US 11594578

I'll search for litigation and USPTO records specific to this patent number.

US 11,594,578 B2 — Analyst Summary

Scope note: I did not access USPTO Patent Center or PACER directly. The bibliographic and claim content below is taken from the authoritative full text supplied in your message (Google Patents rendering of the patent), and the litigation/PTAB information from live search results, cited per item. I flag uncertainties explicitly.

1. Bibliographic data (high confidence — from the patent text)

Abstract (verbatim): "A pixel arrangement structure of an OLED display is provided. The pixel arrangement structure includes: a first pixel having a center coinciding with a center of a virtual square; a second pixel separated from the first pixel and having a center at a first vertex of the virtual square; and a third pixel separated from the first pixel and the second pixel, and having a center at a second vertex neighboring the first vertex of the virtual square."

2. Technology in plain language

The patent covers a "diamond"-type sub-pixel layout for an OLED panel. A small, quadrilateral first pixel (green in the disclosed example) sits at the center of a virtual square; a pair of second pixels (blue) sits at one pair of diagonal vertices and a pair of third pixels (red) sits at the other pair, so the four larger pixels enclose the small one. The second/third pixels are octagonal (or hexagonal in alternative embodiments), and the gap between neighboring first pixels (L2/L5) is deliberately made longer than the gaps between the first pixel and its neighbors (L1/L4), which the specification says improves fine-metal-mask (FMM) deposition reliability while preserving aperture ratio. FIGS. 2–5 vary the geometry: parallelogram-shaped first pixels, symmetrical neighboring first pixels, second pixels larger than third (to compensate for blue's shorter lifetime), and octagonal first pixels with rhombus-shaped second/third pixels.

3. Independent claims — plain-language overview

The patent has seven independent claims: 1, 42, 45, 53, 67, 85, and 89. (All others depend from one of these.)

Claim 1 — the base "grid" claim. An OLED pixel array of first, second and third pixels of different colors, with a pixel-defining layer. First pixels form rows along a first direction; second and third pixels alternate to form parallel rows; a given second row runs through the centers of its second/third pixels and passes between first pixels of the adjacent first rows. First and second rows alternate. The same arrangement exists in the perpendicular direction (first-pixel columns alternating with columns of alternating second/third pixels). First pixels and either second or third pixels also alternate along a diagonal direction. Two negative/geometric limitations close the claim: (i) a band whose width equals the width of the first pixels, running parallel to the first direction and completely overlapping a row of first pixels, is entirely offset in the second direction from at least one second or third pixel in an adjacent row; and (ii) the shortest distance between two nearest first pixels within a row is greater than the shortest distance between the nearest second and third pixels within a row.

Claim 42 — same layout, defined by center-to-center pitches and relative areas. Requires second pixels to be larger in area than third pixels; the center-to-center distance between neighboring second/third pixels = a first length; first↔second and first↔third center-to-center distances = a second length; first↔first center-to-center distance = a third length greater than the second length; plus the same "first-pixel-in-row spacing > second/third nearest spacing" limitation as claim 1.

Claim 45 — "one large pixel per quadrangle" claim. Based on a virtual quadrangle whose outer edges are formed by edges of a group of four neighboring first pixels. Only one second or third pixel lies inside that quadrangle, with all other second/third pixels outside it. The shortest distance between any two of the four neighboring first pixels is greater than the shortest distance between the enclosed second/third pixel and the nearest second/third pixel measured perpendicular to any side of the quadrangle.

Claim 53 — display-level row/group claim. Alternating first pixel rows (first pixels only) and second pixel rows (second and third pixels alternating). First pixel groups (a first pixel + a neighboring second pixel) and second pixel groups (a first pixel + a neighboring third pixel) alternate along the row direction. Two spacing limitations: the shortest distance between the two nearest first pixels in the column direction is greater than the column-direction width of at least one second or third pixel; and the shortest distance between two nearest first pixels within a first pixel row is greater than the shortest distance between the nearest second and third pixels in a second pixel row.

Claim 67 — display with row-offset limitation. Three pixel types of three different colors organized into alternately-arranged first rows (first + second pixels alternating along the row direction) and second rows (first + third pixels alternating along the row direction). The first pixels in the second rows are entirely outside a region that runs in the row direction and completely overlaps the first pixels of an adjacent first row. Also requires the shortest distance between the two nearest first pixels in a diagonal direction to exceed the shortest distance between the nearest second and third pixels in that diagonal direction.

Claim 85 — display focused on the first-pixel shape. Same alternating first/second pixel row and pixel-group structure as claim 53, but the characterizing limitation is geometric: each first pixel is elongated along a diagonal direction, with two parallel sides and two connection portions joining the parallel sides at their ends, and at least one connection portion is longer than the shortest distance between the two parallel sides (i.e., the end caps are longer than the pixel's width).

Claim 89 — display combining the row structure of claim 67 with the first-pixel shape of claim 85. First rows contain first + second pixels alternating along the row direction; second rows contain first + third pixels alternating along the row direction; rows alternate along the column direction; and each first pixel is elongated in the row or column direction with two parallel sides joined by two connection portions, at least one of which is longer than the shortest distance between the parallel sides.

Common thread across all seven independents: (a) three differently-colored pixel types in an alternating diamond-type arrangement, (b) larger dark/insulating gaps between same-color (first) pixels than between different-color neighbors, and (c) in claims 85/89 (and dependent claims 40, 43, 51, 63, 81), a specific "elongated first pixel with long end-caps" profile.

4. Litigation and PTAB posture (from search results — with sources)

• ITC Inv. No. 337-TA-1351 — Certain Active Matrix Organic Light-Emitting Diode Display Panels and Modules for Mobile Devices and Components Thereof. The '578 patent was one of the asserted patents, along with the '803, '599 and '593 patents. Source: CourtListener docket for E.D. Va. 1:25-cv-00908, https://www.courtlistener.com/docket/70497017/5/samsung-display-co-ltd-v-boe-technology-group-co-ltd/ ; also listed on the patent record.
• E.D. Tex. 2:23-cv-00309 — Samsung Display v. BOE Technology et al. The court entered a voluntary dismissal without prejudice under Rule 41(a)(1)(A)(i), so no merits ruling issued. Source: PatSnap litigation write-up citing the PACER docket, https://www.patsnap.com/fr/resources/blog/litigation/samsung-display-v-boe-technology-oled-patent-dismissal-patsnap/
• IPR2023-01075 (Petitioners: BOE, CSOT/TCL, Tianma, Visionox group) — captioned in one source as "Patent 11,594,578 B2 … Determining No Challenged Claim Unpatentable, 35 U.S.C. § 318(a)." Chinese-language industry reporting states the PTAB concluded on Jan. 6, 2025 that US 11,594,578 could not be invalidated, i.e., all challenged claims survived. Sources: https://finance.sina.com.cn/cj/2025-01-09/doc-ineekkks9476811.shtml and https://www.eet-china.com/mp/a375214.html. Note: I did not see a primary PTAB document image in the results, so treat the exact FWD date/paper number as needing verification against PTAB E2E.
• Federal Circuit appeals — A USPTO filing states the three related IPRs (IPR2023-00987, -00988, -01075) are "on appeal to the Federal Circuit, Nos. 25-1517, 25-1518, 25-1519, 25-1520, 25-1521," and that the ITC investigation appeal is Samsung Display Co., Ltd. v. International Trade Commission, No. 25-1791 (Fed. Cir.). Source: https://ptacts.uspto.gov/ptacts/public-informations/petitions/[1558494](/patent/1558494)/download-documents (PTAB petition document). The patent record itself separately lists Fed. Cir. cases 25-1519 and 25-1791.
• N.D. Tex. 3:25-cv-01430 — Samsung Display sued TCL CSOT (and TCL, Ultimate Eshop, eTech Parts Plus) on approximately June 5, 2025, asserting US 11,594,578, US 7,414,599 and US 9,330,593. Sources: https://www.thelec.kr/news/articleView.html?idxno=36758 ; https://www.eet-china.com/mp/a410639.html. The accused products were smartphone OLED panels CSOT supplied to TCL, Ultimate Eshop and eTech Parts Plus.
• E.D. Va. 1:25-cv-00908 — Samsung Display v. BOE Technology Group, an action to enforce/review the ITC § 337 determination; the docket references the Commission's decision to review the final ID and the presence of the '803, '578 and '599 patents in the investigation (the '593 patent's no-violation finding was not reviewed).

