AI CRITIQUE

Teknik şema çizimlerinin üzerinde, üç siyah çip bulunan küçük bir bellek modülü duruyor. Sıcak ışıklı arka planda bir kalem ve rulo halinde bir kağıt görünüyor. — Temsilî görsel — FLUX.1-dev ile bilgisayarımda üretildi; haberin gerçek fotoğrafı değil.Illustrative image — generated with FLUX.1-dev on my computer; not a real photograph of the story.

CXMT to launch 4F² architecture DDR5 RDIMM by end of 2026

Read the article English (automatic translation)

Summary

Chinese memory maker CXMT announced it will ship DDR5 RDIMM modules with a 4F² architecture for data centers by the end of 2026. The company claims that its vertical transistor design reduces cell area by approximately one-third compared to the industry-standard 6F² layout. This announcement triggered a drop of over 7% in Samsung and SK hynix shares and approximately 4% in Micron stock.

Why it matters

If CXMT releases 4F² server DRAM two years ahead of Samsung's stated 2028 target, it will directly challenge the dominance of Korean and American rivals in the highest-margin DRAM segment. This move also signals a shift in CXMT's strategy from traditional consumer and appliance markets to the server market, where pricing power and volumes are significantly higher. A reliable Chinese player in server DRAM could reshape supply chain balances and price dynamics for cloud giants and OEMs.

Highlights

The 4F² vertical transistor design theoretically offers about 50% more cells per chip compared to 6F². However, speed, capacity, or efficiency figures have not yet been disclosed.
CXMT cites over 300 patent applications and articles dated 2021, 2023, and 2025. Nevertheless, the first commercial product remains a roadmap claim rather than a shipped SKU.
Following the news, Samsung and SK hynix shares fell by more than 7%, while Micron dropped by about 4%. This reaction may be considered disproportionate for a product that has not yet been proven in mass production.

Critical take

The article is essentially a restatement of a single executive's roadmap claim. There are no independent benchmarks, yield data, or verified speed and capacity specifications. Despite this, the news is framed as a 'breakthrough' that has already moved global markets. Even the analysts quoted in the article warn that stabilizing mass production will take a 'significant amount of time,' emphasizing that the 2026 launch date is a target, not a guarantee. Stock sales may be pricing in a scenario that is still years away from reality.

Bu kritik, yerel yapay zeka modeli Qwen3.8-27B tarafından yazılmıştır. Haberi ben seçtim; kritik metnini yayımlanmadan önce tek tek denetlemiyorum.This critique was written by the local AI model Qwen3.8-27B. I chose the story; I do not review each critique before it is published.

Back to the issue