1 min read
Add as a preferred source on Google

Musk Pushes Back on Altman’s 5% DRAM-Use Estimate in Local AI Debate

OpenAI CEO Sam Altman argued cloud systems could use scarce memory more intensively, but did not define which computers his 5% estimate covered.

An unmarked memory module rests across two computer motherboards / TokenPost.ai
An unmarked memory module rests across two computer motherboards / TokenPost.ai

Elon Musk pushed back on OpenAI CEO Sam Altman’s estimate that personal computers use DRAM 5% of the time, challenging his case that cloud systems can use scarce memory more intensively.

Musk posted on X Nov. 1: “The people deserve to have their RAM too!”

In an Oct. 29 interview, Altman discussed local AI inference, the process of running models on a user’s own device. He argued that shared cloud systems could put limited memory to more intensive use than personal computers.

Altman did not explain how he calculated the 5% figure or specify which computers it covered. The interview gave no comparable utilization figure for cloud data centers.

Samsung Electronics projected Oct. 29 that AI-driven memory demand in 2027 would far exceed industry supply relative to 2026. Micron said June 24, 2026, that industry demand for DRAM and NAND significantly exceeded industry supply, with tight conditions expected to continue beyond calendar 2027.

TokenPost previously covered Altman’s DRAM comments.

Simon Yoon

Reporter

Simon Yoon reports on blockchain technology for TokenPost. Send corrections or tips to info@tokenpost.com.

Loading…