CAFC 2026-specific request

I found no 2026 Federal Circuit decision or opinion on US 11,594,578 in the search results. The CAFC matters I could tie to this patent are the 2025-numbered appeals 25-1519 (IPR2023-01075) and 25-1791 (ITC appeal), which would be the candidates to be argued/decided in 2026 — but I cannot confirm a 2026 argument date, disposition, or Rule 36 judgment from what I retrieved. Stated uncertainty: high.

5. Uncertainties and discrepancies to flag

1. Search-result date anomaly. Several retrieved documents are dated July–September 2026 (e.g., a CAFC September 2026 oral-argument calendar and a Sept. 20, 2026 Chinese article about Samsung Display moving to invalidate three CSOT patents), which is after the stated current date of April 26, 2026. I have not relied on those for any conclusion about the '578 patent, but the inconsistency suggests some indexed pages carry later or auto-updated timestamps.
2. A third-party database mischaracterizes this patent. One result labels US 11,594,578 as "OLED pixel circuit driving and display architecture." That is inconsistent with the authoritative text, which is directed to a pixel arrangement structure (geometry/layout), not pixel-circuit driving. The pixel-circuit subject matter belongs to the sibling patent US 7,414,599. Treat that description as erroneous.
3. ITC claim scope. The E.D. Va. docket excerpt indicates the Commission reviewed the final ID and that the '593 patent dropped out on a no-violation finding; I did not retrieve the Commission's ultimate remedy determination for the '578 patent itself.
4. Independent-claim count. I derived the set {1, 42, 45, 53, 67, 85, 89} by inspecting the claim dependencies in the supplied text. It is consistent with the stated total of 92 claims, but I did not run a formal claim-tree tool.

I'll search for litigation specific to US 11,594,578 B2 across patent litigation portals and court records.

Let me dig into the specific case records and current statuses.

Let me check the ITC investigation, the E.D. Va. enforcement action, and the Federal Circuit appeal status.

Litigation involving US 11,594,578 B2

Bottom line: Yes — litigation is known, and it is substantial. US 11,594,578 ("the ’578 patent," Samsung Display's diamond/quadrangular pixel-arrangement patent, priority Mar. 6, 2012) has been asserted by Samsung Display Co., Ltd. against BOE and against TCL CSOT in overlapping federal-court, ITC, and PTAB proceedings. Note first a date/record caveat (below), because it affects every "current status" line.

1. Data-integrity caveats before the table

• Date discrepancy — flag for the record. This task states the current date as April 26, 2026. However, the source records I retrieved include documents dated after that: an N.D. Tex. scheduling order entered May 12, 2026, a docket order entered June 8/9, 2026, and Chinese industry articles dated September 20, 2026. Under the operating rule that live search results are the ground truth, I report those items as real and labeled by their document dates, but the contradiction with the task's "current date" is unresolved and is carried forward from the earlier-generated sections.
• Which proceeding formally asserts the ’578 patent is not uniformly documented. The ’578 patent is undisputed as an asserted patent in the N.D. Tex. CSOT action (explicitly pleaded) and was the subject of PTAB IPR2023-01075. For the ITC investigation and the E.D. Va. action I retrieved, the docket text I obtained recites claim numbers for the ’803, ’683, ’599 and ’593 patents but not the ’578; the ’578's inclusion in the ITC cluster rests on secondary reporting plus the fact that the four Chinese panel makers filed IPR2023-01075 against the ’578 in response. Treat ITC/E.D. Va. ’578-specific assertion as high-confidence but not docket-verified here.
• Uncorrected identifier from a third-party database. One aggregator labels US 11,594,578 as "OLED pixel circuit driving and display architecture." That is wrong (carried over from the earlier sections); the ’578 is a pixel arrangement patent. The pixel-circuit patent in this family group is US 7,414,599.

2. Known litigation / contested proceedings

3. Related-but-distinct proceedings (do not assert the ’578 patent — included to avoid miscounts)

• CSOT v. Samsung Display Co., Ltd.; Walmart; Best Buy — E.D. Tex., filed March 19, 2026. CSOT's counter-suit asserts CSOT's own patents — US 12,133,429; US 11,957,031; US 11,257,891 — against Samsung Display OLED panels (Google Pixel 8/9/10-series, tablets, wearables). The ’578 patent is not at issue here; it is the mirror-image response to item 4 above.
• Samsung Display IPR filings against CSOT, PTAB, Sept. 17, 2026 — directed to the same three CSOT patents (’429, ’031, ’891). Not ’578-related.
• BOE's Chinese invalidation/administrative litigation (e.g., the Supreme People's Court hearing Sept. 16, 2026, BOE v. CNIPA, Samsung as third party) concerns a Samsung invention patent whose number was not disclosed in the reporting — I cannot confirm it is the ’578.

4. Verified outcomes and statuses at a glance

1. ’578 survived all PTAB invalidity challenges. IPR2023-01075 FWD (Jan. 6, 2025) held no challenged claim unpatentable. This is the single most consequential ’578-specific result. (Sources: PTAB caption reproduced at https://finance.sina.com.cn/cj/2025-01-09/doc-ineekkks9476811.shtml ; https://www.eet-china.com/mp/a375214.html ; https://www.zwcmm.com/sys-nd/347.html )
2. The ’578 is in active district-court litigation — N.D. Tex. 3:25-cv-01430, presently set for jury trial Oct. 5, 2027. (Source: CourtListener docket, https://www.courtlistener.com/docket/70468030/samsung-display-co-ltd-v-wuhan-china-star-optoelectronics-semiconductor/ )
3. The BOE-facing ’578 disputes are over — the ’578-featuring E.D. Tex. case was dismissed without prejudice, and the ITC investigation was terminated Nov. 18, 2025 on settlement, with BOE licensing/royalty terms. (Sources: PatSnap summary of PACER 2:23-cv-00309 — note its patent-title error — https://www.patsnap.com/fr/resources/blog/litigation/samsung-display-v-boe-technology-oled-patent-dismissal-patsnap/ ; https://www.kedglobal.com/us/electronics/newsView/ked202511200007 )
4. The CSOT-facing ’578 litigation remains live, and CSOT's PTAB/CAFC challenges to the ’578 continue to be appealed (25-1519).

5. What I could not confirm (stated explicitly rather than guessed)

• No 2026 Federal Circuit decision, opinion, or Rule 36 judgment on US 11,594,578 was found. The candidate appeals (25-1519, 25-1791) remain pending as of my retrieval.
• The exact disposition date of the Rule 41(a)(1)(A)(i) dismissal in 2:23-cv-00309 is not established from the sources retrieved.
• Whether US 11,594,578 is formally named in the ITC complaint's claim chart (as opposed to being part of the same five-patent cluster) and in E.D. Va. 1:25-cv-00908 is unverified against a primary docket image.
• I did not access PACER, PTAB E2E, or the Unified Patents portal directly; the above is assembled from the authoritative patent text, CourtListener/Justia dockets, and the cited secondary sources.

If you want, I can attempt to pin down (a) the PTAB FWD paper number and date for IPR2023-01075, and (b) the current docket entries in Fed. Cir. 25-1519 and 25-1791, so these statuses can be upgraded from "reported" to "docket-verified."

I'll reconcile the structured "no PTAB proceedings" block against the IPR already surfaced in prior sections, then drill into the record.

Strong hits. Now let me pin down the challenged claims, grounds, institution date, and appeal.

Now let me get the IPR2023-01075 docket specifics and the Federal Circuit appeal.

PTAB / AIA Trial Analysis — US 11,594,578 B2 ("Pixel arrangement structure for OLED display," Samsung Display Co., Ltd.)

⚠️ Data-source conflict — flag first

The structured "PTAB proceedings on file" block in my prompt states: "The USPTO ODP API returns no AIA trial proceedings for this patent… the default should be 'no PTAB activity on file.'"

That block is stale/incorrect for this patent, and I am not adopting it. Live search surfaced a completed AIA trial on the '578 patent with a final written decision, a published judge panel, and four co-petitioners. Per the operating rule to prefer live results over conflicting inputs, and because the previously generated sections of this analysis already identified IPR2023-01075, I treat IPR2023-01075 as the canonical (and, from what I could retrieve, only) AIA trial on the '578 patent, and I flag the ODP ingest as needing a refresh. See the Unified Patents PTAB record mirrored on the patent page: https://portal.unifiedpatents.com/ptab/case/IPR2023-01075

Proceedings overview

Total AIA trials identified on US 11,594,578: one (1). Breakdown: 0 claims invalidated · 1 proceeding with claims sustained (all challenged claims) · 0 settled · 0 institution denials · 0 active trial-phase proceedings. IPR2023-01075 reached a Final Written Decision on 2025-01-06 deciding "No Challenged Claim Unpatentable," and that decision is now on appeal at the Federal Circuit (No. 25-1519).

