Semiconductor Value Chain on Korea Invest Insights

ARM's Rally — The Next AI Bottleneck Is CPU Orchestration, Optical Links and Power Integrity

Fri, 22 May 2026 21:40:00 +0900

Related series: NVIDIA Q1 FY27 and Korean AI infrastructure / Samsung Electro-Mechanics’ KRW 1.5T silicon-capacitor contract / Understanding MLCCs and silicon capacitors / Samsung Electronics rerating thesis

ARM’s rally is not simply a statement that “CPUs matter again.” More precisely, AI servers are being redefined from GPU boxes into systems that combine CPUs, XPUs, HBM, optical links, power-stabilizing components, high-speed substrates and test infrastructure. ARM is the visible price reaction. But at the current valuation, a large part of the five-year success case is already embedded. For Korean investors, the more useful signal is the next bottleneck: Marvell’s custom silicon and optical links, Samsung Electro-Mechanics’ silicon capacitors, SK hynix and Samsung Electronics memory, and the high-speed substrate / test-socket layer.

Key Takeaways

ARM’s rally is a reclassification event: from mobile IP royalty company toward AI data-center CPU platform company. As AI agents run continuously, call tools, move data and coordinate GPUs, ASICs and memory, the CPU becomes the orchestration layer of the AI rack.

The external catalyst was NVIDIA’s Q1 FY27. NVIDIA reported $81.6B in revenue, $75.2B in Data Center revenue, and Q2 guidance of $91.0B ±2%. The market read that as acceleration, not peak-out, and began repricing the layers below GPUs: CPU, memory, interconnect, power and substrates. (NVIDIA)

ARM’s own story is strong. FY26 revenue was $4.92B, royalty revenue $2.61B, licensing revenue $2.31B, and non-GAAP EPS $1.77. The Arm AGI CPU is positioned as a data-center CPU platform with Meta as lead partner, and Arm says FY27-FY28 customer demand is already above $2B. (Arm)

The problem is price. Research OS local DB shows ARM closed at $298.23 on May 21, 2026. Against FY26 non-GAAP EPS of $1.77, that is roughly 168.5x earnings. Using a market capitalization around the low-$300B range and FY26 revenue of $4.92B, the stock trades around the 60x sales zone. The thesis is real; the stock is not cheap.

The practical alpha is therefore not ARM itself, but the bottlenecks ARM’s move points toward: Marvell’s custom silicon and optical fabric, Samsung Electro-Mechanics’ silicon capacitor, SK hynix / Samsung Electronics HBM and DRAM, high-speed substrates and test sockets.

The one-line conclusion: ARM is the signal. The alpha sits in the bottlenecks.

1. What ARM’s Rally Really Means

Calling this a “CPU renaissance” is only half right. The deeper message is that the constraint is moving.

In 2023-2024, the core AI infrastructure bottleneck was GPUs. In 2025, HBM and advanced packaging rose to the front. In 2026 and beyond, as AI racks become larger systems, CPU orchestration, memory movement, optical links, power integrity, substrates and testing all become limiting factors.

GPU scarcity
→ HBM / packaging scarcity
→ CPU orchestration / memory movement / optical interconnect bottlenecks
→ package-level power integrity / high-speed substrate / test bottlenecks

AI agents are not ordinary web requests. They call models, use tools, read documents, query databases and store intermediate outputs. GPUs handle compute, but CPUs coordinate task order, memory management, network I/O, security, error handling and system operation.

This makes the CPU the control plane of the AI rack. If the GPU is the engine, the CPU is the traffic controller. ARM’s rally reflects the market starting to price that role.

Arm’s FY26 results support the narrative. The company reported FY26 revenue of $4.92B and Q4 revenue of $1.49B, with record annual royalty and licensing revenue. Data-center royalties more than doubled year over year. (Arm)

But the right investment question is not “is ARM’s thesis right?” It is “does the current price still leave enough risk-reward?”

2. Arm AGI CPU: From IP Vendor to Silicon Platform

Arm has put the Arm AGI CPU at the center of the story. It describes the product as production silicon for agentic AI data centers, with Meta as lead partner. Arm also says commercial systems are available to order through ASRock, Lenovo, Quanta and Supermicro. (Arm)

The old Arm model was:

IP licensing
→ customers design and manufacture chips
→ Arm receives license fees and royalties

The AGI CPU extends that model:

IP + compute subsystems + production silicon
→ customers deploy the Arm platform faster at data-center scale
→ Arm gains more platform control

This is powerful, but it changes the risk profile. Arm is no longer only a neutral blueprint provider. As it sells its own silicon, customers and licensees may begin to see it as a potential competitor.

