Developing and applying large language models for generative artificial intelligence requires massive computing power. Nvidia has seen a 175% rally in stock prices due to the computing power required for AI-related projects.
Citi Analyst, Peter Lee, believes that AI-related computing will drive a surge in demand for DRAM memory chips, estimating that memory needs will increase by six times for Nvidia’s DGX100 graphics processor-based servers. As such, Lee predicts that AI-related memory accounts for 16% of global DRAM revenue this year which is set to surge by 31% next year, 41% in 2025, and 56% by 2027.
This trend is anticipated to benefit Samsung Electronics and SK Hynix significantly. Micron Technology is the only major U.S.-based memory-chip manufacturer set to benefit as well. Despite recent softness and a drop in PC sales as a result of the pandemic, any upcoming commentary by Micron on the potential benefits to demand from AI should help boost the stock.
Micron’s stock is currently up by 0.7% at $65.75, with a gain of approximately 32% year-to-date.
The rise in artificial intelligence projects means there will be an increasing demand for DRAM memory chips. Companies who can supply them, such as Samsung Electronics, SK Hynix, and Micron Technology, are set to benefit greatly going forward.
