ICTK (KOSDAQ 456010) _ There is a product and the market is large — so why is this company's share so small?
There is a product and the market is large — so why is this company's share so small? ICTK's products, market, competitors and revenue
Fact disclosures, contracts, papers, observed figures Insight·Estimate facts joined into a reading Company stated what the company itself said Unverified not confirmed in disclosures
［Notice］ expand
- Report base date: 2026-09-11 (disclosures and results after this date are not reflected)
- Holding: none — the author does not hold the security discussed in this piece, as of the base date. The author will not trade the security during the five trading days before and after publication; after that period, the author may buy or sell without prior notice.
- This piece is a private analysis based on publicly disclosed materials. It does not recommend the purchase or sale of any security and is not investment advice. The author is not a licensed financial investment business operator or investment adviser, and does not provide individual consultations.
- All data and figures are drawn from sources such as the electronic disclosure system that are believed to be reliable, but their accuracy and completeness cannot be guaranteed. Estimates reflect the author's assumptions and may differ from actual outcomes; all figures are as of the base date stated in the text. Won amounts carry a US dollar figure converted at a fixed rate of KRW 1,400 = US$1, for the reader's convenience only — it is not a closing, reporting-date or audited rate, and because the same rate is applied to every year, the shape of a trend is identical in won and in dollars. The BTQ contract, denominated in US dollars from the start, is shown at its original amount.
- The analysis was conducted by setting aside the market's short-term assessment and examining the company's core competitiveness and fundamental direction first.
- Investment decisions and their results belong to the investor, principal loss may occur, and this material may not be used as evidence in any determination of legal responsibility.
- The critical analysis in this report is a subjective opinion offered in the public interest, for investors' right to know and for the sharing of information; there is no intent to damage the reputation of, or interfere with the business of, any company.
Four questions, one line each
| Question | Answer |
|---|---|
| Product | Fact A security semiconductor that generates its own key inside the chip (VIA PUF). One other company has a PUF with the same property. Insight·Estimate The moat is not the technology but a seat — "finished chip + Korean certification" |
| Market | Fact Large and old (NXP alone has shipped more than 6bn units). This company's share is small (security-chip revenue KRW 1.07bn / US$0.76m in 2025; cumulative shipments of 15m units are Unverified). Whether a market for finished chips with post-quantum cryptography (PQC) built in exists at all is Unverified |
| Competitors | Fact Above (Synopsys and eMemory, which sell the same technology as design blocks) · beside (NXP and Infineon on the same shelf) · below (software and HSM). If one name each: eMemory for the technology, NXP for the customers, Synopsys for the business itself |
| Revenue | Fact Three years of decline (security chips KRW 2.76bn → 1.07bn (US$1.97m → 0.76m); utilization 91% → 52%). A trace of recovery in the export column in H1 2026 (KRW 0.04bn (US$0.03m) for all of 2025 → 0.22bn (US$0.16m) in one half-year). The only confirmed item is the BTQ service backlog of KRW 11.30bn (US$8.07m) |
SWOT (Strengths · Weaknesses · Opportunities · Threats)
| Type | Content |
|---|---|
| Strengths | ① Fact VIA PUF core technology and 150 registered patents — a passive-element method, so it reproduces the same key without a correction circuit even as temperature and voltage change ② Fact Finished-product supply carried out in house from design through certificate injection and final testing; the G3K holds KCMVP certification — a seat that combines "finished chip + Korean certification" in the domestic market ③ Fact Liquid assets of KRW 44.27bn (US$31.62m) at 2026-06-30 (cash 11.61 + short-term financial instruments 25.63 + short-term financial assets 7.03, in KRW bn) |
| Weaknesses | ① Fact Three customers each above 10% of revenue accounted for 79.5% of half-year revenue; the largest alone for 48.7% ② Fact R&D spending equal to 140.5% of revenue (KRW 3.27bn / US$2.34m in H1 2026) — up from 92.0% in 2024 and 118.0% in 2025 ③ Fact Security-chip revenue fell for three years, from KRW 2.76bn (US$1.97m) in 2023 to 1.07bn (US$0.76m) in 2025, and utilization from 91.1% to 52.5%. The IPO forecast of KRW 19.05bn (US$13.61m) revenue for 2025 compares with an actual KRW 4.02bn (US$2.87m) |
