Micron’s earnings surge sparks memory mania, draws Nvidia comparisons

Micron Technology just posted the kind of earnings report that makes Wall Street forget how to breathe. Revenue hit $41.46 billion for fiscal Q3 2026, a roughly 345% increase year-over-year from $9.3 billion. Adjusted earnings per share came in at $25.11, blowing past analyst estimates. The stock responded by jumping 15% on June 25.

Profits surged nearly 15-fold compared to the same period last year, according to FT Alphaville. Gross margins reached approximately 85%, a figure that would make most chipmakers weep with envy.

The AI memory machine

DRAM prices have been surging as data centers race to build out AI infrastructure, and high-bandwidth memory, or HBM, has become the bottleneck everyone’s scrambling to fill. HBM sits right next to AI processors, and without enough of it, even the most powerful GPU is just an expensive paperweight. Nvidia’s chips need massive amounts of HBM to function, which means Micron, Samsung, and SK Hynix, the three companies that control nearly all global DRAM and HBM supply, are sitting in an extraordinarily comfortable position.

Micron has reportedly sold out its entire HBM capacity through 2026, with supply tightness expected to persist well into 2027. Analysts are forecasting that the three major memory makers will maintain pricing power for at least two more years, thanks to the consolidated nature of the market.

Micron’s Q2 FY2026 results, reported on March 18, showed revenue of $23.86 billion, a 196% year-over-year increase, with gross margins around 75%. The Q3 acceleration from that already-elevated baseline is what’s drawing the Nvidia comparisons.

The trillion-dollar club

Micron crossed the $1 trillion market capitalization threshold in May 2026, reportedly one of the fastest companies ever to reach that milestone.

Nvidia’s margins expanded dramatically as data center demand took off, and the company maintained pricing power because it had no real competition in high-end AI training chips. Micron shares the market with Samsung and SK Hynix, but the trio collectively controls enough supply that the competitive dynamics function more like an oligopoly than a free market.

What this means for crypto and tech investors

Bitcoin miners have been pivoting toward AI and high-performance computing workloads for the past two years, and companies like Core Scientific and Hut 8 have repositioned themselves as AI-adjacent data center operators. When memory chip demand is this intense, it signals that the broader AI infrastructure buildout is accelerating, not plateauing.

DRAM and NAND supply tightness can ripple into pricing for adjacent components, affecting the cost structure for ASIC manufacturers and GPU-based mining operations.

A 345% revenue increase and 15-fold profit growth suggest the AI memory cycle is still in its expansion phase. The consolidated market structure—three companies controlling virtually all supply—supports pricing power as long as demand stays strong, but it also means any demand slowdown would hit all three players simultaneously.