OpenAI CEO warns that consumer RAM is inefficient amid rising AI-driven hardware costs
Sam Altman argued that memory in personal computers is a poor use of scarce DRAM, while market analysts note that AI demand and supply constraints have driven up RAM prices and could reshape the PC hardware market.
OpenAI chief executive Sam Altman told the Cheeky Pint podcast that memory residing in personal computers represents an inefficient allocation of scarce DRAM, which he believes should be dedicated to data-center workloads. He said local AI inference utilizes that memory only about 5% of the time, making it a "terrible" use of resources.
The comment comes as the PC hardware market faces elevated component costs. Prices for a 32 GB DDR5 kit have risen to roughly $500, and high-end laptops and graphics cards are commanding premium prices. Industry observers attribute the pressure to a combination of strong AI-driven demand, limited inventory, geopolitical tariffs and semiconductor foundries shifting capacity toward enterprise servers.
Analysts note that the surge in AI spending has strained memory supplies, with companies such as Anthropic reporting billions in infrastructure expenses and large projected compute commitments. At the same time, some corporate customers are scaling back AI usage, citing cost and privacy concerns. This pullback, along with broader market corrections, could prompt foundries to reallocate production back to consumer-grade memory, potentially easing prices.
However, if enterprise orders continue to decline, manufacturers might reduce overall wafer output, keeping RAM scarce and maintaining high prices. The outcome will influence whether the current hardware cost surge eases for consumers building new PCs.