That is why regulatory risk matters. Bloomberg has reported that the FTC is looking into whether Arm could restrict or degrade license access while expanding its own chip ambitions. (Bloomberg)

The AGI CPU is a long-term call option, but it also makes Arm heavier, more complex and more exposed to channel conflict.

3. Valuation: Correct Thesis, Expensive Stock

The uncomfortable part of the ARM rally is the math. Directionally, the thesis makes sense. Valuation already discounts a long runway.

Research OS local DB shows ARM closed at $298.23 on May 21, 2026. Against FY26 non-GAAP EPS of $1.77:

FY26 non-GAAP P/E
= 298.23 / 1.77
= about 168.5x

Against FY26 revenue of $4.92B and a market cap in the low-$300B range:

FY26 P/S
= roughly low-$300B / $4.92B
= about mid-60x sales

This is not a price for “good current earnings.” It is a price for a 2030-2031 scenario: higher data-center CPU penetration, AGI CPU revenue, higher royalty rates and Armv9 / Neoverse expansion.

If Arm reaches something like $25B of revenue and $9+ of EPS by FY31, today’s price can be explained. But that is a five-year success case. Investors need to separate a correct story from a good entry.

Our view is Avoid at current price / Wait. The company is not the problem. The entry price is.

4. Better Alpha 1: Marvell and the Connectivity Bottleneck

If ARM represents CPU orchestration, Marvell represents custom compute plus optical connectivity plus switching. This may be the more actionable derivative idea.

Marvell reported FY26 revenue of $8.195B and non-GAAP EPS of $2.84. It expects FY27 revenue to approach $11B, with data center leading growth and interconnect revenue growing more than 50%. (Marvell)

Marvell is not just a networking chip company.

Layer	Meaning
Custom AI silicon	Design partner as hyperscalers build their own XPUs
Optical interconnect	Next bottleneck for intra-rack and inter-rack data movement
PCIe / CXL switching	Connects CPU, GPU, ASIC and memory pools
NVLink Fusion partnership	Strategic position inside heterogeneous NVIDIA-connected AI infrastructure

The Celestial AI acquisition adds a photonic scale-up platform. Marvell has said it expects the business to reach a $500M annualized run-rate by Q4 FY28 and a $1B annualized run-rate by Q4 FY29, if milestones are met. (Marvell Celestial)

The caveat: Marvell has already rallied. Research OS local DB shows MRVL closed at $190.69 on May 21, 2026. At this level, the May 27 earnings report needs to confirm the FY27 revenue path, data-center demand, custom silicon and optical interconnect momentum.

Our view: Wait / Buy on pullback.

5. Better Alpha 2: Samsung Electro-Mechanics and Power Integrity

In Korea, the clearest second-order bottleneck is Samsung Electro-Mechanics.

On May 20, 2026, Samsung Electro-Mechanics announced a silicon-capacitor supply contract worth approximately KRW 1.5T over two years, from January 1, 2027 to December 31, 2028. The company explained that silicon capacitors are installed inside high-performance semiconductor packages such as AI server GPUs and HBM to improve power stability. (Samsung Electro-Mechanics)

This is the same direction as ARM’s rally. As AI racks become systems rather than GPU boxes, CPU, GPU, HBM and ASIC packages become denser and draw power faster. Voltage noise and power fluctuations become a bottleneck.

Samsung Electro-Mechanics’ thesis is not “MLCC demand rises.” It is:

Old SEMCO:
MLCC + camera modules + FC-BGA

New SEMCO:
AI package power-integrity components
+ FC-BGA
+ AI-grade MLCC

A silicon capacitor does not replace all MLCCs. It is a high-performance complement used inside or very close to AI GPU / HBM packages. This can reclassify Samsung Electro-Mechanics from a passive-parts cycle name into an AI package power-integrity supplier.

The risks are clear: the customer is undisclosed, 2026 earnings impact is limited, margins are not disclosed, and the stock has already reacted. Our view remains Watchlist → Buy on pullback.