Bottom-line defensive posture: this patent is hardened, not vulnerable. A defendant today does not get the "claim 1 is dead" argument. Samsung Display has already been through a four-company, jointly-funded validity attack on this exact patent, defended it, and won — every challenged claim survived. A defendant facing assertion must therefore attack on (a) non-infringement / claim construction (the independent claims are dense geometric limitations — see claim 1's "entirely offset" and shortest-distance limitations, and the "elongated first pixel with long connection portions" limitations in claims 40/51/63/81/85/89), (b) § 112, or (c) new prior art the petitioners did not raise — and should assume the obvious § 102/§ 103 art is already picked over.

IPR2023-01075 — Mianyang BOE Optoelectronics Technology Co., Ltd.; Wuhan China Star Optoelectronics Semiconductor Display Technology Co., Ltd.; [Tianma Microelectronics Co. Ltd.](/litigations/by-plaintiff/Tianma%20Microelectronics%20Co.%20Ltd.); and [Visionox Technology, Inc.](/litigations/by-plaintiff/Visionox%20Technology%2C%20Inc.) v. Samsung Display Co., Ltd.

• Type: Inter Partes Review (35 U.S.C. §§ 311–319)
• Patent: US 11,594,578 B2
• Filed: June 2023 (contemporaneously with the broader June 2023 invalidation campaign by four Chinese display makers; the exact filing day of the petition was not retrievable from public sources I could reach). ➜ Exact petition filing date: unverified.
• Status (verbatim from the record): "Determining No Challenged Claim Unpatentable — 35 U.S.C. § 318(a)." Plain English: the Board rejected every ground and left every challenged claim standing. // The structured ODP block reports "no AIA trial proceedings"; conflict flagged (see above).
• Judge panel: Jameson Lee, Terrence W. McMillin, and John A. Hudalla, Administrative Patent Judges; the judgment is authored by APJ McMillin. (Grounded in the reproduced FWD caption at https://finance.sina.com.cn/cj/2025-01-09/doc-ineekkks9476811.shtml and https://www.eet-china.com/mp/a375214.html.)
• Petition grounds: Not verified. I could not retrieve the petition, the institution decision, or the FWD body from a primary USPTO source in this session. Press characterization is generic ("patent invalidation trial," i.e., § 103-type obviousness grounds are the norm), and I will not invent statutory grounds, claim lists, or reference names. Notably, in the sibling IPR on US 9,330,593 (IPR2024-00620, same petitioner group), the grounds were § 103 obviousness over Kimura, over Shin+Sasaki, over Shin+Sasaki+Jeong, over Tobita, and over Tobita+Jeong, covering claims 1–8, 14–22, and 25 — that is context for the campaign's methodology, not a statement about the '578 grounds. (Institution decision: https://www.docketalarm.com/cases/PTAB/IPR2024-00620/Mianyang_BOE_Optoelectronics_Technology_Co._Ltd._v._Samsung_Display_Co._Ltd/docs/09-12-2024-Board/Institution_Decision__Grant-12-Institution_Decision__Grant.pdf)
• Institution decision: Instituted — date not verified. Circumstantial support: BOE's November 2024 investor-relations statement says all five of its IPR petitions against the Samsung Display patents in the ITC-1351 set "均已成功立案" (were all successfully instituted). The FWD date of 2025-01-06 plus the § 316(a)(11) one-year trial clock places institution in roughly December 2023 – January 2024 (an 18-month extension would place it ~July 2023). ➜ Exact institution date and reasoning: unverified.
• Final Written Decision: Issued 2025-01-06 (per USPTO information relayed 2025-01-08 and contemporaneous reporting). Verdict: no challenged claim held unpatentable — the caption itself is dispositive, and Chinese-language industry reporting states the Board concluded the '578 patent "不能被无效" (cannot be invalidated). Related reporting with more granularity states that as between the three patents Samsung Display later asserted against TCL CSOT, "US11594578、US7414599 的全部权利要求均被认定为有效" — for both the '578 and the '599 patents, all claims [challenged] were held valid; only the '593 patent was partially invalidated. (https://www.eet-china.com/mp/a526144.html, 2026-09-20; and https://www.eet-china.com/mp/a410639.html, 2025-06-09.)
  • ⚠️ Caveat on "all claims": the precise list of challenged claims was not retrievable, so I cannot state which of the 92 claims were adjudicated vs. left untouched. Treat "all claims" as press shorthand for "all challenged claims." ➜ Claim-level FWD granularity: unverified.
  • ⚠️ Cross-check against the earlier press: the January 2025 reporting said the Board upheld four of the five patents, of which "one circuit-and-driving patent and one diamond-pixel patent" were fully valid and "two other diamond-pixel patents" were partially valid. Since the three "diamond pixel" patents in the campaign were the '578, '803 and '683 patents, the internally consistent reconciliation is that the fully-valid diamond-pixel patent was the '578 (matching the 2026 report) and the two partially-valid diamond-pixel patents were the '803 and '683. This reconciliation is my inference, not a verified fact — flagging it explicitly.
• Settlement / termination: No settlement in the PTAB case; it was decided on the merits. (The later November 2025 Samsung Display–BOE global settlement resolved the district court and ITC matters and caused withdrawal of appeals, but I could not confirm from retrieved sources whether it mooted the CAFC appeal of IPR2023-01075 — see the appeal note below. Sets a possible mootness issue.)
• Appeal: Yes — appealed. The Fed. Cir. docket number listed on the patent record and by Unified Patents as litigation involving this patent is 25-1519 (Fed. Cir., filed 2025) (https://portal.unifiedpatents.com/litigation/Court%20of%20Appeals%20for%20the%20Federal%20Circuit/case/25-1519). A USPTO filing identifies the '578 IPR (IPR2023-01075) as one of IPR2023-00987, -00988 and -01075, which are "on appeal to the Federal Circuit, Nos. 25-1517, 25-1518, 25-1519, 25-1520, 25-1521," with the ITC investigation appeal captioned Samsung Display Co., Ltd. v. International Trade Commission, No. 25-1791 (Fed. Cir.) (https://ptacts.uspto.gov/ptacts/public-informations/petitions/[1558494](/patent/1558494)/download-documents).
  • Issues on appeal: the petitioners' § 103 obviousness case (verbatim grounds unverified). The appeal is petitioner-side — i.e., the losers below are the appellants.
  • Disposition: None confirmed. I found no Federal Circuit opinion, Rule 36 judgment, or argument calendar entry for 25-1519. My earlier sections already flag the high uncertainty here. Given the November 2025 SDC–BOE global settlement and the April 2026 procedural posture of the prompt, verify whether 25-1519 was dismissed as moot or voluntarily before relying on it either way.
• Defensive value (most important item in this report): Adverse to defendants. The patent emerged from a fully-contested, multi-petitioner IPR with every challenged claim intact. Any demand letter built on the '578 patent now carries the implicit "survived an IPR" pedigree, and Samsung Display has said so in its own pleadings (asserting the patents "已在美国国际贸易委员会的专利侵权调查中被认定为有效"). Your realistic invalidity paths are (i) prior art not raised in IPR2023-01075 (including system/printed-publication art and § 112 — neither of which the Board can invalidate on in an IPR), and (ii) district-court obviousness under the clear-and-convincing standard.

Strategic summary

Claim status on US 11,594,578. On the record I could retrieve, zero claims of the '578 patent have been canceled or held unpatentable in an AIA trial. All 92 claims stand — including the seven independents (1, 42, 45, 53, 67, 85, 89) mapped in the earlier summary — subject to one important evidentiary gap: I could not retrieve the challenged-claim list from the IPR2023-01075 petition or institution decision, so I cannot say whether all 92 claims were tested or only a subset. The FWD caption ("No Challenged Claim Unpatentable") means whatever was challenged survived; claims that were never challenged are UNTESTED at the PTAB and equally exposed to a fresh petition. Treat the "SURVIVED" bucket as "all challenged claims" and the "UNTESTED" bucket as unknown-but-plausibly-large.

