AI-related semiconductors — including processors, high-bandwidth memory (HBM), and networking components — accounted for nearly one-third of total sales in 2025. Gartner projects that AI infrastructure spending will exceed USD 1.3 trillion in 2026. 

The buildout of AI infrastructure drove strong demand for AI processors, high-bandwidth memory (HBM), and networking chips. In 2025, HBM represented 23% of the DRAM market, surpassing USD 30 billion in sales, while AI processors exceeded USD 200 billion. Gartner projects that AI semiconductors will account for more than 50% of total semiconductor sales by 2029. 

Nvidia strengthened its market lead, surpassing USD 100 billion in semiconductor revenue for the first time and extending its lead over Samsung Electronics by USD 53 billion. According to Gartner, the company contributed to more than 35% of the industry’s growth last year. 

Samsung retained the second position with USD 73 billion in revenue, driven by a 13% increase in memory sales, while non-memory revenue declined 8% year-over-year. SK Hynix rose to number three with USD 61 billion in revenue, up 37% on strong demand for HBM in AI servers.