memory shortages could persist for up to five years
Samsung Electronics (SSNLF) announced plans to deploy more than 110 trillion won — approximately $73.24 billion — throughout 2026 for research initiatives, development projects, and manufacturing infrastructure. This represents a substantial 22% increase compared to the previous year’s allocation of 90.4 trillion won.
This massive outlay positions Samsung’s semiconductor spending beyond that of competitor Taiwan Semiconductor Manufacturing Company (TSM) for the calendar year.
Last year’s budget was distributed between 52.7 trillion won for capital investments and 37.7 trillion won dedicated to research and development. For 2026, Samsung is escalating both categories as it competes for dominance in artificial intelligence semiconductor technology.
Details emerged from a regulatory filing released Thursday. Samsung additionally disclosed its pursuit of strategic acquisitions and mergers across robotics, healthcare technology, automotive electronics, and climate control systems.
The electronics giant confirmed it will distribute 9.8 trillion won in standard shareholder dividends throughout 2026.
A significant portion of this capital targets high-bandwidth memory, commonly known as HBM — the specialized chip architecture that powers Nvidia (NVDA) artificial intelligence accelerators.
SK Hynix currently maintains a commanding position in HBM manufacturing. Samsung’s aggressive investment strategy represents a deliberate effort to narrow this competitive disadvantage.
During Samsung’s shareholder meeting, co-chief executive Jun Young-hyun highlighted unprecedented demand growth. He noted “the emergence of agentic AI is driving an explosive increase in customer orders,” spanning both memory products and enterprise storage solutions.
Micron (MU) has also entered this competitive arena, transforming the market into a three-competitor battle for AI infrastructure supply.
AI Production Shift Creates Memory Bottleneck
The dramatic increase in AI chip manufacturing is producing unintended consequences. As semiconductor producers redirect manufacturing capacity toward premium AI components, output of standard memory modules has declined.
These conventional chips remain essential for automotive systems, mobile devices, and various consumer electronics — yet availability is becoming increasingly constrained.
SK Group chairman Chey Tae-won addressed this challenge in public remarks, cautioning that conventional memory shortages may extend four to five years because of fundamental production capacity limitations.
Samsung indicates its capacity expansion strategy is partially intended to address this supply imbalance by increasing overall manufacturing volume.
Financial Scale Becomes Competitive Barrier
At this magnitude of capital deployment, only select corporations can maintain competitive investment levels. Samsung, TSMC, and SK Hynix constitute the exclusive group with financial resources to commit tens of billions annually.
Samsung’s $73 billion pledge creates simultaneous competitive pressure against TSMC in foundry operations and SK Hynix in memory production.
The company’s US-traded stock under ticker SSNLF has climbed 54.05% during the recent period as investor focus on Samsung’s artificial intelligence strategy has intensified.
Samsung’s Korea-listed equity (005930) serves as the principal trading instrument for institutional investors monitoring the company’s progress in AI semiconductor competition.