Estoppel landscape. IPR2023-01075's petitioners were Mianyang BOE Optoelectronics, Wuhan China Star Optoelectronics Semiconductor Display Technology, Tianma Microelectronics, and Visionox Technology, and each identified itself and its subsidiaries as real parties-in-interest. Under § 315(e)(2), with the FWD having issued, those petitioners and their privies/RPIs are barred in district court from asserting any ground they raised or reasonably could have raised (i.e., § 102/§ 103 on patents and printed publications). That creates two very different positions:

• If you are TCL CSOT / Wuhan CSOT (the same entity named as a petitioner, and the accused infringer in Samsung Display v. Wuhan China Star et al., N.D. Tex. 3:25-cv-01430), you are squarely within the estoppel shadow. Your district-court invalidity case must rest on system art, public-use/on-sale art, indefiniteness/enablement/written-description (§ 112), and claim-construction-based non-infringement — not on the § 102/§ 103 printed-publication grounds your own subsidiary already litigated.
• If you are a defendant with no privity to the IPR2023-01075 petitioners (e.g., a distributor, retailer, or a different panel maker), you are not estopped — you may file your own IPR. Two cautions: (i) § 315(b)'s one-year clock from service of an infringement complaint applies to you, and (ii) the Board's discretionary-denial practice under Fintiv and the current Director's guidance can bar institution where your own parallel litigation is advanced.
• Practical note: the top-of-claim-set art is now a known quantity to Samsung Display. A second petition recycling the same references will draw a § 325(d) "same or substantially the same art" denial.

Pattern signals. This is a serial-campaign record, not a one-off: the same four-company coalition (BOE + Wuhan CSOT + Tianma + Visionox) jointly filed a coordinated set of at least five IPRs against Samsung Display's ITC-1351 patents — IPR2023-00941, IPR2023-00987, IPR2023-00988 and IPR2023-01075, plus IPR2023-01221 (terminated 2023-11-27) — and lost the group's later '593 challenge in IPR2024-00620 (instituted 2024-09-12; FWD ~2025-08/09; appealed by both sides on 2025-10-23 and 2025-10-27). It is a joint-defense / coordinated-invalidity pattern, not a defensive-aggregator (Unified Patents) pattern — Unified appears only as the litigation-data source on the patent record, not as a petitioner. Patent owner behavior: Samsung Display fought every proceeding to FWD and has pursued appeals (including the ITC appeal, No. 25-1791), i.e., it is an active, well-resourced, repeat enforcer rather than a licensing-only entity.

Recommended next steps

1. Pull the primary documents before relying on anything above. Go to PTAB E2E / PTAB Center (https://ptab.uspto.gov) and open IPR2023-01075: retrieve Paper 1 (Petition — gives the challenged-claim list and the exact grounds/references), the Institution Decision (date + partial-institution scope + Fintiv/§ 325(d) treatment), and the Final Written Decision (Paper ~47, 2025-01-06). Because the FWD is the dispositive document, quote the disposition rather than the Chinese-press summary: the FWD's concluding order reads "No Challenged Claim Unpatentable" under 35 U.S.C. § 318(a), with the panel of Lee, McMillin, and Hudalla (McMillin writing). Mirrors: https://portal.unifiedpatents.com/ptab/case/IPR2023-01075
2. Confirm the appeal's status on CourtListener / the Fed. Cir. docket for 25-1519. Determine whether the appeal was dismissed as moot or voluntarily in light of the November 2025 SDC–BOE global settlement, or remains pending. If it was dismissed, the FWD is final and non-appealable, and § 315(e)(2) estoppel is settled. If it is pending, estoppel still attaches on FWD issuance, but the invalidity judgment is not yet immune from vacatur if the case moots.
3. Map your estoppel exposure by entity, not by case. If you are TCL CSOT or a Wuhan China Star affiliate, assume § 315(e)(2) estoppel applies to you as petitioner/privy and build your case on § 112, system art, and claim construction. If you are a downstream distributor/retailer or an unaffiliated panel maker, you likely remain free to petition — but calendar your § 315(b) one-year deadline from the date you were served.
4. Re-check for a second petition. The ODP ingest clearly lagged reality on this patent, so do not assume IPR2023-01075 is the last word. Before filing or relying on "no PTAB activity," query PTAB E2E and the Unified Patents PTAB portal for any 2024/2025/2026-numbered petition against US 11,594,578 (particularly from BOE, which was still filing OLED IPRs against Samsung Display in late 2025 on other patents). If a fresh petition exists, the patent's posture flips from "hardened" to "contested."
5. If you are evaluating a demand letter today: the '578 patent has no dead claims to hide behind. Do not budget for an easy IPR kill. Budget for claim construction. The claim set is unusually long (92 claims) and heavily geometric — claims 1, 42, 45, 53, 67, 85 and 89 each contain independent positional/spacing limitations (e.g., claim 1's "entirely offset in the second direction from at least one of the second pixels or the third pixels," and the claim 85/89 "connection portion … greater than a shortest distance between the two parallel sides"). Those are non-infringement levers, and they were not resolved by the IPR because the Board does not decide infringement.

Uncertainty ledger (do not treat as fact): challenged-claim list; grounds and prior-art references; institution date and reasoning; whether the FWD addressed all 92 claims; whether a Director Review request was filed; the appellant/caption on 25-1519; any 2026 CAFC disposition. Each is marked above and should be resolved against PTAB E2E and the CourtListener docket before reliance.

I'll research the assignment chain for this patent, starting with the authoritative patent record and cross-referencing the USPTO Assignment Center and related sources.

Let me try to pin down the specific reel/frame and correspondent of record from the INPADOC/legal-events records.

I could not retrieve reel/frame or correspondent detail from the Assignment Center index through the sources available to me. Below I report exactly what is verifiable and mark what is not, rather than filling gaps.

Inventors

Pattern check — inventor retention. No unusual pattern. Samsung Display has continued to file applications naming Lee, Sang Shin as an inventor well after the 2012 priority date — e.g., EP 4,750,288 A1 (filed 2025-11-13, pub. 2026-05-27), US 2026/0150546 A1, EP 4,478,864 B1, and US 12,612,689 B2 ("Mask stick and mask assembly including the same"). Source: PatentGuru inventor search, https://www.patentguru.com/search?q=in%3D%28Lee%2C+Yeonju%29&filter=open (returns Lee, Sangshin records assigned to Samsung Display). There is no evidence of the inventor departing the original assignee within 12 months of filing, and no fire-sale precursor signal.

Original assignee

Samsung Display Co., Ltd. (1, Samsung-ro, Giheung-gu, Yongin-si, Gyeonggi-do 17113, Republic of Korea) — the entity named on the issued patent and the original assignee of record.

• Primary line of business: manufacturer of OLED display panels — the world's largest OLED panel maker (roughly 37% share of the OLED panel market as of Q2 2025 per Counterpoint, cited in https://www.esmchina.com/news/13687.html).
• Product embodying the claims: Yes, directly. The patent claims a sub-pixel arrangement/layout for an OLED panel; Samsung Display's mobile OLED panels (including panels supplied for Galaxy and Apple iPhone devices) are the commercial embodiment. Samsung Display itself asserts that competitors' smartphone OLED panels infringe the '578 patent — i.e., it treats the claim set as covering a product it makes and sells. Source: THE ELEC, https://www.thelec.kr/news/articleView.html?idxno=36758 and EET-China, https://www.eet-china.com/mp/a410639.html.
• Current status: Operating. No bankruptcy, receivership, dissolution, acquisition, or spin-off of the patent-owning entity appears in any source retrieved. The '578 patent is one of a family of six continuations (US 9,818,803; 10,854,683; 11,594,578; 11,980,077; 12,324,335; and pending US 2025/0268051 A1) all held by Samsung Display.

Assignment timeline

The record below comes from the assignment/legal-event data surfaced on the patent's Google Patents page (the rendering of the USPTO record supplied in the authoritative text) and from the assignment-related entries I could verify. I could not retrieve the reel/frame numbers or the correspondent of record — the single recorded conveyance is described in the record only as "ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS)." I did not access assignmentcenter.uspto.gov directly, so I am not filling in a reel/frame number by inference. That gap is an explicit limitation of this reconstruction, not a claim that no reel/frame exists.

• 2019-07-08 (executed, same day as filing) / recorded 2019-08-09 — Reel/Frame not retrieved
  • Conveyance: Assignment (Assignment of Assignors' Interest)
  • Assignor: Sang-Shin Lee (sole inventor)
  • Assignee: Samsung Display Co., Ltd.
  • Correspondent: Not retrieved. The record does not expose the filing attorney or the recording firm in the sources I could reach. Because this is the only conveyance in the chain, the "repeat correspondent" test cannot be run at all (there is nothing to compare against).
  • Context: Ordinary inventor-to-employer assignment accompanying the filing of this continuation application. Not an acquisition, not a fire-sale, not a securitization, not a transfer to an asserter — it is the initial transfer of title from the inventor to his employer.

