T-REX 2X Long SKHY Daily Target ETF

HYNX

23.5901.30% ( -0.310 )

Trading: Oct 9, 11:57:39 (EDT)

Share your thoughts

What's on your mind?

0 / 2,000

933 followers · 0 opinions

▲Bullish50%

50%Bearish▼

D
Dolphin ResearchOct 9, 2026 at 08:33 AM
PinnedSK hynix: Can LTAs smooth the cycle? How far can valuation go?
HBM prices have surged, and SK hynix shares have jumped as much as 10x over the past six months, but Samsung Electronics and Micron Technology are rapidly grabbing supply-chain share.
In Part 1, we mapped the DRAM/HBM supply-demand gaps and identified the cycle inflection; please refer to that piece. However, an industry upcycle does not guarantee steady profits at the company level.
Here we stay focused on SK hynix, analyzing its partnerships with NVIDIA and major ASIC customers. We also compare long-term agreement strategies across the Big Three. We then model forward earnings and assess the current valuation, as incremental memory demand in the AI era is concentrating with a handful of XPU vendors and cloud players...
+12
SK HynixDeep ResearchAppraisal value
D
Dolphin ResearchOct 8, 2026 at 10:52 AM
PinnedSK hynix: How Long Can the AI-led Memory Supercycle Last?
The last two pieces from Dolphin Research outlined the underlying logic and trajectory of the 'memory' rally from a tech lens, while this one shifts back to the company level. It zeroes in on the most representative HBM leader of this cycle — $SK Hynix(SKHY.US).
SK hynix’s current position is delicate. Memory prices have surged and HBM remains tight, driving a step-change in earnings.
The stock at one point rallied ~10x within six months.
On the other hand, Samsung and Micron are rapidly plugging into the supply chain of core customers such as Nvidia, shifting Hynix’s edge from exclusive to merely leading.
As a manufacturing company...
+23
SK HynixDeep ResearchAppraisal value
B
Boss's BossOct 9, 2026 at 07:51 AM
$XL2CSOPHYNIX(07709.HK) added a position of 8,000 shares at 33.8 today; currently holding: 19,800 shares. Stage cost: 36.88
D
Dolphin ResearchOct 8, 2026 at 06:03 AM
$Taiwan Semiconductor(TSM.US) TSMC does not produce a single DRAM chip, yet it is becoming central to high‑performance HBM. From HBM4 onward, two critical steps — a customized base layer and co‑packaging of memory with the GPU — run through TSMC's lines.
Entry 1: the base layer. It serves as the bottom‑stack 'relay station' in HBM. HBM3e and earlier used simpler 20nm flows that DRAM vendors could make in‑house; with HBM4, nodes move to 12nm and flagships to 2/3nm, forcing outsourced fab to TSMC (Samsung can self‑manufacture).
The bar is higher because a doubled bus width raises PHY precision requirements, and functions like the memory controller that lived in the XPU are migrating to the base layer. As the base layer shifts from off‑the‑shelf to customer‑specific parts — e.g., NVDA's NVHBM uses a proprietary protocol claiming +30% bandwidth and −15% power — reliance on TSMC's advanced nodes deepens.
Entry 2: CoWoS packaging. HBM's high bandwidth depends on co‑packaging with the XPU on a silicon interposer, primarily via TSMC's CoWoS, where CoWoS‑L is already mainstream with a >60% share.
Capacity cadence matters more: DRAM wafers will not meaningfully free up until 2027 (+15%~20%), while CoWoS capacity in 2027 doubles vs. 2026 — packaging is expanding faster than DRAM bits.
The bottleneck and value are shifting toward packaging and foundry. How much HBM a DRAM vendor can deliver now depends not only on its own lines, but also on how much base‑layer wafer capacity and CoWoS slots it can lock at TSMC.