6. Better Alpha 3: SK hynix and Samsung Electronics Memory

ARM’s rally is positive for Korean memory as well. More CPU orchestration means more CPU-side memory, more data movement to HBM, more LPDDR / DDR / CXL memory pooling, and more server DRAM.

GPU-centric server
→ GPU + CPU + DPU + NIC + switch ASIC + HBM + DRAM
→ higher rack-level memory capacity and bandwidth

SK hynix is the cleanest HBM and server DRAM beneficiary, but it is also a crowded winner. The view is Hold / Pullback Buy.

Samsung Electronics is different. It is the catch-up option: HBM4, SOCAMM2, server DRAM, eSSD, HBM4 base die and foundry optionality. Since Arm AGI CPU manufacturing is tied to TSMC 3nm, one should not overstate direct Samsung Foundry exposure. But AI rack diversification still improves the value of Samsung’s memory and packaging ecosystem.

Company	Role	View
SK hynix	Proven HBM winner	Hold / Pullback Buy
Samsung Electronics	HBM catch-up + broad IDM option	Watchlist / Buy on confirmation

Samsung needs evidence: HBM4 qualification, volume, margins and foundry loss reduction.

7. Better Alpha 4: High-Speed Substrates and Test Sockets

As AI racks evolve from single GPU boards into dense systems, substrates and testing become second-order winners.

High-speed substrates and PCBs. More CPUs, GPUs, ASICs, HBM, NICs and switch ASICs mean more high-speed signaling and more complex board design. Isu Petasys, Daeduck Electronics, TLB, Korea Circuit and Simmtech are candidates. But direct orders, customer qualification, ASP and margin must be verified.

Test sockets. As chips become faster, hotter and more pin-dense, sockets and probe-card requirements rise. ISC, LEENO Industrial and TSE are relevant candidates. This layer can carry higher margins and less capex cyclicality than substrates, but product mix matters.

Layer	Candidates	What to Verify
High-speed PCB	Isu Petasys, Daeduck, TLB, Korea Circuit, Simmtech	AI-server revenue, layer count, ASP, OPM
Test sockets	ISC, LEENO, TSE	ASIC/HBM/CXL customers, product mix, new socket revenue
Package substrate	Samsung Electro-Mechanics, Daeduck, Korea Circuit	FC-BGA utilization, qualification, margins

The rule: do not buy the phrase “AI exposure.” Buy verified customers, orders and margins.

8. Strategy

Name / Domain	View	Reason
ARM	Avoid at current price / Wait	Correct thesis, but very rich valuation
Marvell	Wait / Buy on pullback	Custom silicon + optical + switching bottleneck
Samsung Electro-Mechanics	Watchlist → Buy on pullback	Silicon capacitor can reclassify it as AI package power-integrity supplier
Samsung Electronics	Watchlist	HBM4, SOCAMM, server DRAM, eSSD, foundry option
SK hynix	Hold / Pullback Buy	Cleanest HBM winner, but crowded
High-speed substrates	Watchlist	Beneficiary, but verify customers and margins
Test sockets	Watchlist	Chip-complexity consumables beta

Catalysts

Marvell earnings on May 27, 2026.
Arm AGI CPU supply expansion and conversion from demand to revenue.
NVIDIA Vera Rubin second-half ramp.
Samsung Electro-Mechanics follow-on silicon-capacitor customers or capacity announcements.
Samsung Electronics HBM4 qualification.
Korean substrate / test-socket companies separating AI revenue in quarterly results.

Invalidation

Arm AGI CPU demand fails to convert because of supply, margins or channel conflict.
FTC / regulatory risk damages Arm’s licensing model.
Marvell loses custom design wins to Broadcom or optical revenue ramps too slowly.
Samsung Electro-Mechanics’ silicon-capacitor contract carries weak margins or lacks follow-on customers.
Korean substrate / test names report AI revenue growth without OPM support.

Final Line

ARM’s rally is a correct revaluation signal. AI servers are no longer GPU boxes. They are CPU, XPU, HBM, optical, power-integrity, substrate and test systems.

But buying ARM after the move may confuse the signal with the asset. The better question is: which bottleneck has not yet been fully priced?

Our answer: ARM is the signal. The alpha sits in the bottlenecks.