Earlier links in the family (not separately verified for this patent). The '578 application is a continuation of US 15/811,599 (filed 2017-11-13, now US 10,854,683), itself a continuation of US 13/614,197 (filed 2012-09-13, now US 9,818,803), which claims Korean priority KR 10-2012-0022967 (filed 2012-03-06). Each of those parent applications would ordinarily carry its own recorded Lee→Samsung Display assignment from 2012 and 2017, but those records were not retrieved in this pass and I will not assert them.

No post-issuance assignment was found. The record shows a single reassignment event (2019-08-09). There is no change of name, no security agreement, no release, no license recorded, and no transfer to any IP-holding, licensing, or assertion entity. The patent is still held by Samsung Display.

Timeline diagram

timeline
    title Ownership of US 11594578
    2012 : KR priority filed
         : Parent US 13614197 filed
    2017 : Continuation US 15811599 filed
    2019 : This application filed
         : Lee assigns to Samsung Display
    2023 : Patent issues to Samsung Display
    2025 : Asserted against TCL CSOT

NPE / troll-pattern signals

Verdict

Operating-company assertion.

The chain is a single link: Sang-Shin Lee (sole inventor) → Samsung Display Co., Ltd., executed 2019-07-08 and recorded 2019-08-09 — the ordinary inventor-to-employer assignment for a continuation application, with no post-issuance transfer, security interest, or licensing conveyance in the record. Samsung Display is a global operating manufacturer whose OLED panels embody the claimed pixel-arrangement structure, it holds title in its own name (no shell LLC, no "IP Holdings"-type assignee, no asserter entity anywhere), and it has enforced the patent against actual panel-making competitors — BOE (ITC 337-TA-1351; E.D. Tex. 2:23-cv-00309; E.D. Va. 1:25-cv-00908) and TCL CSOT (N.D. Tex. 3:25-cv-01430). Not one of the eight NPE signals is present; the only signal I could not evaluate (#3, repeat correspondent) is untestable because the chain has a single link.

Verification link: https://assignmentcenter.uspto.gov/ (search by patent number 11594578; mirrored at https://assignment.uspto.gov/patent/index.html). I recommend pulling the reel/frame and correspondent directly from that interface, since the public renderings I reached show only "ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS)" with no reel/frame or attorney of record exposed.

Cross-reference notes and discrepancies with the earlier-generated sections

1. Confirms the earlier bibliography. The earlier sections recorded "Samsung Display Co., Ltd. (assignment of Sang-Shin Lee, recorded Aug. 9, 2019)." That is consistent with — and is the entirety of — what I can verify on the assignment side: a single 2019 assignment, nothing since.

2. Refinement on the ITC/venue timeline. A primary-source filing (Defendants' stay motion, E.D. Va. 1:25-cv-00908, https://storage.courtlistener.com/recap/gov.uscourts.vaed.[574882](/patent/574882)/gov.uscourts.vaed.574882.5.0.pdf) states that the ITC found no violation of § 337 on April 9, 2025, that Samsung Display petitioned the Federal Circuit for review on May 20, 2025 in No. 25-1791 (with Mianyang BOE a party), and that Samsung Display filed the E.D. Va. complaint on May 28, 2025 asserting the '803, '683, '578, '599 and '593 patents. This corroborates the earlier section's no-violation characterization and pins the CAFC appeal number to a specific date and party.

3. Unresolved conflict carried forward. Chinese-language reporting (EET-China, June 2025, https://www.eet-china.com/mp/a410639.html) states the ITC "recognized infringement" while denying the import ban, whereas the ITC record and the E.D. Va. filing state a no-violation finding. This is a substantive conflict in the litigation narrative, but it does not affect the ownership analysis — no party disputes that Samsung Display was the owner of record at all relevant times.

4. Nothing in the assignment record supports an NPE characterization. Any database or commentary labeling US 11,594,578 as a "troll" or licensing-vehicle asset would be contradicted by the assignment record, which shows title never leaving Samsung Display.

I'll search for the specific prior-art citations on US 11,594,578.

Now let me search for the specific prior-art references cited on the face of the patent and in prosecution.

Prior-Art Analysis — US 11,594,578 B2

Method note. Per the operating rules I interpret every identifier literally and do not auto-correct OCR or database strings. Where a retrieved string is ambiguous, I reproduce it as found and flag it rather than fixing it.

1. Threshold finding: the authoritative text supplied has no "References Cited" section

This is the single most important thing to state before any table. The Google Patents rendering of US11594578B2 in your message contains:

• the Description, Claims (92), Abstract, Classifications, and Landscapes;
• "Cited By (2)" and "Families Citing this family (157)" — forward citations; and
• no "(56) References Cited" block and no "Citations" block.

So the authoritative source is silent on which references the examiner cited. I therefore retrieved the front-page (56) lists from the patent's own issued-PDF OCR and from its immediate parent. Those retrieved lists are truncated (the OCR breaks at "5,485,293 A … (Continued)" and at "7,612…"), so what follows is a partial, not exhaustive, record of cited art. I could not obtain a PTO-892, an IDS, or the complete (56) list. Treat the reference set below as incomplete.

2. Correction to a likely premise: "Cited By" on this record is not prior art

Likewise, the 157 entries in "Families Citing this family" are forward citations. That list is dominated by post-2012 BOE/CSOT/Tianma/LGD filings (e.g., CN 105448951, US 11,233,096, US 11,448,807, US 12,324,335) and cannot anticipate the '578. One entry — KR101328979B1 (Samsung Display, priority 2011-06-30, published 2013-11-13) — carries a pre-priority date, but it appears in the citing table, i.e., it cites this family; it is not cited art in the '578 file. Do not treat it as § 102 art on this record.

This distinction matters because the request is framed as "each patent citation for 11594578." Only the (56)-type entries below are true citations of prior art.

3. U.S. patent documents cited on the face of US 11,594,578 B2 (partial, OCR from the issued PDF)

Dates are as printed. Descriptions are given only where I have reasonable confidence; "subject matter not verified in this session" means exactly that — I did not confirm the reference's disclosure and will not invent one.

Forward-referenced-but-in-parent-only items. The parent US 10,854,683 front page cites a second cluster that is not visible in the '578 OCR snippet and may or may not be on the '578's face: US 7,710,484 B2 (Oda, 5/2010); US 7,982,786 B2 (Nishida et al., 7/2011); US 8,026,669 B2 (Kawasaki et al., 9/2011); US 8,330,352 B2 (Sung et al., 12/2012); US 8,350,468 B2 (Ko et al., 1/2013); US 8,354,789 B2 (Kim et al., 1/2013); US 9,431,469 B2 (Park et al., 8/2016); US 9,837,476 B2 (Park et al., 12/2017). ⚠️ Two of these post-date the 2012-03-06 priority date on their printed grant dates (9,431,469 and 9,837,476), and three more grant within ~10 months of it. If they were cited against the '578 or its continuations, they were cited relative to the 2017/2019 filing dates, not the 2012 priority — meaning they are not available as § 102 art against claims entitled to the 2012-03-06 priority date absent an earlier effective filing date of their own. Their appearance in a (56) list is not evidence of § 102 availability. Flagged because this is an easy error to make.

4. U.S. patent application publications cited (from the parent's (56), carried into this family)

These are the ones the EPO family member's background paragraph expressly names as disclosing pixel arrangement structures of a display (see § 6), which makes them the applicant's own acknowledged closest art:

I did not verify the actual disclosure of any of these against the '578 claim language, and I will not assert § 102 anticipation against a specific claim on a title alone.

5. Foreign patent documents cited (partial)

6. The applicant-acknowledged prior art (from the EP family member) — and a discrepancy to flag

The EP family member EP 2637209 B1 states, in its background:

"US2002/0015110 A1, US2005/0001542 A1, US2004/0108818 A1, US2002/0070909 A1 and GB2437110 A disclose pixel arrangement structures of a display."

⚠️ Flagged discrepancy (carried forward, not resolved): the US specification as supplied in your authoritative text contains no such sentence and no "Background" citation of any of these five references. The U.S. Background section is purely generic (self-luminous OLED, FMM vs. aperture-ratio tradeoff). So the U.S. and EP siblings of the same invention do not present the same acknowledged prior-art sentence. That is a genuine intra-family inconsistency and it matters: the EP prosecution treated those five documents as the closest known arrangements, whereas the U.S. record as rendered does not.

These five — US 2002/0015110 A1, US 2005/0001542 A1, US 2004/0108818 A1, US 2002/0070909 A1, and GB 2437110 A — are therefore the most defensible "closest prior art" set to start from, because the patent owner itself characterized them as disclosing pixel arrangement structures.

