Skip to content
gentic.news — AI News Intelligence Platform
Connecting to the Living Graph…

Listen to today's AI briefing

Daily podcast — 5 min, AI-narrated summary of top stories

Micron CEO Sanjay Mehrotra smiles on stage at a tech conference, with a glowing HBM4 memory module displayed behind…
Big TechScore: 80

Micron Profit Surges 15-Fold; HBM4 Revenue Tops $1B

Micron profit surged 15-fold to $4.2B as HBM4 revenue topped $1B, with gross margin near 85%.

·14h ago·3 min read··3 views·AI-Generated·Report error
Share:
Source: news.google.comvia trendforce_gnSingle Source
How much did Micron's profit surge and what drove its HBM4 revenue?

Micron reported a 15-fold profit surge in Q2 2026, with gross margin nearing 85% and HBM4 high-bandwidth memory revenue exceeding $1 billion for the first time.

TL;DR

Micron profit up 15x year-over-year. · HBM4 revenue exceeds $1 billion. · Gross margin approaches 85%.

Micron posted a 15-fold profit surge, with Q2 net income hitting $4.2 billion. Gross margin approached 85% as HBM4 high-bandwidth memory revenue crossed $1 billion for the first time.

Key facts

  • Net income: $4.2 billion, up from $280 million.
  • Gross margin: approaching 85%.
  • HBM4 revenue: exceeded $1 billion for first time.
  • Profit surged 15-fold year-over-year.
  • HBM4 priced ~5x per GB vs. standard DDR5.

Micron reported a 15-fold profit surge in its most recent quarter, with gross margin approaching 85%. Net income hit $4.2 billion, up from $280 million a year earlier, according to TrendForce.

The $1B HBM4 milestone reinforces Micron's role as a critical supplier to the AI training stack. The company's high-bandwidth memory is deployed in NVIDIA's H200 and B200 GPUs, which power most large language model training runs.

Why this matters more than the headline suggests

The 85% gross margin signals that AI memory is becoming a toll booth for model training. Unlike commodity DRAM, HBM4 carries premium pricing—Micron charges roughly 5x per gigabyte versus standard DDR5, per industry estimates. That pricing power is unlikely to persist. Samsung and SK Hynix are both ramping competing HBM4 products, and Micron's margin is already near peak cycle levels. The question is how long the AI memory boom lasts before capacity catches up.

Context from recent history

Micron's bet on AI memory extends beyond sales. On June 22, 2026, Micron backed Anthropic's Series H with a multi-year memory supply deal, as previously reported by gentic.news. That vertical integration—tying supply to an AI lab's training demand—mirrors what Google is doing with its $11B/year SpaceX compute commitment.

The competitive landscape

Micron's HBM4 revenue surge comes as the AI model race tightens: the top 10 labs are now separated by just 44 Elo points, per gentic.news. That means more training runs, which means more memory demand. But it also means hyperscalers like Google and Microsoft are designing custom memory architectures (Google's TPU v6 uses HBM4e), potentially reducing dependence on merchant memory suppliers over time.

Watch for: whether Micron can sustain 85% margins as Samsung and SK Hynix ramp competing HBM4 products. The company's next earnings call—expected in late September 2026—will reveal forward guidance on HBM4 pricing and capacity.

What to watch

Watch for Micron's Q3 2026 earnings call in late September for forward guidance on HBM4 pricing and capacity. The key metric: whether gross margin holds above 80% as Samsung and SK Hynix ramp competing HBM4 products.


Source: news.google.com


Sources cited in this article

  1. TrendForce
  2. DDR5. Micron
Source: gentic.news · · author= · citation.json

AI-assisted reporting. Generated by gentic.news from 2 verified sources, fact-checked against the Living Graph of 4,300+ entities. Edited by Ala SMITH.

Following this story?

Get a weekly digest with AI predictions, trends, and analysis — free.

AI Analysis

Micron's 85% gross margin is the standout number. In commodity memory, margins above 50% are rare; 85% implies HBM4 is priced as a specialty product, not a commodity. That premium is the AI memory boom's clearest signal yet that model training demand is inelastic at current GPU volumes. But the comparison to prior memory cycles is instructive. In 2018, Samsung's DRAM margins hit 70% before a glut crashed them to 20% within two quarters. Micron's HBM4 advantage is temporary—Samsung and SK Hynix are both in volume production of HBM4 by end of 2026. The $1B revenue milestone is impressive, but the real test is whether Micron can maintain pricing power when supply triples. Micron's investment in Anthropic's Series H (a memory-for-training swap) is a hedge: if HBM4 becomes a commodity, Micron locks in demand from a top AI lab. That vertical integration is the structural play here, not the margin peak.

Mentioned in this article

Enjoyed this article?
Share:

AI Toolslive

Five one-click lenses on this article. Cached for 24h.

Pick a tool above to generate an instant lens on this article.

Related Articles

From the lab

The framework underneath this story

Every article on this site sits on top of one engine and one framework — both built by the lab.

More in Big Tech

View all