Samsung commits $73bn to AI chip expansion in 2026

Samsung will spend $73 billion on chip manufacturing in 2026, a 22% jump that tops TSMC's planned $50 billion. The push targets AI memory chips, where Samsung is fighting to reclaim ground from SK Hynix.

Categorized in: AI News IT and Development
Published on: Mar 20, 2026
Samsung commits $73bn to AI chip expansion in 2026

Samsung Commits $73 Billion to AI Chip Manufacturing in 2026

Samsung Electronics plans to spend more than ₩110 trillion (US$73 billion) on chip capacity expansion and research this year, a 22% increase from 2025 that exceeds Taiwan Semiconductor Manufacturing Company's planned US$50 billion capital expenditure.

The investment reflects a strategic pivot toward AI-driven demand, with Samsung focusing heavily on next-generation AI chips and advanced foundry processes. The capital commitment represents more than half of the company's projected operating profit for the year.

The HBM Competition

Samsung is targeting a comeback in high-bandwidth memory (HBM) chips, where SK Hynix currently dominates as the primary supplier to NVIDIA. Earlier this year, Samsung became the first company to commercially ship HBM4 chips to customers, breaking through qualification delays that had hindered its position.

The company unveiled its HBM4E chip at NVIDIA's GTC event in March, earning endorsement from NVIDIA President and CEO Jensen Huang. Samsung's 4-nanometer technology will manufacture Groq 3 processors, and the company recently signed a deal to supply HBM4 chips to Advanced Micro Devices (AMD).

Memory Supply Constraints Drive Pricing Power

Memory has become a bottleneck in AI inference. Demand for high-end memory essential to NVIDIA accelerators has created supply constraints where long lead times for new fabrication facilities reinforce suppliers' pricing power.

Macquarie analysts forecast that DRAM and NAND prices will remain strong for at least two more years, with net profit potentially rising tenfold between 2025 and 2028. The explosive surge in orders for server-grade storage, driven by agentic AI, underscores the immediate need for capacity expansion.

The Broader Memory Shortage

As Samsung shifts production toward high-end AI memory, conventional memory chips have become scarce. These chips supply most modern devices-cars, smartphones, laptops, and data centers-creating a structural deficit that is inflating prices and disrupting production across industries.

Samsung's capacity expansion may help address this global shortage, though analysts view the issue as long-term and structural. Jun Young-hyun, Vice Chairman and CEO of Samsung, said the company will "strengthen technology competitiveness and manage risks" despite headwinds from US tariff concerns and manufacturing costs.


Get Daily AI News

Your membership also unlocks:

700+ AI Courses
700+ Certifications
Personalized AI Learning Plan
6500+ AI Tools (no Ads)
Daily AI News by job industry (no Ads)