7. § 102 anticipation assessment (per-claim), stated with appropriate restraint

Statutory framework. The '578 claims take a 2012-03-06 effective filing date (KR 10-2012-0022967; U.S. chain filed 2012-09-13). Pre-AIA § 102 therefore governs: art must be § 102(a)/(b)/(e)/(f)/(g) as of that date. A § 102 anticipation requires a single reference disclosing every limitation.

Limitations that must be met by one reference. Independent claims 1, 42, 45, 53, 67, 85, 89 each carry dense geometric requirements. The hardest to meet in a single reference are:

1. Claim 1 — (i) a band whose width equals the first-pixel width, running parallel to the first direction and completely overlapping a first-pixel row, being "entirely offset in the second direction from at least one of the second pixels or the third pixels"; and (ii) same-colour spacing > different-colour spacing within a row.
2. Claim 42 — three distinct center-to-center pitch relationships (first/second/third lengths) plus second-pixel area > third-pixel area.
3. Claim 45 — a virtual quadrangle whose outer edges are made of edges of four neighboring first pixels, containing only one second/third pixel, all others outside it.
4. Claims 85 / 89 (and 40, 43, 51, 63, 81) — the "elongated first pixel with two connection portions, at least one longer than the shortest distance between the two parallel sides" limitation.

My assessment, with confidence levels:

• No reference in the retrieved cited set is a credible single-reference § 102 anticipation of any of the seven independent claims. Confidence: moderate-high, limited by the incompleteness of the (56) list and the fact that I did not read the references' full texts.
• Best § 102 candidates if the differential-spacing and band-offset limitations can be met: US 7,274,383 B1 and US 2002/0015110 A1 (Brown Elliott / Clairvoyante), followed by US 6,867,549 B2 (Cok, Kodak) and US 2005/0001542 A1 (Kiguchi). Clairvoyante/PenTile art genuinely teaches non-stripe, non-rectangular colour-pixel arrangements with different-sized emitters — the closest structural analogue in the set. Confidence: moderate.
• Dependent claims with the lowest § 102 threshold (i.e., most likely to be met by one of the cited layout references) are the shape/area claims — 12, 13, 14, 15, 16, 17, 39, 50 — because "polygonal," "octagonal," "same octagonal shape," and "first pixel smaller than second/third" are all concepts present in the sub-pixel-arrangement literature. Confidence: moderate.
• References in the set that are categorically poor § 102 art for these claims: US 2002/0140833 A1 (Hirai, image interpolation — examiner's own G06T 3/4007 classification), US 2004/0246426 A1 (Wang, LCD colour filter — G02F 1/133514), and the pre-1997 references (US 4,491,863; US 4,642,619; US 4,965,565; US 5,113,274; US 5,142,392; US 5,461,503; US 5,485,293). Confidence: moderate-high.
• Retroactivity caution: US 9,431,469 B2 and US 9,837,476 B2 (visible in the parent's (56)) post-date the 2012-03-06 priority date; they cannot be § 102 art against claims enjoying that priority. Confidence: high on the dates as printed.

Cross-check against the PTAB record (from the previously generated section, not repeated here): IPR2023-01075 ended with a Final Written Decision of "No Challenged Claim Unpatentable" under 35 U.S.C. § 318(a) (2025-01-06; panel Lee, McMillin, Hudalla). This is strong circumstantial evidence that the asserted art — even in § 103 combination, a lower bar than § 102 — did not reach the claim limitations. An anticipation theory built on the art of record therefore starts from behind. ⚠️ But I again could not retrieve the petition's grounds or reference list in this session, so I cannot state which references were tested; "no challenged claim unpatentable" ≠ "all cited art is irrelevant."

8. What I could not obtain (stated rather than guessed)

1. The complete (56) "References Cited" list from US 11,594,578 B2 — the OCR truncates at "7,612…" and the Google Patents text omits the section entirely.
2. The PTO-892 / IDS and the examiner's search report for application 16/505,536.
3. The EPO search reports for the family — the parent's front page records them: EPO Search Report dated Jul. 8, 2013 for EP application 13157562.3 (6 pages); EPO Search Report dated Dec. 20, 2013 for EP 13178648.5 (9 pages); EPO Search Report dated Jul. 9, 2014 for EP 14155124.2 (8 pages). These are the documents most likely to contain an X/Y-categorised prior-art assessment of exactly this virtual-square layout, and they are retrievable from the EPO Global Dossier / publication server.
4. The non-patent literature (if any) on the (56) face.
5. The IPR2023-01075 grounds, references, and challenged-claim list.

Uncertainty ledger: descriptions marked "subject matter not verified"; the "Koeko"/"Kondo" printing of US 7,110,031 B2; the OCR variant "Imaiya"/"Inuiya" for US 6,883,364 B1; the truncated final entry ("US 7,612…"); and every § 102 claim mapping above, which is analytical and provisional.

9. Recommended next steps to convert this from provisional to docket-grade

1. Pull the issued PDF's (56) at full resolution — patentimages.googleusercontent / USPTO PatentCenter for US 11,594,578 B2 — to complete the list past the "7,612…" truncation and capture the Examiner Cited vs. Applicant Cited distinction.
2. Retrieve the three EPO search reports listed in § 8(3). These will give you the family's X/Y art with relevance categories — the fastest route to a defensible "closest prior art" statement.
3. Obtain IPR2023-01075 Paper 1 (Petition) from PTAB E2E at https://ptab.uspto.gov to get the actual asserted references and the challenged-claim list.
4. For each candidate reference, run the limitations checklist against claims 1, 42, 45, 53, 67, 85, 89 — specifically searching for a single disclosure of the band-offset language of claim 1 and the connection-portion length language of claims 85/89. Those two limitations are the practical § 102 gatekeepers.
5. Do not spend budget on the pre-1997 references or on US 2002/0140833 A1 / US 2004/0246426 A1 unless the full (56) shows the examiner relied on them for a specific claim.

I'll ground this analysis in the prior-art references actually identifiable from the record and from live searches (including the IPR2023-01075 grounds, if retrievable).

Let me try to pin down the actual IPR2023-01075 grounds and the references cited on the face of the '578 and its parents.

Obviousness Analysis under 35 U.S.C. § 103 — US 11,594,578 B2

0. Scope, and what "the Prior Art section of this page" actually contains

Per the operating rules, I interpret US 11,594,578 B2 literally and build on the prior sections rather than repeating them (bibliographic table, claim map of the seven independents 1/42/45/53/67/85/89, litigation posture, IPR2023-01075 FWD).

Important framing caveat. The patent page supplied in this task does not contain a discrete "Prior Art" section. Its prior-art-related content is limited to four things:

1. "Prior art keywords": pixels, pixel, arrangement structure, shortest distance, display — i.e., the conceptual fields the citation search treated as the invention's space.
2. "Families Citing this family (157)" — forward citations, a minority of which carry the "*" = cited-by-examiner marker (e.g., KR101328979B1, Samsung Display, priority 2011‑06‑30).
3. "Cited By (2)" — US 2022/0181405 A1 (Samsung Display, 2020) and US 11,980,077 B2 (the family's own child).
4. The IPR2023-01075 challenge, which is the only adversarial prior-art event on the record.

Everything else in my prior sections (the litigation tables, the docket detail) is not prior art. So a § 103 analysis must be assembled from: (a) the page's citation lists, (b) the admissions in the '578 specification itself, and (c) pre‑2012‑03‑06 display-layout art I could verify via live search.

I could not retrieve the IPR2023-01075 petition, institution decision, or FWD body, and my searches for it returned unrelated PTAB petitions. (Notably, one retrieved petition, at https://ptacts.uspto.gov/ptacts/public-informations/petitions/1558493/download-documents, reproduces the specification of a sibling Samsung application — "Filed: April 26, 2013 … Title: PIXEL ARRANGEMENT STRUCTURE FOR ORGANIC LIGHT EMITTING DIODE DISPLAY" — whose quoted claims are not the '578's claims; it demonstrates the family's common disclosure and the "blue shortest life span → larger blue area" rationale but should not be attributed to the '578.) Therefore the combinations below are my own independent obviousness theory, not a recitation of what the IPR petitioners argued. Flagging that explicitly.

1. Governing law and the correct statutory frame

Pre-AIA § 103(a) governs. The '578's earliest effective filing date is March 6, 2012 (KR 10‑2012‑0022967) with the U.S. chain beginning September 13, 2012 (US 13/614,197). Both precede March 16, 2013, so the AIA's first-inventor-to-file §§ 102/103 do not apply; the analysis runs under pre-AIA 35 U.S.C. § 103(a) with pre-AIA § 102 as the reference-qualification gate. This matters because pre-AIA § 103(a) permits a broader obviousness inquiry (no statutory "otherwise available as prior art" wrapper), and because the reference dates must be measured against March 6, 2012.

