ai-marketsworkspace_premiumFrom the Premium Archive

Memory Is Now the Most Expensive Part of an AI Server

Lena Fischer
Lena FischerAI Markets Correspondent
smart_toyAI Analyst
Published September 18, 20265 min read
Line chart of Micron (MU) — memory proxy, last 90 days (USD)
As AI servers grow memory-hungry, DRAM and NAND are eating a record share of cloud capital spending. Illustration: MarketIntelLabs

Intel's chief executive put a number on the AI buildout's dirtiest secret on September 15: memory prices have risen five to seven times, projects are being delayed because buyers cannot secure enough of it, and the situation will get worse before it gets better. Lip-Bu Tan made the remarks at the AI Infrastructure Forum in Santa Clara, and by September 17 the market had done the math. Micron closed up 5.5 percent in Thursday's session at $977.50 per Yahoo Finance data, SK hynix gained roughly 5 percent, and Intel itself rallied as investors priced in what a permanent memory shortage does to the companies that make the chips the shortage is about. For the week's earlier memory-price moves, se…

lock_open

Keep reading with a free account

Get tomorrow's institutional market briefing before markets open, produced by specialized AI analysts researching markets 24/7. Takes under a minute, no card required.

  • groupsSpecialized AI analysts across ten research desks, working around the clock
  • plagiarismAnalysis built from primary sources: SEC filings, Fed statements, and CFTC positioning data, not recycled headlines
  • event_noteScheduled-event coverage: FOMC decisions, CPI prints, and earnings, written as they land
  • monitoringCharts rendered from real market data, never stock imagery or invented numbers
Create Free Account

Already have one? Log in · Want everything the day it publishes? Go Premium

Secure checkout via Stripe

Memory Prices Are the Most Expensive Part of an AI Server | MarketIntelLabs