| Opportunities | ① Fact A regulatory calendar that demands hardware — Korea's post-quantum transition (preparation complete by 2030, transition complete by 2035), the EU Cyber Resilience Act (CRA) fully applicable from 2027-12-11, and a KCMVP requirement for Korean AMI (advanced metering infrastructure) operators ② Fact Ministry of Trade, Industry and Energy national project, 2026-04 to 2029-12, KRW 7.0bn (US$5.0m), with KT as the demand-side partner ③ Fact BTQ (BTQ Technologies, a Canadian quantum-security company) contract backlog of KRW 11.30bn (US$8.07m), contract ending 2029-11-30 — with design completion announced on 2026-07-10, this may be the step before volume supply |
| Threats | ① Fact Entry of global IP (Intellectual Property — design assets) majors into the PUF field, such as Synopsys's acquisition of Intrinsic ID (2024-03-20). NXP has sold chips with PUF since 2013, and Taiwan's eMemory published in 2018 that its NeoPUF needs no correction either — a "correction-free PUF" is not unique to this company ② Insight·Estimate The largest half-year customer (disclosed only as "Company A") reads as BTQ, which is also a shareholder (452,058 shares) — a structure with no down payment or advance ③ Fact Convertible bonds outstanding of KRW 13.0bn (US$9.29m) — subject to early redemption requests from 2027-11-23. Grace periods for KOSDAQ administrative-issue designation also expire in turn (continuing-operations loss 2026-12-31 · revenue below KRW 3.0bn (US$2.14m) 2028-12-31) — though on 2025 figures the company meets neither condition today |
The number that changes the judgement — Q3 2026 security-chip revenue
Insight·Estimate The yardstick is a return to the 2023 level (security chips KRW 2.76bn (US$1.97m) a year → roughly KRW 0.7bn / US$0.5m a quarter). The IPO plan (KRW 7.7bn (US$5.50m) a quarter) is already too far from reality to serve as a yardstick.
| Q3 security chips | How to read it | Look at alongside |
|---|---|---|
| KRW 0.3–0.4bn (US$0.21–0.29m) a quarter (H1 pace maintained) | Noise, not a recovery. Read it toward "the share is not growing" | If utilization does not rise from 47.9%, same direction |
| KRW 0.5–0.7bn (US$0.36–0.50m) a quarter | On the way back to the 2023 level. Check whether the export column continues | Do security-chip exports continue quarter after quarter |
| KRW 0.7bn (US$0.50m) or more a quarter + exports at least half | Read as the big-tech door having opened. A revenue structure different from 2023's (dependence on domestic carriers) | Utilization recovery + a single-sales/supply-contract disclosure |
Insight·Estimate "It can deliver" is still a hope, and "it cannot" ignores the trace — the honest answer is that neither can be said yet. In the same report, utilization is the signal that comes first, and the security-chip export column is the cell to look at before any other.
Main text — tap a chapter to expand
1. What is the product
A security semiconductor that generates its own key inside the chip — the technology is real but not unique, and the moat is a seat: "finished chip + Korean certification".
1-1. In one phrase Fact
Fact A security semiconductor that makes its own key inside the chip. An ordinary security chip stores its key (the encryption key) in memory, and a stored key can be extracted. This company's chip uses the tiny random variations that arise during chip manufacture (differences in the size of vias — the small holes that connect one chip layer to the next) as the key. Nothing is stored, so there is nothing to extract. This approach is called PUF (physical anti-cloning), and the company's version is VIA PUF. (Semi-annual report II-2)
1-2. Four layers of what it sells Fact
| Layer | What | Who buys it | 2025 revenue |
|---|---|---|---|
| ① Design blocks (IP) | VIA PUF IP; post-quantum cryptography (PQC, Post-Quantum Cryptography — encryption built to resist attack by quantum computers) IP | Other chip-design companies | Negligible (an IP supply-service order of KRW 0.29bn / US$0.21m) |
| ② Security chips | G3 (mass-produced since 2018), G3K (the NIS-certified KCMVP version) — the mainstay | Mobile carriers (USIM, Universal Subscriber Identity Module — the subscriber identity card in a handset), IoT (Internet of Things) device makers | KRW 1.07bn (US$0.76m) |
| ③ Security modules and devices | Modules carrying the chip; eSIM/USIM communication modules | Device makers | KRW 0.23bn (US$0.16m) |
| ④ Development services and solutions | Custom security-chip design for a customer; software | BTQ (Canada), carriers, research institutes | KRW 2.72bn (US$1.94m) |
Source — semi-annual report (20260814003391) II-2 · II-4; [Amended] annual report (20260723000522).
Reading the position of the services correctly Insight·Estimate Two thirds of the money is services, but the largest of them, the BTQ contract, is "designing a security chip for BTQ's exclusive use". Once the design is finished (completion announced 2026-07-10), the next step is to mass-produce and supply that chip. So this service may not be a sideline outside the core business but the step before product revenue. Unverified There is, however, no volume-supply contract yet.