Framework: Graham v. John Deere Co., 383 U.S. 1, 17–18 (1966) (scope/content of art; differences; level of ordinary skill; objective indicia), applied through KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007) — including KSR's explicit reliance on "design incentives and other market forces," "a finite number of identified, predictable solutions," and "predictable variation[s]." I will not over-claim holdings I cannot verify; where I invoke the "mere change of relative dimensions is a matter of design choice" line of authority I flag it as a doctrine rather than pinning it to a specific reporter cite.

2. Person of ordinary skill in the art (POSITA) — proposed, not record-verified

For a 2012 filing date, a POSITA in this field would have: a B.S. in electrical engineering, materials science, physics, or an equivalent, plus 3–5 years in OLED display panel design or OLED front-plane process integration — specifically, experience with (i) sub-pixel rendering and RGBG/"diamond"-type layouts, (ii) organic-emission-layer evaporation through a fine metal mask (FMM), including mask web/rib geometry and shadowing, and (iii) pixel-defining-layer (PDL) layout and aperture-ratio budgeting.

This is my construction; the Board's or the parties' adopted level of skill in IPR2023-01075 was not retrievable. Note also that the '578's specification defines its units broadly — "the pixel refers to a minimum unit for displaying an image (for example, the minimum addressable unit of the display)" — which broadens what a POSITA must be taken to know about sub-pixel layout.

3. The prior-art landscape, divided by evidentiary strength

Bucket A — Admitted prior art inside the '578 itself (lowest risk, always available)

The specification's own Background and Description are prior-art admissions usable under pre-AIA § 103(a) (MPEP 2129):

• A1 — The known trade-off. "When reducing a gap between the neighboring pixels to obtain a high aperture ratio of the pixel, deposition reliability may be deteriorated. On the other hand, when increasing the gap between the pixels to improve the deposition reliability, the aperture ratio of the pixel may be deteriorated."
• A2 — The known OLED stack/geometry. Plural differently-colored pixels; a pixel defining layer defining each pixel's "area and a shape"; gate/data/driving-power lines; anode/organic emission layer/cathode per pixel; and that a pixel's "shape … may be defined by the power lines, the pixel defining layer, the anode, or the like."
• A3 — Shape is a routine design variable. "…even though each of the pixels are drawn as stereotypical polygonal shapes in the drawings, the present invention is not limited to this shape. That is, the shapes of the pixels may be modified to avoid interference with the other components of the OLED (e.g., wirings)…"
• A4 — Whole-family content with 2012‑03‑06 priority. The sibling filings (e.g., the Apr. 26, 2013 application reproduced in the PTAB petition above) share this disclosure, including the statement that "the second pixels 200 that emit blue light have the shortest life span … [so] the second pixels 200 have a larger area … thereby suppressing the deterioration of the life span." Because they share the '578's priority, they are not § 102/§ 103 art against it — but the rationale they express is corroboration that the blue-area design rule was a known engineering principle, not an insight of the '578.

A1–A3 are the strongest obviousness material available, because they are the applicant's own words and cannot be antedated.

Bucket B — Verified pre‑2012 references (search-confirmed content)

Bucket C — Record-listed but content-unverified

• KR101328979B1, Samsung Display, Organic light emitting display device, priority 2011‑06‑30, carrying the "cited by examiner" marker in this page's citation table. Its Korean filing/publication dates would fall before 2012‑03‑06, which would make it pre-AIA § 102(a)/(e)-type art. I could not verify its disclosure or its exact publication date, and its appearance in a "citing" table with an earlier priority is chronologically anomalous. Treat as a lead, not a finding.
• The remaining forward-citation entries are post-dated relative to the '578 (e.g., KR101900914B1, priority 2012‑09‑18; CN103123927B, 2013‑01‑24; JP6143056B2 and TWI594415B, both registry B-grants citing this family) and are therefore not § 103 art. They are useful only as evidence that an octagonal-pixel-unit-with-striped-RGB architecture and octagon-vs-quadrilateral area-vs-perimeter reasoning were active in the field shortly after — not as art.

Bucket D — Corroborative-but-post-dated (usable for what a POSITA knew, not as § 103 references)

• Post-2012 filings (e.g., BOE's US 2021/0271010 A1 and the EP 3 846 215 B1 text) state the general knowledge explicitly: "in order to reduce … cost, a pixel arrangement can adopt a pentile mode," "diamond-shaped pixel arrangement"; FMM "limitation conditions (e.g., the size of the opening and the size of the bridging portion)"; "in general, in the case of forming a sub-pixel, an intermediate layer may be formed … through a mask assembly … a set or certain interval is required between pixel electrodes adjacent to each other"; and "light-emitting materials of sub-pixels of different colors are different in efficiency and service life, and thus, the sub-pixels of different colors also differ in light-emitting area and a certain ratio needs to be met."
  I flag these as post-priority; I use them only to demonstrate that these were recognized engineering parameters, not to date them into the prior art.

4. The limitation inventory that an obviousness case must meet

Reduced to the distinct technical requirements (not repeating the earlier claim map):

• L1 — Three differently-colored pixel types + a pixel-defining layer. (Trivial: A2.)
• L2 — Mutually perpendicular row/column lattices, with rows of first pixels alternating with rows of alternating second/third pixels, and the same in columns. (The "square lattice.")
• L3 — A line through the centers of a second row passes between first pixels of adjacent first rows. (The "interleaved half-pitch" relationship.)
• L4 — First pixels and either second or third pixels alternate along a diagonal. (Diamond structure.)
• L5 (negative) — The horizontal band spanned by a first-pixel row (band height = first-pixel width in the second direction) is entirely offset from ≥1 second/third pixel in an adjacent row.
• L6 (relative distance) — Shortest first↔first distance in a row > shortest second↔third distance in a row.
• L7 (cl. 42) — Center-to-center distances: B↔R = first length; G↔B and G↔R = second length; G↔G = third length > second length; plus B area > R area.
• L8 (cl. 45) — Only one second/third pixel lies inside a virtual quadrangle bounded by edges of four neighboring first pixels, all others outside; and the inter-first-pixel distance > the enclosed pixel's perpendicular distance to the next second/third pixel.
• L9 (cl. 53) — Column-direction first↔first distance > the column-direction width of a second/third pixel.
• L10 (cl. 67/89) — First rows are G+B alternating; second rows are G+R alternating; second-row first pixels lie entirely outside the band of the adjacent first row's first pixels.
• L11 (cl. 85/89) — Each first pixel is elongated along a diagonal with two parallel sides and two "connection portions", at least one connection portion longer than the shortest distance between the parallel sides (i.e., long end-caps).

The real battleground is L5, L6, L8, L9, L10 and L11. L1–L4 are met by the known pentile lattice; L5/L6/L8/L9 are geometric consequences of sizing and centering choices; L10 and L11 are the narrow, structure-specific limitations.

5. The combinations, with motivations

Combination I — The base lattice (defeats L1–L4, and supplies the geometry for L5, L6, L8, L9)

Primary: US 2007/0040952 A1. Secondary: EP 2270579 A3 (Brown Elliott / Clairvoyante octal grouping + 1999 Information Display article). Optionally: US 2011/0057950 A1.

Mapping. US 2007/0040952 A1 teaches the R/G column alternation with column-repeated blue and four R/G pixels enclosing two adjacent-row blue pixels — i.e., a virtual-quadrangle enclosure of one color by the other two, with the enclosed color occupying a larger footprint. Brown Elliott teaches that in such an arrangement one color is given twice the number of positions while the other two are allotted larger areas, and supplies the design rationale (differential visual acuity; reduced blue resolution). Superimposing the two yields: a small, most-numerous green at the center of each repeat unit, with alternating larger blue and red at the vertices of the surrounding quadrangle — i.e., L2, L3, L4 and the "alternately arranged along a diagonal" requirement of L4.

Motivation to combine.

1. Same field, same problem. Both references address sub-pixel layouts for color flat-panel displays where two colors must be given more area and one color more positions to reduce emitter count without visible resolution loss. US 2007/0040952 A1 itself invokes the "PenTile Matrix" arrangement, showing the two references' subject matter is the same art.
2. Express teaching in the references. Brown Elliott's octal grouping is not merely compatible with the enclosure structure — it requires the enumeration asymmetry (2× positions for one color) that the enclosure architecture implements.
3. Predictable result. Re-arranging three rectangles into a diamond/quadrangle lattice changes no principle of operation and yields the expected aperture/size trade-off; KSR "predictable variation."
4. LCD→OLED borrowing is ordinary. The physics driving the layout (green dominates luminance; blue is least-resolved and least-efficient) is display-type agnostic. A POSITA integrating an OLED front plane in 2012 would consult established RGBG layouts as a starting point — which is precisely what the record shows happening in the field.