Evidence Map

[Fact]

NVIDIA Q1 FY27 revenue was $81.6B, Data Center revenue was $75.2B, and Q2 revenue guidance was $91.0B ±2%. (NVIDIA)
Arm FY26 revenue was $4.92B, royalty revenue $2.61B and licensing revenue $2.31B. Arm said AGI CPU customer demand for FY27-FY28 was above $2B. (Arm)
Marvell FY26 revenue was $8.195B, non-GAAP EPS was $2.84, and the company expects FY27 revenue approaching $11B. (Marvell)
Samsung Electro-Mechanics announced a KRW 1.5T silicon-capacitor supply contract on May 20, 2026. (Samsung Electro-Mechanics)
Research OS local DB shows ARM closed at $298.23 and MRVL at $190.69 on May 21, 2026.

[Inference]

ARM’s rally signals a bottleneck shift from GPUs toward CPU orchestration, memory movement, optical interconnect and power integrity.
The better risk-reward may sit in Marvell, Samsung Electro-Mechanics, Korean memory, high-speed substrates and test sockets rather than ARM itself.

[Blocked]

Arm AGI CPU customer-level orders and margins.
Marvell customer-level custom silicon revenue.
Samsung Electro-Mechanics silicon-capacitor customer, ASP, margin and exact package position.
Direct ARM/NVIDIA/Marvell exposure for Korean substrate and test names.

This article is research commentary, not investment advice. US price data is based on Research OS local DB closing prices for May 21, 2026. Company data is based on official NVIDIA, Arm, Marvell and Samsung Electro-Mechanics materials. Scenario views are analyst frameworks and may be wrong. Data cut: May 22, 2026 KST.

Understanding MLCCs and Silicon Capacitors — What Samsung Electro-Mechanics' ₩1.5 Trillion Contract Reveals About the AI Package Power Bottleneck

Fri, 22 May 2026 00:00:00 +0000

📚 Related Series Samsung Electro-Mechanics Silicon Capacitor ₩1.5 Trillion Contract / Samsung Electro-Mechanics Hybrid Challenger / Samsung Electro-Mechanics MLCC & FC-BGA Business Deep Dive / Semiconductor Value Chain Hub / AI Substrate & PCB Hub

An MLCC is one type of capacitor — an ultra-compact ceramic component found in virtually every electronic device, functioning as a power stabilizer. A silicon capacitor is placed closer to the semiconductor chip than an MLCC, in some cases inside the package itself, to suppress instantaneous power fluctuations in AI GPUs and HBM. From an investment standpoint, what matters is the signal that “passive components” are graduating from simple commodity parts to AI packaging and power-integrity bottleneck components.

Key Summary

A capacitor stores electrical charge briefly and releases it on demand. For non-engineers, think of it as a water tank, shock absorber, or noise filter for an electrical circuit. An MLCC is the highest-volume ceramic multilayer version of that capacitor. Samsung Electro-Mechanics describes MLCCs as playing the role of a “dam” — holding electricity and releasing it in controlled amounts.¹

A silicon capacitor is also a type of capacitor. Unlike an MLCC, which stacks hundreds of ceramic layers, a silicon capacitor forms dielectric and electrode layers on a silicon wafer. According to Samsung Electro-Mechanics’ product documentation, it can be thinned to under 100 micrometers, embedded inside a package, and benefits from low parasitic inductance that is favorable for power stabilization.²

The essential point is not substitution but a shift in location. MLCCs are mounted primarily on the PCB surface or around the chip. Silicon capacitors can go inside the package, beneath the substrate, or directly adjacent to the die. Because AI GPUs and HBM draw sudden, large current spikes, proximity matters as much as raw capacitance — “how quickly can power be delivered” becomes as important as “how much can be stored.”

Samsung Electro-Mechanics’ May 20, 2026 silicon capacitor supply contract worth approximately ₩1.5 trillion is the first large-scale evidence of this structural shift. The contract runs from January 1, 2027 through December 31, 2028. The company stated that the product is embedded inside high-performance semiconductor packages — including AI server GPUs and HBM — to enhance power-supply stability.³

Investment judgment, however, must remain cold. This week (May 18–22), the three core MLCC names averaged +35.6%, handily outpacing the five core AI substrate names at +14.0%. Samsung Electro-Mechanics gained +32.7% on the week; Samwha Capacitor surged +56.4%. Theme momentum is unmistakable, but the efficiency of fresh capital deployment has diminished sharply.