1-3. How it is made Fact
Fact The company is fabless (it has no fab, i.e. no factory). Design is done in house; wafer production, packaging and testing are outsourced; only the final certificate injection and final inspection are done on the company's own equipment. The "utilization rate" the company discloses is the utilization of this inspection equipment. (Semi-annual report II-3) Insight·Estimate Because the equipment runs just before a chip ships, it is a gauge that moves ahead of revenue.
1-4. Is the technology real — yes, but not unique
| Company claim | Check | Verdict |
|---|---|---|
| Same key without a correction circuit, even as temperature and voltage change | The company's filing (semi-annual report II-7) and the joint press release with BTQ (2026-07-10) agree | Fact |
| Competitors' PUFs use active elements and need correction | Taiwan's eMemory also presented its NeoPUF at ISSCC (International Solid-State Circuits Conference) 2018 with zero errors from −40°C to 150°C and no correction needed | Insight·Estimate Weak grounds — there is one more company with the same property |
| Listed in the global standard ISO/IEC 20897 | True (2020-12). Synopsys and PUFsecurity have passed the same standard's tests | Insight·Estimate A threshold, not a moat |
| International certification | Domestic KCMVP level 2 (2022). No international CC or PSA (Platform Security Architecture, Arm's IoT security certification) — a 2025 expense item reads "additional work for CC certification", so it is in preparation | Fact Domestic certification only · Unverified a competitor's (PUFsecurity's) CC/PSA certification could not be confirmed from a certification document |
Insight·Estimate In sum — not a technology moat but a seat moat: "finished chip + Korean certification". Another company holds a PUF with the same property and several have passed the standard, so what sets this company apart is not the technology itself but the seat it has taken by turning that technology into a finished chip and obtaining domestic certification for it.
2. Does the market exist
The market is large. This company's share is small. Whether a market for finished PQC chips exists at all is unconfirmed, and this company has no transaction in it.
2-1. Cut the market into three rings and ask only "is there real trade"
Insight·Estimate Buyers do not buy a "PUF chip"; they buy a secure element (SE) chip. PUF is one function of that chip. So the rings are cut by SE chips. (Cutting by PUF chips would make the market look smaller than it is.)
| Ring | Market | Real trade | Verdict |
|---|---|---|---|
| Outer | All security chips (cards, eSE, secure MCU) | Fact NXP alone has shipped more than 6bn units. Intrinsic ID's PUF is in 650m devices | A large, old market |
| Middle | Buyers of a separate finished SE chip (cards, payment terminals, meters, cars, IoT retrofit) | Fact The market is large (held by incumbent majors). This company's 2025 security-chip revenue is KRW 1.07bn (US$0.76m). Unverified Cumulative shipments of 15m units could not be confirmed from a document | The market is large; this company's share is small |
| Inner | Finished SE chips with PQC built in | Unverified Whether incumbent majors ship PQC-embedded SEs. Fact This company has one BTQ development-service contract; product sales and public-procurement awards: zero | Existence of the market unconfirmed; no transaction by this company |
eSE (embedded Secure Element) is a secure-element chip built into a device; MCU (Microcontroller Unit) is a small control chip. Shipment figures — NXP press release (SmartMX family, over 6bn units as of 2016), Intrinsic ID press release.
2-2. Why the market exists Fact
| Reason | Grounds |
|---|---|
| Where to keep the key is a problem software cannot solve | A stored key can be extracted. That a hardware root of trust is needed is the shared view of the standard (ISO/IEC 20897) and the US NIST (National Institute of Standards and Technology) zero-trust guidance (SP 800-207) |
| Regulation demands hardware | Korea: AMI (Advanced Metering Infrastructure) operators must use KCMVP-certified cryptographic modules (2012 ministry guideline); post-quantum transition prepared by 2030, complete by 2035 (Ministry of Science and ICT, 2025-09-03). Europe: the Cyber Resilience Act (CRA) applies in full from 2027-12-11, in effect requiring a hardware root of trust and an unclonable device identity, with no exemption for small firms |
| Connected devices multiply and authentication splits down to the device level | A seat for a chip that can be planted cheaply in volume |
2-3. Why is this company's share small when the market is large Insight·Estimate
Insight·Estimate The large volumes go two ways.
- Where volumes are large — smartphones, cars — the PUF is put inside the main chip as a design block (the customers of Synopsys and eMemory)
- Where the business is old — cards, passports, payments — incumbent security-chip majors hold it on certification and procurement track records
Insight·Estimate The seat this company can take lies between the two — when security is retrofitted to an existing device (meters, CCTV, industrial IoT), when domestic certification is a procurement condition, when volumes are too small for a design-block licence to make sense, when a supply chain must be split. The seat exists, but the past three years are the story of this company failing to widen it. The problem is not the market; it is the company's share.
