The memory sector has taken a sharp hit recently.
Western Digital has fallen significantly, and Micron is also down. A massive circuit-breaker-triggering drop in South Korean equities has severely damaged market sentiment. Many retail investors are now torn over whether memory stocks have already topped out and whether AI demand has been overhyped.
During today’s live session, our analyst shared a perspective that starkly diverges from the market consensus—not simply arguing whether prices 'can rise' or 'can’t rise,' but reframing the issue from an entirely different angle.

1. The analyst first posed a question: How is this memory cycle different from 2010?
Between 2010 and 2012, the memory sector also experienced a major rally. At that time, smartphones exploded in popularity—Iphone 4 went global, and the shift from feature phones to smartphones suddenly spiked demand for memory chips.
But what happened afterward? That memory rally didn’t last long. Consumer electronics like smartphones have a key trait—they’re discretionary purchases. If memory prices are high this year, I can just delay upgrading my phone by a year with little impact. This is consumer (B2C) demand, which can wait.
However, AI-driven memory demand this time follows a completely different logic than smartphones.
The analyst put it bluntly during the members-only live session: 'If large AI model companies don’t buy memory and computing power this year, they’ll fall behind competitors next year.' This is enterprise (B2B) demand—it can’t wait.
The instructor also gave an analogy—like during the 2020 pandemic, when securing masks meant survival and failing to do so meant being left behind. Today’s large AI model companies face the same situation: those who secure chips and storage this year will surge ahead, while those who don’t will fall behind. In the U.S. market, at most only two players in this industry will ultimately survive—no one dares lag.