The bottom line: the significance of silicon capacitors is not MLCC replacement — it is the expansion of the passive component battlefield from the PCB surface into the interior of AI semiconductor packages. Premium product lines at Samsung Electro-Mechanics and Murata may be re-classified as AI infrastructure supply chain assets. But a good industry shift and a good entry price remain entirely separate questions.

1. The Relationship Between Capacitors and MLCCs

A capacitor stores electrical charge and releases it rapidly when needed. When a semiconductor chip demands a sudden surge of current, a nearby capacitor supplies it, preventing voltage from collapsing. Conversely, when voltage spikes, the capacitor absorbs some of it, stabilizing the circuit. This is why capacitors serve the roles of energy storage, voltage smoothing, noise rejection, and power stabilization in electronic circuits.

An MLCC — Multi-Layer Ceramic Capacitor — is an ultra-compact capacitor constructed by stacking hundreds of thin ceramic dielectric layers alternating with metal electrode layers. Samsung Electro-Mechanics references production capability for high-capacitance MLCCs with up to 600 layers, noting that MLCC importance grows alongside the expansion of 5G, consumer electronics, autonomous vehicles, and the Internet of Things.¹

The hierarchy is straightforward:

Capacitor = the entire category of electrical stabilization components
MLCC = a mass-market, multilayer ceramic variant within that category
Silicon Capacitor = a high-performance, silicon-wafer-based variant optimized for ultra-close proximity

Both MLCCs and silicon capacitors are capacitors. They differ in where they are deployed and what performance demands they are designed to meet.

2. Capacitor vs. MLCC vs. Silicon Capacitor

	General Capacitor	MLCC	Silicon Capacitor
Category	Broad family	One type of capacitor	One type of capacitor
Simple analogy	Circuit water tank	Ultra-compact ceramic water tank	Ultra-proximity water tank embedded next to or inside the chip
Key material	Ceramic, aluminum electrolyte, tantalum, film, silicon, etc.	Ceramic dielectric + internal metal electrodes	Silicon wafer-based structure
Primary role	Energy storage, voltage smoothing, noise rejection	High-frequency noise filtering, power stabilization around chips, miniaturization	Suppression of instantaneous power transients in AI GPUs, HBM, and high-performance chips
Typical location	Power stages, boards, modules, motors, inverters	Primarily on the PCB, around chips, around modules	Inside packages, within substrates, immediately adjacent to the die
Advantages	Wide selection across applications	Small, low-cost, mass-producible	Extremely thin; can be placed close to the die; favorable for power stabilization
Limitations	Large performance variance by type	Limited in ultra-high-speed, ultra-proximate package power stabilization	High manufacturing complexity and cost; current addressable market concentrated in high-performance semiconductors
Primary growth vectors	Automotive, power grid, industrial, AI server	Automotive, AI server, smartphone, 5G, IoT	AI GPU, HBM, high-performance computing, high-performance mobile chips

Even a cursory look at TDK’s capacitor product portfolio illustrates the breadth of the category: small MLCCs, high-voltage ceramic capacitors, film capacitors, aluminum electrolytic capacitors, and power capacitors all sit under the same “capacitor” umbrella.⁴ “Capacitors are good” is too broad a statement. In investment terms, what matters is which capacitor, in which location, for which application.

3. Where They Are Used

3.1 General Capacitors

General capacitors appear in virtually every electrical circuit, from consumer electronics to power infrastructure.

Application	Primary Capacitor Type	Role
Smartphone, PC	MLCC, tantalum, polymer	Power stabilization around application processors, memory, and PMIC
Server, AI server	MLCC, polymer, silicon capacitor	Suppression of power transients in CPUs, GPUs, and HBM
Automotive electronics	MLCC, film, aluminum electrolytic, tantalum	Stabilization of ECUs, ADAS, inverters, chargers, and BMS
EV power systems	Film, high-voltage MLCC, aluminum electrolytic	High-voltage DC link, charging, power conversion
Industrial, power grid	Film, aluminum electrolytic	Power factor correction, motor drives, energy storage
Telecom, high frequency	Ceramic, silicon capacitor	High-frequency filtering, impedance matching

3.2 MLCC

The core application for MLCCs is power stabilization around chips. They appear in smartphones, automobiles, servers, network equipment, home appliances, and industrial machinery. As devices shrink, semiconductors switch faster, and the number of sensors and communication modules grows, the number of points requiring voltage stabilization increases proportionally.

