Loading…
Loading…
Samsung Electronics is the only company in the world simultaneously operating five revenue franchises each individually material to global industrial policy: (1) memory (DRAM #1 globally with ~42% share, NAND #1 globally with ~33% share; HBM trailing SK Hynix on HBM3E qualification but pulled level on the HBM4 NVIDIA Rubin race); (2) foundry (#2 globally after TSMC, ~12% market share, persistently loss-making at the Samsung Foundry segment level through FY24-FY25, 3nm GAA process running below internal yield targets); (3) System LSI (mobile application processors — Exynos 2400/2500 lines, image sensors — the Galaxy + Apple iPhone CIS share split); (4) Samsung Display (wholly-owned subsidiary, ~85% global flexible-OLED share, primary OLED panel supplier to Apple's premium iPhone line, expanding into automotive cockpit and tablet displays); (5) Device eXperience — Galaxy mobile (#1 global smartphone shipments through FY25), Samsung TVs / appliances, and the Harman International US-incorporated automotive-electronics + audio-brands subsidiary (JBL, Harman Kardon, AKG, Bang & Olufsen licensing, plus automotive-OEM infotainment + telematics — acquired 2017 for ~USD 8bn).
Group exposure score is 74 including 5 tracked subsidiary — standalone is 71.
Named counterparties identified by open-source research and written only where a primary source states the relationship — never inferred from sector or co-mention. 2 suppliers, 2 customer-side. This is a partial view: it covers the relationships we could evidence, not the company's full supply base.
The Elec (2026-04-08): '린데와 에어프로덕츠는 이란의 카타르 라스라판 공습 이후 삼성전자와 SK하이닉스에 헬륨 공급 부족에 전면 대응하기로 하고 별도 장기 계약을 맺었다' (Linde and Air Products signed separate long-term contracts with Samsung Electronics and SK hynix); '두 업체 모두 미국 현지에서 원료 헬륨을 가져오는 비중이 높아' and a presidential-office official said the secured volumes are US-origin ('이번에 확보한 물량은 미국산'). Origin US helium sources; destination the buyer's Korean fabs. Trade press only; the article names both suppliers and both buyers jointly without an itemised supplier-buyer matrix. Air Products sources helium in the US.
The Elec (2026-04-08): '린데와 에어프로덕츠는 이란의 카타르 라스라판 공습 이후 삼성전자와 SK하이닉스에 헬륨 공급 부족에 전면 대응하기로 하고 별도 장기 계약을 맺었다' (Linde and Air Products signed separate long-term contracts with Samsung Electronics and SK hynix); '두 업체 모두 미국 현지에서 원료 헬륨을 가져오는 비중이 높아' and a presidential-office official said the secured volumes are US-origin ('이번에 확보한 물량은 미국산'). Origin US helium sources; destination the buyer's Korean fabs. Trade press only; the article names both suppliers and both buyers jointly without an itemised supplier-buyer matrix. Linde FY2025 report: helium sourced from helium-rich gas streams in the United States.
UBI Research unit-shipment figures reported via OLED-Info (2025-08-29): Samsung Display shipped 124 million OLED panels to Apple in 2024 versus LG Display's 67.5 million and BOE's 43 million — a computed 52.9% unit share, not a company-disclosed percentage. Neither Samsung nor Apple names the other or discloses a per-customer percentage directly; Samsung's own 2025 Sustainability Report states only that the Americas (where Apple is based) is 45.6% of Display-segment revenue, an aggregate that is NOT a per-customer share and is not used as one here. (Trade-pairs refresh 2026-09-29: OLED display panels are a finished component, not trade in any raw material; no plant-to-plant route is stated, so no corridor is credited.)
DIFFERENT BASIS from the 2024 row above — not a directly comparable vintage pair, so no delta is claimed. The 2024 figure is a unit-count share computed from raw shipment totals; this 2025 figure is Samsung's residual (100% − LG Display 30.3% − BOE 16.4%) because UBI/Display Daily reported LG's and BOE's shares directly but stated Samsung's absolute 2025 unit volume was not publicly disclosed. The residual is arithmetic on two disclosed percentages, not an invented number, but it inherits any imprecision in those two figures and assumes no fourth supplier. Confirms Samsung remained the largest of the three suppliers through 2025. (Trade-pairs refresh 2026-09-29: OLED display panels are a finished component, not trade in any raw material; no plant-to-plant route is stated, so no corridor is credited.)