2-4. The answer
Insight·Estimate The market is large. This company's share is small. Whether a market for finished PQC chips exists at all is unconfirmed, and this company has no transaction in it.
3. Who are the competitors
Above (Synopsys and eMemory, selling the technology as design blocks) · beside (NXP and Infineon on the same shelf) · below (non-chip approaches) — this company is the smaller party in all three directions.
3-1. Three directions Fact
| Direction | Counterpart | Why it is competition | Distance today |
|---|---|---|---|
| Above — sellers of the same technology as design blocks | Synopsys (acquired Intrinsic ID 2024-03-20) · eMemory/PUFsecurity (Taiwan) | Different customers, so no head-on collision. But if design tools and PUF are sold as a bundle, chip designers choose the bundle over a finished chip | Unverified Synopsys may be this company's tool supplier — supplier and competitor at once |
| Beside — finished chips on the same shelf | NXP · Infineon · Microchip · Samsung | Sold to the same customers (device makers) from the same shelf. NXP has sold chips with PUF since 2013 | Insight·Estimate Ahead in scale, price and certification history. Unverified A February 2025 interview in which the company's CEO reportedly placed these firms "on the same line" could not be traced to an outlet and date |
| Below — approaches that are not a chip | Software key management · HSM (Hardware Security Module — a separate security appliance kept in a server room) | Lowest cost. But the key is stored | Pushed back the more regulation demands hardware |
3-2. If you had to pick one Insight·Estimate
| To see | This company |
|---|---|
| Whether the technology is real | eMemory — the direct counterpoint and the only listed company whose numbers are visible |
| The risk of losing customers | NXP · Infineon |
| The risk of the business being erased outright | Synopsys |
| A comparable | None in Korea — KCS, Wooriro and others are grouped by theme but do different things |
3-3. A caution Insight·Estimate
Insight·Estimate eMemory is a royalty company (almost no cost of goods); this company sells finished products (wafer and packaging costs). The margin ceilings differ, so the numbers must not be carried across as they are.
4. What about this company's revenue
Three years of decline is correct; there is a faint trace of recovery in the H1 2026 export column; and the only confirmed money is the BTQ service backlog of KRW 11.30bn (US$8.07m).
4-1. So far — three years down, a sign of recovery in H1 2026 Fact
| (KRW bn / US$m) | 2023 | 2024 | 2025 | H1 2026 | H1 annualized |
|---|---|---|---|---|---|
| Security chips | 2.76 US$1.97m | 1.88 US$1.34m | 1.07 US$0.76m | 0.74 US$0.53m | 1.48 US$1.06m |
| – domestic | 2.69 US$1.92m | 1.64 US$1.17m | 1.03 US$0.74m | 0.52 US$0.37m | 1.04 US$0.74m |
| – export | 0.07 US$0.05m | 0.24 US$0.17m | 0.04 US$0.03m | 0.22 US$0.16m | 0.44 US$0.31m |
| Modules / devices | 0.87 US$0.62m | 0.22 US$0.16m | 0.23 US$0.16m | 0.07 US$0.05m | 0.14 US$0.10m |
| Development services | 2.56 US$1.83m | 4.59 US$3.28m | 2.72 US$1.94m | 1.52 US$1.09m | 3.04 US$2.17m |
| Total | 6.19 US$4.42m | 6.68 US$4.77m | 4.02 US$2.87m | 2.33 US$1.66m | 4.66 US$3.33m |
| Utilization | 91.1% | 78.2% | 52.5% | 47.9% | — |
Source — semi-annual report (20260814003391) II-2 · II-3; [Amended] annual report (20260723000522) II-2 · II-3. At KRW 1,400 = US$1, KRW 1bn ≈ US$0.71m. The annualized column is H1 × 2, calculated by the author and not an observed figure Insight·Estimate.
How to read it
- Fact 2023 → 2025 is a decline, correctly read. Security chips KRW 2.76bn → 1.07bn (US$1.97m → 0.76m), utilization 91% → 52%. The company's stated reasons were "reduced orders from end customers" and "adjustment of an overseas customer's product-launch timing" (semi-annual report XI-4).
- Insight·Estimate The direction may have changed in H1 2026. Annualized, security chips come to KRW 1.48bn (US$1.06m), more than 2025. The export column in particular — overseas security chips were KRW 0.04bn (US$0.03m) for all of 2025, and KRW 0.22bn (US$0.16m) in one half-year. This could be the first trace of the big-tech volume supply announced in May. But a half-year cannot be compared directly with a full year (H1 product breakdowns for 2024 and 2025 were not disclosed), and utilization has not yet risen.