MLCC importance in AI servers is also rising. TDK notes that as AI and cloud demand drives higher integration density and performance in data center servers, power density per rack and per server increases — making the performance and footprint of passive components a genuine design constraint.⁴ This is why the MLCC thesis has expanded beyond a simple smartphone-cycle recovery story into AI server power stabilization.

3.3 Silicon Capacitors

The core application for silicon capacitors is power stabilization inside packages. Samsung Electro-Mechanics describes silicon capacitors as ultra-compact, high-performance devices built on silicon wafers, embedded inside high-performance semiconductor packages — including AI server GPUs and HBM — to enhance power-supply stability.³

The company’s product documentation reinforces this: silicon capacitors form dielectric and internal electrodes on silicon, can be thinned to under 100 micrometers through wafer processing, and are well-suited for embedding inside packages. Low parasitic inductance is the key advantage for power stabilization.²

To demystify the terminology:

Parasitic inductance = an interference factor that delays the instantaneous response of current flow
Parasitic resistance = an interference factor that causes energy loss as current passes through
Power integrity = the ability to deliver stable, ripple-free voltage and current to a chip exactly when needed

AI semiconductors draw power in sudden, large spikes. A capacitor far from the die responds too slowly. The solution is to move the capacitor immediately adjacent to the chip — or inside the package itself. The investment case for silicon capacitors is not “higher capacitance” but faster response from a closer location.

4. Key Suppliers and Customer Base

4.1 MLCC Key Suppliers

Region	Key Companies	Strengths
Japan	Murata, TDK, Taiyo Yuden	Ultra-small, high-reliability, automotive-grade MLCC leadership
Korea	Samsung Electro-Mechanics	High-value MLCC for IT, automotive, and AI server; expanding into silicon capacitors
Taiwan / Greater China	Yageo, Walsin	Commodity, industrial, and mid-to-low price passive component portfolio
US / Japan-affiliated	KYOCERA AVX, KEMET/Yageo, Vishay	Specialty capacitors, automotive, industrial, and high-frequency product lines

The MLCC customer base is extremely broad: smartphone OEMs, PC and server manufacturers, automotive OEMs, Tier-1 suppliers, telecom equipment makers, and semiconductor module companies. For MLCC investing, demand mix by end-market and the share of high-value products matter more than exposure to any single customer.

4.2 Silicon Capacitor Key Suppliers

	Key Companies	Positioning
Murata	Leader in silicon capacitors and integrated passive devices	High-performance mobile, telecom, and high-performance computing silicon capacitors
Samsung Electro-Mechanics	Large-scale new entrant	Secured approximately ₩1.5 trillion supply contract for 2027–2028 delivery
KYOCERA AVX et al.	Specialty MOS capacitors	High-frequency, microwave, and specialty industrial product lines
Advanced packaging ecosystem	Captive or co-design options	Linked to interposer and in-package power integrity technology

The customer dimension is critical. MLCCs have a broad, substitutable customer base. Silicon capacitors, by contrast, must be designed into a package from the outset. They require customer qualification, integration into package architecture, and validation of power integrity. Once designed in, they are difficult to replace — the switching cost is high.

Samsung Electro-Mechanics’ contract counterparty is officially disclosed only as a “large global corporation.” It would therefore be incorrect to identify the customer as NVIDIA, AMD, Broadcom, Google, Meta, Microsoft, Amazon, or any other specific company. That remains unconfirmed.

5. What Silicon Capacitors Actually Signify

5.1 Technical Significance — The Power Bottleneck Migrates Inside the Package

As AI GPUs and HBM scale in compute intensity and data movement, instantaneous current demand has grown substantially. The problem is that power arriving from a distance is slowed by parasitic elements in the power traces, substrate, and package — degrading the transient response. The solution is to move capacitors closer to the die.

Conventional MLCCs sit primarily on the PCB or around the package perimeter. Silicon capacitors can go inside the package or immediately adjacent to the die. With capacitors of equivalent capacitance, placement determines performance.