Take SK hynix's HBM4: it only handles die fabrication and stacking, while the base layer is co‑designed with NVDA and manufactured by TSMC, and final packaging is also completed at TSMC. If the industry moves toward 'fewer layers, more dies', it saves DRAM wafers but consumes more packaging capacity.
The choke point shifts further from 'DRAM vendor expansion' to 'TSMC ramp'.
D
Dolphin ResearchOct 8, 2026 at 06:01 AM
$SK Hynix(SKHY.US) Stacking over a dozen memory dies and then bonding them, SK hynix can finish the job in one shot. Samsung and Micron have to press and bond layer by layer.
Its proprietary bonding process is one of SK hynix's key competitive edges.
SK hynix uses MR-MUF: align all layers once, complete bonding in a single pass, then fill gaps with resin. Samsung and Micron use TC-NCF, bonding each layer through repeated heat and pressure.
The toolsets are fundamentally different.
Two gaps stand out. One-shot bonding reduces steps, boosting efficiency and yield, while layer-by-layer bonding means one defect can scrap the entire stack.
Repeated thermal compression also warps and cracks ultra-thin wafers more easily, pushing yields even lower.
This flows straight to GPM, giving SK hynix a structural edge.
So why don't Samsung and Micron just copy it?
First, SK hynix has locked up the process and key liquid resin materials with patents. Second, Samsung and Micron have sunk heavy capex into the TC route, making a pivot costly.
So they are effectively skipping this generation and betting on next-gen 'hybrid bonding' to leapfrog.
A bonding process itself is a ready-made moat. Bonding technology will be a hidden line separating winners from laggards among memory makers going forward.
Samsung and Micron appear more inclined to develop next-gen hybrid bonding directly to attempt a leap. $SKHY 2X Short ETF(SKDD.US) $T-REX 2X Long SKHY Daily Target ETF(HYNX.US) $SKHY 2X Long ETF(SKHX.US)
B
Boss's BossOct 8, 2026 at 01:59 AM
$XL2CSOPHYNIX(07709.HK) The distance to a pullback isn't too far. In principle, the profitability of storage companies remains very strong. It's just that market divergence is significant. Yesterday in Singapore, I discussed this topic with a friend who does macro trading. One idea came up: the reason for such wide divergence right now is concern over the growth rate and sustainability of AI-driven earnings. Is there a hidden issue behind this? Currently, enterprises (and their owners) struggle to effectively measure the "value-add" brought by AI adoption. For instance, you use AI—you've created how much new value? Saved how much cost? You might feel you can't live without it and love Claude or ChatGPT, but what is its actual monetary value? This problem is even more complex for enterprises. It will take some time for at least some companies and owners to figure this out. This is also an issue the industry needs to ponder.
B
Boss's BossOct 7, 2026 at 05:18 AM
$XL2CSOPHYNIX(07709.HK)阶段性 buying sufficient. The positions that were previously cleared have returned.
D
Dolphin ResearchOct 6, 2026 at 11:42 AM
AI can't rewrite the DNA — how long can SanDisk's rally last?
In two earlier pieces, 'From Leaves to Flowers: How AI Inference Can Revive NAND?' and 'NAND’s Prolific Nature: Why Can SanDisk Defend 80% GPM?', Dolphin Research argued that AI will not only lift NAND demand but could also break its long-standing 'silicon cycle' fate.
Under an inference-led paradigm, NAND is shifting from a peripheral data warehouse to an essential element on the compute path.
With the move from training to inference and longer context windows, today’s SSDs have effectively become a 'token battery' that stores compute...
+36
SandiskDeep Research
B
Boss's BossOct 6, 2026 at 08:24 AM
$XL2CSOPHYNIX(07709.HK) has reached the water level at close 😄. Continuing to oscillate within the range, the core issue is excessive divergence.
B
Boss's BossOct 6, 2026 at 06:02 AM
$XL2CSOPHYNIX(07709.HK)Surging soon, time for a happy afternoon tea
B
Boss's BossOct 6, 2026 at 05:06 AM
$XL2CSOPHYNIX(07709.HK) Time is tight and the task is heavy, just get it done.
B
Boss's BossOct 6, 2026 at 02:57 AM
$XL2CSOPHYNIX(07709.HK) Quickly built a position of 2,800 shares. The joy of ups and downs.

T-REX 2X Long SKHY Daily Target ETF

PinnedSK hynix: Can LTAs smooth the cycle? How far can valuation go?

PinnedSK hynix: How Long Can the AI-led Memory Supercycle Last?

AI can't rewrite the DNA — how long can SanDisk's rally last?