- Fact The distance from the plan submitted at the IPO (KRW 19.05bn / US$13.61m for 2025, KRW 30.95bn / US$22.11m for 2026) was already settled by the 2025 actual of KRW 4.02bn (US$2.87m). The 2026 plan of KRW 30.95bn (US$22.1m) is hard to reach from an H1 of KRW 2.33bn (US$1.66m).
4-2. Ahead — to whom, and when Fact
| Counterpart | What | When | How much | Certainty |
|---|---|---|---|---|
| BTQ (Canadian quantum security; also a shareholder) | Quantum-security chip design service → possible volume supply afterwards | to 2029-11-30 | US$10m (KRW 14.36bn at the contract-date rate · 14.63bn at the half-year-end rate), delivered KRW 3.33bn (US$2.38m), backlog KRW 11.30bn (US$8.07m) | Confirmed — timing of the five instalments not stated; BTQ's funding capacity needs checking |
| Global big-tech (undisclosed) | Finished security chips | Volume supply announced 2026-05 Company stated | Undisclosed — H1 security-chip exports of KRW 0.22bn (US$0.16m) may be the first trace Insight·Estimate | Announcement + a small trace — the absence of a disclosure is normal under confidentiality practice |
| KT (ministry national project) | PQC security SoC (System on Chip — many functions on one chip) development | 2026-04 to 2029-12 | Project budget KRW 7.0bn (US$5.0m) | Confirmed but not revenue (a subsidy). Actual chip revenue after 2030 |
| Domestic carriers | eSIM/USIM communication modules | Supply planned after development completes in H1 2026 | Backlog KRW 0.086bn (US$0.06m) | Plan |
| Public institutions | KCMVP-certified PQC chips | Deadlines 2030 · 2035 | None | Regulatory clock — procurement notices so far: zero |
Source — single-sales/supply-contract disclosure (20251028900086), semi-annual report II-4 backlog, national-project announcement 2026-06-09, company announcement 2026-05-18.
4-3. For revenue to stand up — one of two doors Insight·Estimate
| Door | Evidence it has opened | When we will know |
|---|---|---|
| Big-tech reorders | Security-chip exports continue quarter after quarter, and utilization rises | Third-quarter report, November 2026 |
| Public PQC procurement | A "hardware + KCMVP" condition appears in procurement notices | 2027–2029 |
Insight·Estimate If both stay shut when the end of 2029 arrives, the BTQ service and the national project end together. If either opens, revenue stands up. Whether they open is decided not by the company but by customers' device launches and the government's procurement.
4-4. The answer
Insight·Estimate The seat has been taken, it narrowed over three years, and there is a small trace of recovery in the H1 2026 export column. "It can deliver" is still a hope, and "it cannot" ignores the trace. The honest answer is that neither can be said yet.
5. SWOT and implications
For each of strengths, weaknesses, opportunities and threats: three lines of fact, then what they mean, side by side. Implications are written as conditionals.
S — Strengths
| # | Fact | Source |
|---|---|---|
| ① | Fact VIA PUF core technology and 150 registered patents — a passive-element method, so it reproduces the same key without a correction circuit as temperature and voltage change. The company's filing and the joint press release with BTQ agree | Semi-annual report II-2 · II-7 |
| ② | Fact Finished-product supply carried out in house from design through certificate injection and final testing. G3K holds KCMVP hardware level 2 certification; CC EAL 6 development-environment certification in July 2025. A seat that holds a finished chip and Korean certification together | Semi-annual report II-3, I-2 |
| ③ | Fact Liquid assets of KRW 44.27bn (US$31.62m) at 2026-06-30 — cash 11.61 + short-term financial instruments 25.63 + short-term financial assets 7.03 (KRW bn). Of this, ordinary deposits are KRW 11.12bn (US$7.94m); the rest is held in time deposits, bonds, private funds and the like | Consolidated statement of financial position, semi-annual report III-7-2 |
Insight·Estimate The three strengths differ in kind. ① and ② are things built over time, which a competitor cannot simply buy with money; ③ is something that shrinks as time passes. But as chapter 1 showed, ① is not a unique property (eMemory has also published a correction-free PUF), so the real support is not ① but the seat in ②. In one line, ③ is buying the time for ② to widen into revenue — and whether the seat widens within that time is the true test of the strengths.
