Why the AI boom could keep memory prices high until 2030 (www.techradar.com)

In a provocative statement, Sam Altman, CEO of OpenAI, suggested that DRAM memory should be allocated mainly to data centers to optimize efficiency for AI applications, as he believes it is underutilized by consumers. His comments come amid significant shortages in memory supply, primarily driven by the rising demand for AI technologies. As OpenAI's upcoming Stargate datacenter project is projected to consume 40% of the world’s DRAM supply, Altman's view highlights a broader concern: the continued prioritization of AI over consumer memory needs could lead to sustained high prices through 2030. The implications of this shortage are considerable for the AI/ML community and beyond. With major memory manufacturers like SK Hynix predicting that the shortage could last for years, there is a potential shift in production toward high-bandwidth memory tailored for AI applications rather than consumer markets. Additionally, emerging technologies in robotics and autonomous systems are set to further strain memory resources as they increasingly rely on advanced AI. This could exacerbate affordability issues for consumers while favoring corporate reallocations of memory supplies, fundamentally reshaping the landscape of memory production and pricing in the near future.