Scope. Absence of a counterparty here is not evidence that none exists — most commercial sourcing is confidential and never becomes public. Rows without a share figure mean no share was disclosed, not that the relationship is small.
6 other tracked dossiers name this company as a counterparty in its own sourced disclosure — this platform never wrote these rows on this company's own page, they are read here inverted from the dossier that disclosed them. This is a partial view: 997 additional named-counterparty rows across the corpus could not be matched to a tracked dossier by name at all (mostly non-Latin legal names), so the true count is higher than what renders here.
Advantest's 83rd Annual Securities Report (IFRS, FY2023 = fiscal year ended 2024-03-31), customer-concentration table: Samsung Electronics Co., Ltd. JPY 55,325m / 11.4% of net sales, mainly semiconductor-and-component-test-system and mechatronics segments. The filing's separate Note 6 'Information of Main Customers' cites a broader 'Samsung Group' figure (JPY 57,725m, >=10% threshold) for the same year -- a wider consolidated-group scope than this entity-level row, not a conflicting number.
Same report, FY2024 (fiscal year ended 2025-03-31) column: Samsung Electronics Co., Ltd. JPY 82,795m / 10.6% of net sales. Note 6's broader 'Samsung Group' figure for the same year is JPY 87,734m.
AMAT's FY2015 10-K names Samsung Electronics and TSMC as its two 10%+ customers (Samsung: 18/12/13% for FY2015/14/13; TSMC: 15/21/27%). Later 10-Ks keep naming them through FY2024 (see FY2023/FY2024 rows); only the FY2025 10-K (filed Dec 2025) drops the names ('Customer One'/'Customer Two', ~19%/15%). Corrected 2026-09-26: this note previously said FY2015 was the most recent filing that names them.
FY2024 10-K (filed 2024-12-13) note on customer concentration and Item 1 table: customers at >=10% of net revenue (FY2024/FY2023/FY2022: Samsung 12/15/12%, TSMC 11/19/20%; Intel 10% in FY2022 only). Base labelled 'net revenue' in this filing, 'net sales' in earlier ones.
FY2023 10-K (filed 2023-12-15) Item 1 table: customers at >=10% of net sales (FY2023/FY2022/FY2021: Samsung 15/12/20%, TSMC 19/20/15%). Prior-vintage row for the FY2024 comparison.
Qualcomm FY2025 Form 10-K: "The primary foundry suppliers for our various digital, analog/mixed-signal, RF and PM integrated circuits include Taiwan Semiconductor Manufacturing Company (TSMC), Samsung Electronics and Global Foundries." Named as a foundry supplier; no volume or share is disclosed and none is implied here. Re-fetched and re-quoted 2026-09-17, unchanged since 2026-06-24 — still the live FY2025 10-K (fiscal year ended 2025-09-28).
Scope. This section is inverted evidence: each linked dossier wrote the original row about itself, describing its own relationship to this company. It is never a claim this company itself disclosed. Absence here is not evidence of no relationship — most of the corpus has no named-counterparty block at all yet.
Every exposure is read as a buyer dependency; the top 3 are its binding chokepoints. If this company actually produces one of these materials rather than buying it, that would be misclassified here — no producer-classifier entry exists for 2 of them.
Dossier discloses a neodymium exposure with no Dy/Tb/Sm tag and a sector where doping is not the rule — whether this company's magnets are Dy/Tb-doped is not on record. Resolved by a disclosed magnet grade, or a named magnet supplier in named_counterparties.
A61's suspension (Announcement 70) lapses 10 Nov 2026 — on the plain text, controls revive automatically. The US DoD CN/RU/IR/KP sourcing bar follows 1 Jan 2027. See the underlying extraterritorial export-control action. We supply the provenance file; we never compute your ratio for you.
As a buyer, policy pressure across its consumed materials is rising.
Descriptive trend in official policy actions on this company’s materials — a policy-pressure trend, not a price or trading signal. No forward probability is implied. Direction is measured on actions we have discovered, and discovery lags events: where the recent window is too thin to support a calm reading, the row says coverage-limited instead of easing.
🇺🇸 US accounts for 100% of severity-weighted pressure.