W — Weaknesses
| # | Fact | Source |
|---|---|---|
| ① | Fact Three customers each above 10% of revenue accounted for 79.5% of H1 2026 revenue; the largest alone for 48.7% | Semi-annual note 25 |
| ② | Fact R&D spending at 140.5% of revenue (KRW 3.27bn / US$2.34m in H1 2026). The ratio kept rising: 92.0% in 2024 → 118.0% in 2025 → 140.5% in H1 2026 | Semi-annual report II-6 |
| ③ | Fact Security-chip revenue KRW 2.76bn in 2023 → 1.88bn in 2024 → 1.07bn in 2025 (US$1.97m → 1.34m → 0.76m); utilization 91.1% → 78.2% → 52.5%. The IPO forecast of KRW 19.05bn (US$13.61m) revenue for 2025 compares with an actual of 4.02bn (US$2.87m), a 78.9% shortfall; the 2026 forecast of 30.95bn (US$22.11m) compares with an H1 actual of 2.33bn (US$1.66m) | Periodic reports II-2 · II-3, semi-annual report XI-4 |
Insight·Estimate These are not three separate weaknesses; they share a single root — product (security-chip) revenue is small relative to the size of the market and shrank further over three years. Customers are concentrated because revenue rests on one service contract; the R&D ratio is high not so much because the numerator is large as because the denominator (revenue) is small; and the forecast missed for the reason the company itself gave — "reduced orders, launch timing adjusted". It is the same conclusion as chapter 2 (not the market's fault, but the company's share). So these weaknesses are not to be fixed one at a time: if product revenue comes up, they ease together; if it does not, they stay together.
O — Opportunities
| # | Fact | Source |
|---|---|---|
| ① | Fact A regulatory calendar that demands hardware — Korea's post-quantum transition, preparation complete by 2030 and transition complete by 2035; KISA (Korea Internet & Security Agency) pilot transition project launched in 2026. EU Cyber Resilience Act (CRA) fully applicable from 2027-12-11. KCMVP-certified cryptographic modules mandatory for Korean AMI operators | Ministry of Science and ICT 2025-09-03; Boan News 2026-02-12; EU Regulation 2024/2847; ministry guideline (2012) |
| ② | Fact Lead organization of the Ministry of Trade, Industry and Energy national project for an integrated PQC·KpqC (Korean post-quantum Cryptography) security SoC — 2026-04 to 2029-12, KRW 7.0bn (US$5.0m), KT as demand-side partner | Announcement 2026-06-09 |
| ③ | Fact BTQ contract backlog KRW 11.30bn (US$8.07m) at 2026-06-30, contract ending 2029-11-30. Completion of the BTQ-exclusive chip design announced 2026-07-10 — volume supply is the next step | Semi-annual report II-4; 20251028900086; joint press release with BTQ 2026-07-10 |
Insight·Estimate What marks this company's opportunities is that all three carry a date — not "things will improve some day" but entries on a calendar. Their weight differs, though. ① is an opportunity opening to the whole market, and this company's share of it is not yet decided (procurement notices: zero); ② and ③ are opportunities whose share is fixed in amount and period. There is also a fourth, left out of the table — the small trace of KRW 0.22bn (US$0.16m) in H1 2026 security-chip exports. It is too small yet to call an opportunity, so it sits in chapter 4, but of the two doors in 4-3 it is the one that could open first. Laid over one another, the three dates cluster in 2027–2029.
T — Threats
| # | Fact | Source |
|---|---|---|
| ① | Fact Synopsys completed its acquisition of the PUF IP company Intrinsic ID on 2024-03-20 — PUF joined its design-tool and IP portfolio. NXP has sold "finished chips with PUF inside" since embedding Intrinsic ID's SRAM (Static Random Access Memory) PUF in SmartMX2 in 2013 and in LPC55Sxx in 2019. Taiwan's eMemory published in 2018 that its NeoPUF also reproduces the same key without correction (whether its subsidiary PUFsecurity holds CC/PSA certification is Unverified) | Synopsys press release 2024-03-20; NXP and Intrinsic ID press releases 2013-02-21 · 2019-03-04; eMemory ISSCC 2018 paper |
| ② | Insight·Estimate The largest half-year customer is disclosed only as "Company A", so its identity is not confirmed. From the transaction structure, this piece reads that customer as BTQ, which is at the same time a shareholder (third-party allotment of 452,058 shares, paid in 2025-12-19, with a voluntary two-year no-sale agreement). Fact The development-service contract has no down payment or advance | Semi-annual note 25; 20251215000349; 20251028900086 |
| ③ | Fact Fifth-series convertible bonds outstanding KRW 13.0bn (US$9.29m) — if converted, the share count rises; if not, they become subject to early redemption requests every three months from 2027-11-23. Related valuation loss of KRW 11.72bn (US$8.37m) in H1 2026. Grace periods for KOSDAQ administrative-issue designation expire in turn — the continuing-operations-loss condition on 2026-12-31, the revenue-below-KRW-3.0bn (US$2.14m) condition on 2028-12-31 | Semi-annual note 17; semi-annual report XI-4 |
Insight·Estimate ① is a threat from outside; ② and ③ are threats from inside (transaction structure and funding). Reading ① through the three directions of chapter 3 — above (Synopsys) is the risk of the seat being swallowed; beside (NXP and the eMemory family) is the risk that the very point of difference, a "correction-free PUF", wears thin. ② rests on an inference, since the disclosure hides the customer's name, but the double edge of the structure is clear — a customer who is also a shareholder makes the relationship firm, yet if one side wobbles, revenue and the shareholder register wobble together. ③ is the easiest threat to watch: 2026-12-31 · 2027-11-23 · 2028-12-31, all three dates are fixed. On 2025 figures, however, the pre-tax loss was about 15% of equity, so there is still distance to the continuing-operations-loss condition.
