Blog

DEEP RESEARCH · SAMSUNG ELECTRONICS

[Samsung Electronics] Buy — Prior-High Breakout Attempt, HBM4 as the Swing Factor

A small entry on the prior-high breakout attempt — and notes on the HBM4 1-C race and how the market values five to ten years out

Written: 2025-06-27 · Trade note + HBM4 competitive landscape · Original Naver Blog post

You are responsible for your own investment decisions. This material is research and is not a recommendation to buy or sell.

0. Bottom line first

The chart looked like a prior-high breakout attempt, so I entered. In HBM4, while Hynix runs on 1-B, Samsung running on 1-C is a comparison worth watching — but Samsung has historically over-promised on HBM, so we need to see real customer results. Hynix's relatively cheaper valuation might already reflect the market's expectation that the technical moat will narrow over the next five to ten years.

1. Why I entered

The chart appears to be attempting a breakout above the prior high, so I entered with a small position. This is a bet on the price action itself rather than a fresh fundamental thesis.

Interpretation: This is not a fundamentally driven buy. I am riding the chart trigger first, and I will use upcoming events (HBM4 sample shipments, customer qualification results) to re-judge sizing.

2. HBM4 — 1-B (Hynix) vs 1-C (Samsung)

SK Hynix

HBM4 on 1-B

Current market leader, perceived to be progressing HBM4 on the 1-B node.

Samsung

HBM4 on 1-C

One node finer (1-C). If it works, a chance to claim a comparative edge.

What to watch

Sample results & qualification

NVIDIA / AMD evaluation results will decide actual share.

Official fact: According to The Bell, Samsung Electronics is in the final stage before shipping 12-high samples of 6th-generation high-bandwidth memory (HBM4). If internal evaluations are positive, the plan is to supply HBM4 12-high samples to major customers such as NVIDIA and AMD in early July. Samsung HBM4 12-high sample shipment 'imminent' (The Bell)

The Bell report: Samsung HBM4 12-high sample shipment imminent

Interpretation: If Samsung can really build HBM4 on 1-C, it could push density and power efficiency ahead of a 1-B competitor. But Samsung has a track record in HBM of strong roadmaps that arrived late, so I stay conservative until 12-high samples actually pass customer evaluation.

3. Valuation — another reading of "Hynix looks cheap"

People often say Hynix looks cheap on valuation. But I think it is also possible that the market always values not on the short term but five to ten years out, so it may be pricing Hynix on the expectation that its technical moat will soon erode.

Interpretation: What looks cheap on near-term EPS could be the market already partially pricing a future scenario where share and margins compress. In that case, Samsung actually delivering on HBM4 1-C itself becomes the trigger that closes the valuation gap.

4. Related read

For a deeper write-up on HBM4 technology competition and custom HBM, see the prior post: [Semis] HBM4 Samsung vs SK Hynix and Custom HBM.

Thumbnail for HBM4 Samsung vs SK Hynix and Custom HBM post