5.2 Business Significance — Passive Components Re-rated as AI Packaging Components

MLCCs have traditionally been priced as cyclical components. When smartphone and PC demand weakens, inventory adjustments follow, and commodity products face price competition. Silicon capacitors, in contrast, are high-complexity components embedded inside AI GPU and HBM packages. They require customer qualification, integration into package design, and power-integrity validation.

For Samsung Electro-Mechanics, this contract carries strategic significance beyond the revenue figure. The company officially stated that it leveraged its accumulated precision process capabilities from the MLCC and package substrate businesses to enter the AI semiconductor core supply chain.³

5.3 Investment Significance — Possible Re-classification as “AI Server Component” Plays

The market has historically viewed Samsung Electro-Mechanics as an MLCC, camera module, and substrate company. If silicon capacitor revenue scales into mass-production levels, that classification partially changes.

Prior Market Perception	Possible Post-Silicon-Capacitor Perception
Smartphone component play	AI server component play
MLCC cycle stock	High-value power integrity component stock
Automotive MLCC growth story	AI GPU / HBM packaging supply chain entrant
Exposure to commodity passives	Exposure to high-complexity in-package components

Overclaiming risks remain, however. Silicon capacitors will not replace the entire MLCC market. Growth will come first in AI package interiors, high-performance computing, high-performance mobile chips, and select telecom and industrial applications where extreme performance is required. The vast majority of power stabilization demand in mainstream smartphones, automobiles, and home appliances will likely remain the domain of MLCCs for the foreseeable future.

6. This Week’s Tape — MLCCs Outran AI Substrates

This week (May 18–22, 2026), the market aggressively re-rated the MLCC theme. The three core MLCC names averaged +35.6%; the five core AI substrate names averaged +14.0%.

6.1 Core MLCC Three-Name Basket

Name	Weekly Return	20D	60D	2026E PER	Foreign	Institution	Foreign + Inst.	Program
Samwha Capacitor	+56.4%	+56.0%	+68.0%	40.8x	-₩5.88B	+₩16.09B	+₩10.21B	N/A
Samsung Electro-Mechanics	+32.7%	+65.0%	+200.5%	73.2x	-₩292.90B	+₩111.79B	-₩181.11B	+₩170.96B
Amotech	+17.6%	-18.4%	+55.2%	18.9x	+₩4.91B	-₩0.87B	+₩4.05B	+₩5.10B

Samsung Electro-Mechanics posted three consecutive sessions of strength: +7.5% on May 20, +13.5% on May 21, and +11.3% on May 22. Samwha Capacitor was more extreme: +23.0% on May 20, +29.9% on May 22. This is less a pure earnings-driven move than a re-rating of AI power-integrity components that originated with Samsung Electro-Mechanics and radiated outward into the MLCC ecosystem.

6.2 Core AI Substrate Five-Name Basket

Name	Weekly Return	20D	60D	2026E PER	Foreign	Institution	Foreign + Inst.	Program
심텍 (심텍)	+31.4%	+52.2%	+146.8%	38.7x	-₩24.58B	+₩61.81B	+₩37.22B	-₩5.35B
TLB	+18.0%	+44.4%	+89.2%	25.3x	-₩7.80B	+₩15.30B	+₩7.50B	-₩3.98B
Daeduck Electronics	+11.4%	+49.5%	+132.8%	38.9x	-₩2.93B	+₩29.05B	+₩26.12B	+₩5.28B
Korea Circuit	+10.7%	+5.6%	+66.7%	25.5x	-₩2.71B	-₩0.58B	-₩3.29B	+₩0.44B
ISU Petasys	-1.5%	-21.9%	+12.7%	35.6x	-₩119.17B	-₩0.74B	-₩119.90B	-₩125.28B

The AI substrate basket was not a uniformly rising tide. Simtech, TLB, and Daeduck Electronics were strong, but ISU Petasys finished the week at -1.5% amid heavy foreign and program selling. Within AI substrates, the market this week appeared to favor SoCAMM, memory module, and FC-CSP exposure over high-layer-count network boards.