Overall implications — as conditionals
- Insight·Estimate If the third-quarter report in November 2026 shows quarterly security-chip revenue above KRW 0.7bn (US$0.5m) with exports at least half, the seat in strength ② changes character from "dependence on domestic carriers" to "overseas finished-chip supply", and the three-year decline in weakness ③ reads as over. If instead the figure stays at KRW 0.3–0.4bn (US$0.21–0.29m) a quarter, the 2026-05-18 volume-supply announcement has not yet turned into volume.
- Insight·Estimate If the BTQ backlog of KRW 11.30bn (US$8.07m) is recognized evenly quarter by quarter and a volume-supply contract follows in a disclosure, opportunity ③ is confirmed as the step before product revenue rather than a service, and the customer concentration in weakness ① also takes on the character of a "predictable base". If instead recognition slips later, the double edge of threat ② shows first on the revenue side.
- Insight·Estimate If a "hardware + KCMVP" condition appears in public procurement notices, opportunity ① turns from an opportunity for the whole market into one in which this company has a share. If instead regulation stops at "apply PQC algorithms" and does not require hardware, the direction below in chapter 3 (software) becomes the cheaper answer and threat ① widens.
- Insight·Estimate A structure in which R&D exceeds revenue (weakness ②) can be sustained only while the KRW 44.27bn (US$31.62m) of strength ③ holds it up. Of the three dates in threat ③, the first on the calendar is 2026-12-31, and the first with cash actually at stake is 2027-11-23. If neither of the two doors in 4-3 opens before then, the pace at which strength ③ — which has been buying time — shrinks becomes the axis of judgement.
- Fact The focus strategy the company itself has stated in disclosures and press releases is ① an integrated security SoC combining PQC on top of PUF, ② putting international standards (ML-KEM · ML-DSA — the post-quantum algorithms selected by the US NIST) and Korean standards (SMAUG-T · HAETAE — algorithms selected in Korea's KpqC competition) on one chip and pursuing KCMVP certification for it, and ③ using the big-tech supply record as a reference to expand to customers in North America, Europe and Asia.
6. The number that changes the judgement
One cell in the Q3 2026 report — security-chip revenue — plus the utilization that precedes it and the export column: these three are the next answer.
Insight·Estimate Yardstick = a return to the 2023 level (security chips KRW 2.76bn (US$1.97m) a year → roughly KRW 0.7bn / US$0.5m a quarter). The IPO plan (KRW 7.7bn / US$5.5m a quarter) is already too far from reality to serve as a yardstick.
| Q3 security chips | How to read it | Look at alongside |
|---|---|---|
| KRW 0.3–0.4bn (US$0.21–0.29m) a quarter (H1 pace maintained) | Noise, not a recovery. Read it toward "the share is not growing" | If utilization does not rise from 47.9%, same direction |
| KRW 0.5–0.7bn (US$0.36–0.50m) a quarter | On the way back to the 2023 level. Check whether the export column continues | Do security-chip exports continue quarter after quarter |
| KRW 0.7bn (US$0.50m) or more a quarter + exports at least half | Read as the big-tech door having opened. A revenue structure different from 2023's (dependence on domestic carriers) | Utilization recovery + a single-sales/supply-contract disclosure |
- Insight·Estimate Why KRW 0.7bn (US$0.50m) a quarter — it is the 2023 annual figure of KRW 2.76bn (US$1.97m) divided by four. Above that line, one can say the three-year decline is over; below it, the annualized H1 figure of KRW 1.48bn (US$1.06m) was a wobble, not a recovery.
- Insight·Estimate Why the export share — of 2023's KRW 2.76bn (US$1.97m), KRW 2.69bn (US$1.92m) was domestic. The same size with half from exports means "a new customer has appeared", not "domestic carriers bought again" — a different character.
- Insight·Estimate Why utilization first — as 1-3 showed, the certificate-injection and testing equipment runs just before a chip ships. If utilization in the same report rises from 47.9%, the security-chip cell in the following quarter is likely to rise; if it stays flat, the revenue cell is likely to stay flat too.
Insight·Estimate Whether utilization rises from 47.9% in the same report is the signal that comes first, and the security-chip export column is the cell to look at before any other.
Appendix — unverified items and sources
Sources for the figures in the text (filing numbers and dates), and items that could not be confirmed from disclosures
Items not confirmed from disclosures or public materials
| # | Item | Why it could not be confirmed |
|---|---|---|
| 1 | Volume-supply contract for the BTQ-exclusive chip | No supply-contract disclosure since the 2026-07-10 design-completion announcement |
| 2 | Timing of the remaining KRW 11.30bn (US$8.07m) of BTQ payments, and BTQ's funding capacity | Only the five-instalment condition is disclosed; the timing is not stated |
| 3 | Whether incumbent security-chip majors ship finished SEs with PQC built in | No confirmed shipment disclosure or report — the very existence of the inner-ring market is unconfirmed |
| 4 | Whether Synopsys is this company's design-tool (EDA) supplier | No supplier names in filings — only "outsourced SoC design development", software intangible-asset purchases and fee expenses as clues |
| 5 | Name of the global big-tech customer, supply volume and contract value | Withheld by the company under a confidentiality agreement; no single-sales/supply-contract disclosure |
| 6 | H1 revenue by product for 2024 and 2025 | The semi-annual report gives only the current half-year and full years — half-years cannot be compared directly |
| 7 | Unit prices and shipment quantities by security-chip product | Omitted on the grounds that costs could be inferred (semi-annual report II-2 · II-3) |
| 8 | Identities of the three customers above 10% of revenue (A, B, C) | Anonymized in the notes — "largest customer = BTQ" in the text is an inference, not a confirmed fact |
| 9 | PUF-segment revenue and share of competitors (Synopsys, eMemory etc.) | Not separately disclosed and outside this piece's scope of verification |
| 10 | Cumulative security-chip shipments of 15m units | Appears to be a company figure, but no document (disclosure or dated press release) could be identified |
| 11 | A February 2025 interview in which the CEO placed NXP, Infineon and others "on the same line" | Outlet and date could not be identified |
| 12 | Whether PUFsecurity holds CC/PSA certification | No certification announcement could be identified |
Sources
| Type | Document | Filing no. · date |
|---|---|---|
| Periodic | Semi-annual report (June 2026) | 20260814003391 · 2026-08-14 |
| Periodic | [Amended] Annual report (December 2025) | 20260723000522 · 2026-07-23 |
| Ad hoc | Single-sales/supply-contract disclosure (BTQ) | 20251028900086 · 2025-10-28 |
| Ad hoc | [Amended] Material-facts report (decision on paid-in capital increase) | 20251215000349 · 2025-12-15 |
| Ad hoc | Change of 30% or more in revenue or profit structure | 20260213901914 · 2026-02-13 |
| Company | Announcement of the start of volume supply of security chips to a global big-tech customer | Newspim · Daily Secu · 2026-05-18 |
| Company | Joint press release with BTQ on completion of the exclusive quantum-security chip design | 2026-07-10 |
| Government | Ministry of Science and ICT "quantum security innovation site meeting", press coverage | Boan News · 2025-09-03 |
| Government | Selection for the ministry's integrated PQC·KpqC security SoC national project | ZDNet Korea, Edaily and others · 2026-06-09 |
| Regulation | EU Cyber Resilience Act (CRA, Regulation (EU) 2024/2847) — fully applicable from 2027-12-11 | Official Journal of the EU, 2024-11-20 |
| Regulation | KCMVP-certified cryptographic modules required for AMI operators | Ministry guideline · 2012 |
| Standard | ISO/IEC 20897-1, security requirements for PUFs | 2020-12 |
| Competition | Synopsys completes acquisition of Intrinsic ID | Synopsys press release · 2024-03-20 |
| Competition | Intrinsic ID SRAM PUF in NXP SmartMX2 / integration in LPC55Sxx | NXP press release 2013-02-21 · Intrinsic ID press releases 2016-11-15 · 2019-03-04 |
| Competition | eMemory NeoPUF — stable without correction across temperature and voltage (zero errors from −40°C to 150°C) | ISSCC 2018 paper |
This piece contains only facts and estimates about the business, market and competition; it contains no valuation and no investment judgement. Company stated marks content the company itself stated in disclosures or press releases. Base date 2026-09-11 · no holding.
댓글
댓글 쓰기