6.3 Price Leadership to MLCCs; Flow Quality Favors Select AI Substrate Names

Basket	Foreign	Institution	Foreign + Inst.	Retail	Program
Core MLCC (3 names)	-₩293.86B	+₩127.01B	-₩166.85B	+₩159.40B	+₩176.05B
Core AI Substrate (5 names)	-₩157.19B	+₩104.84B	-₩52.35B	+₩52.05B	-₩128.89B
AI Substrate ex-ISU Petasys	-₩38.02B	+₩105.57B	+₩67.55B	-₩65.47B	-₩3.60B

On price, MLCCs dominated. On flow quality, the AI substrate basket excluding ISU Petasys looks cleaner. Simtech, TLB, and Daeduck show clear institutional accumulation with retail sellers or neutral retail — a more constructive configuration. In MLCC names, the ₩292.9 billion of foreign selling in Samsung Electro-Mechanics alone is too large to ignore. The stock price was strong, but this looks more like institutional, program, and retail-driven re-rating than foreign-led accumulation.

7. Investment Assessment

Name	View
MLCC	Strongest price momentum. Overheating risk and foreign selling overhang present.
AI Substrates	Lower absolute return, but institutional flow and earnings linkage are more stable.
Samsung Electro-Mechanics	Cross-category leader across MLCC, FC-BGA, and silicon capacitors. Holding is defensible; chasing is not.
Samwha Capacitor	Highest theme expansion intensity. +56% in one week makes new entry extremely difficult.
Daeduck Electronics	Core AI substrate holding. Valuation discount relative to Samsung Electro-Mechanics is meaningful.
Simtech / TLB	Led this week’s AI substrate expansion. Next week, monitor for a post-heat consolidation before re-entry.
ISU Petasys	Relative laggard within AI substrates. Lower priority until foreign flow stabilizes.

This week’s market was one in which attention expanded from AI substrates into MLCCs. The AI substrate thesis has not been invalidated. Rather, the market is now disaggregating “the next bottleneck after memory” more finely — extending the re-rating into AI server power integrity components.

In practical terms:

Hold: Samsung Electro-Mechanics, Daeduck Electronics
Do not chase: Samsung Electro-Mechanics, Samwha Capacitor
Monitor on pullback: Simtech, TLB
Await flow recovery: ISU Petasys
Key confirmation needed: Whether MLCC price increases and AI-server revenue lift 2Q–3Q earnings revisions

8. One-Line Takeaway

An MLCC is the electronic industry’s “ultra-compact ceramic water tank.” A silicon capacitor is an “ultra-proximity power stabilizer” embedded inside AI GPU and HBM packages. Silicon capacitors do not replace MLCCs. The battlefield for passive components is expanding from the PCB surface into the interior of AI semiconductor packages.

The real meaning of Samsung Electro-Mechanics’ ₩1.5 trillion contract is not “selling one more type of capacitor.” It is the signal that a passive component maker has entered the bottleneck component supply chain inside AI semiconductor packages. This shift can re-classify premium product lines at Samsung Electro-Mechanics and Murata into AI infrastructure value chain assets.

The market has, however, already responded quickly. The three core MLCC names averaged +35.6% this week; Samsung Electro-Mechanics is up +200.5% over 60 days and trades at 73.2x 2026E PER. The company is excellent and the industry shift is real. Whether today’s price is also a good new entry point is a separate question entirely. The next catalyst to monitor is not price action but 2Q–3Q earnings revisions, disaggregated AI-server revenue disclosure, follow-on silicon capacitor orders, and foreign flow recovery.

This post is for research and commentary purposes only and does not constitute investment advice. Technical descriptions of capacitors, MLCCs, and silicon capacitors are based on official materials from Samsung Electro-Mechanics, TDK, Murata, and other publicly available technical sources. The Samsung Electro-Mechanics silicon capacitor supply contract (approximately ₩1.5 trillion, January 1, 2027 through December 31, 2028) reflects the company’s official announcement. The customer has not been publicly disclosed; no specific GPU, ASIC, or cloud company has been identified as the counterparty. Price, flow, and valuation data for the May 18–22, 2026 period are sourced from the user’s Research OS local database and may differ from official exchange or brokerage data. 2026E PER, foreign/institutional/program flow figures, and basket averages for the core MLCC three-name and core AI substrate five-name groups reflect the analyst’s classification. Analysis may be wrong. Data reference date: May 22, 2026, KST.

Disclaimer: For research and information purposes only. Not investment advice. Names cited are for analytical illustration; readers should perform their own due diligence and consult licensed advisors before any investment decision.