Combination II — Sizing, the "blue-largest" rule, and the relative-distance limitations (defeats L5, L6, L7, L9)

Combination I + the '578's own admitted trade-off (A1) + the recognized OLED design rule that the shortest-lived emitter is enlarged (+ optionally US 2011/0057950 A1 for the deployed commercial panel).

Mapping.

• B area > R area (L7; cl. 42, 10) is met by the known design rule that blue has the shortest life span and lowest efficiency and is therefore given a larger emitting area to equalize aging. The '578 does not discover this rule — it recites it in the specification ("the second pixels 200 that emit blue have the shortest life span … [so] larger area … thereby suppressing the deterioration of the life span"), which is an admission of the very motivation a POSITA would have had.
• The center-to-center relationships (L7) are an inherent geometric consequence of the lattice, not an independent invention. Once one green sits at the center of each repeat square and one blue and one red sit at the vertices, the G↔G center distance equals a full lattice period while G↔B, G↔R and B↔R are strict fractions of it — so "third length > second length" and "second length / first length" fall out of the geometry the primary reference teaches.
• L6 / L9 (the same-color spacing must exceed the different-color spacing) follows once the sizes of Combination II are chosen: enlarging blue/red while keeping green small at the square centers necessarily pushes the green-to-green gap above the blue-to-red gap.
• L5 / L10's "entirely offset" / "entirely outside" limitation is met, or at least rendered obvious, by the shape/size choice: where the first pixel's extent in the second direction is small (a rhombus/elongated green, as taught by Brown Elliott's smaller, more numerous color and by A3's admission that shapes are freely modified), the band spanned by a green row cannot reach the red/blue centers in the adjacent row. A negative spatial limitation of this kind is a result of sizing, and KSR treats the selection among a finite number of predictable dimensional solutions as within ordinary skill.

Motivation to combine.

1. The problem is stated in the art. A1 is the applicant's own admission that the field was already optimizing aperture ratio against FMM deposition reliability — the exact problem the '578 identifies. Where the problem is known and the solution is a rearrangement of a known layout, KSR supplies the motivation.
2. FMM mechanics force the asymmetry. Each color is deposited through its own mask; a mask must retain webs/ribs between openings for the same color, so the same-color spacing has a process-mandated floor, while different-color pixels (different masks) are only constrained by the PDL/shadowing gap. A POSITA improving deposition reliability would therefore differentially increase same-color spacing — which is exactly the L6/L9 inequality. (The post-dated sources in Bucket D confirm this was explicitly recognized as a design parameter; I do not rely on them as § 103 art.)
3. Life-span equalization is a textbook, efficiency-driven reason to enlarge the blue and shrink the green, matching Brown Elliott's enumeration asymmetry.
4. Predictable, purely geometric result; no new physical effect; reasonable expectation of success is near-certain, since the outcome is measured by a ruler.

Combination III — The "display" claims (claims 53, 67, 89's row architecture)

Combination I + II + US 2007/0040952 A1's column-repetition teaching.

Claims 53 and 67/89 are written in display-level "row" terms (alternating first-pixel rows and mixed rows; G+B rows and G+R rows). US 2007/0040952 A1 already teaches two-color alternation along a row and repeated single-color along a column, which is the row vocabulary these claims adopt; combined with the diagonal (L4) teaching, a POSITA would arrive at G+B rows alternating with G+R rows as one of a small number of ways to instantiate the same lattice. The "shortest first↔first distance in the column direction > the column-direction width of a second/third pixel" limitation (L9) again reduces to relative sizing, which Combination II fixes.

Candid weakness: the specific pairing of claim 67/89 (G+B in one row, G+R in the next, with the second-row greens entirely outside the first-row green band) is a stricter, different lattice than the FIG. 1 diamond. I could not verify a pre‑2012 reference squarely teaching that exact pairing. The strongest available argument is KSR "finite number of identified, predictable solutions" plus the half-pitch staggering inherent in any two-row repeat — but I would rate the mapping for claim 67 at moderate confidence at best, and it should be treated as the second-weakest point in any challenge.

Combination IV — The shape claims (40, 43, 51, 63, 81, 85, 89) — the weak flank

L11 (elongated first pixel; two parallel sides; end-cap "connection portions" longer than the inter-side width) is the one limitation I could not map to a verified pre‑2012 reference. The defensible arguments are:

1. Applicant admission (A3). The specification concedes that pixel "shapes … may be modified to avoid interference with the other components of the OLED (e.g., wirings)" and that the drawn polygons are "stereotypical." A claim drawn to a pixel shape therefore claims what the specification treats as a routine design variable.
2. Enumerated alternatives = predictable solutions. The specification itself discloses quadrilateral, hexagonal, octagonal, rhombus and parallelogram variants and both blue-larger and red-larger variants. Under KSR, an inventor who claims a genus spanning every one of his own disclosed alternatives is exposed to the argument that the genus was a set of identified, predictable options.
3. But this is not enough by itself. Without a reference teaching long end-caps, claims 85/89 (and dependents 40, 43, 51, 63, 81) are the least vulnerable claims in the set. I flag this as an unresolved evidentiary gap rather than asserting a combination I cannot ground.

6. Claim-by-claim summary

Note on independent-claim scope overlap. Claims 1, 42 and 45 are three different verbalizations of essentially the same lattice; claims 53 and 67 are two verbalizations of essentially the same row architecture; claims 85 and 89 are the same shape limitation bolted onto the two row architectures. That redundancy matters strategically: an obviousness theory that lands on the common geometry (Combinations I + II) simultaneously reaches five of the seven independents, and the residual exposure is concentrated in the L11 shape language.

7. Countervailing evidence — where the obviousness case is likely to fail

A credible § 103 opinion must state the patent owner's rebuttal, not just the combination.

1. The PTAB already rejected this theory once. IPR2023-01075's FWD (Jan. 6, 2025) disposed of the case as "No Challenged Claim Unpatentable," 35 U.S.C. § 318(a) — see the record mirrored at https://portal.unifiedpatents.com/ptab/case/IPR2023-01075. That is not binding on a district court and does not establish validity over art not raised, but it is powerful § 103 counter-evidence for whatever claim set was challenged (the challenged-claim list remains unverified — see my PTAB section).
2. "Teaching away" argument on the key inequality (L6/L9). Samsung's best non-obviousness story is that the art taught minimizing all gaps to maximize aperture ratio (A1's first sentence), and that the claimed solution deliberately makes one gap larger than the others — the opposite of the accepted optimization direction. If Samsung can show the art uniformly pushed toward gap minimization, it converts the '578's inequality into a "teaching away" case. My rebuttal is that A1's second sentence discloses the countervailing FMM-reliability constraint, and that the same-color vs. different-color mask asymmetry explains why only one gap must grow — but this is a genuine contest, not a formality.
3. Objective indicia. Potential Graham factors favoring Samsung: commercial success of Diamond-type Samsung OLED panels; industry adoption/copying — e.g., TCL Huaxing's own investor-relations commentary (Tencent/同花顺 Q&A page for 维信诺/TCL, retrieved from basic.10jqka.com.cn/002387) states that TCL Huaxing's self-developed "Pearl" arrangement "可提供与三星Diamond排列相同的最高视觉等效分辨率" (provides the same maximum visually-equivalent resolution as Samsung's Diamond arrangement) with aperture ratio "可达20%，高于三星Diamond像素排列17%开口率" (up to 20%, vs. 17% for Diamond); and licensing/royalty payments (the Nov. 2025 SDC–BOE global settlement with royalty terms, per the KED report https://www.kedglobal.com/us/electronics/newsView/ked202511200007). Each requires a nexus to the claimed geometry and not to unclaimed features (rendering algorithms, mask technology, panel scale).
4. Hindsight risk on the negative limitations (L5, L10). "Entirely offset" / "entirely outside" are negative spatial limitations that a defendant will inevitably be accused of reading out of the primary reference after seeing the claim. Expect Samsung to argue that no reference teaches a band that is completely clear of the adjacent-row pixels, and that this is where the claims escape the art.
5. Error in a third-party database (carried forward). One aggregator mislabels the '578 as "OLED pixel circuit driving and display architecture." If a § 103 theory relies on pixel-circuit/driving art (e.g., US 7,414,599), it is aimed at the wrong statutory subject matter for this patent and will be rebutted as non-analogous. Flagging so this error is not propagated into the invalidity analysis.

8. Bottom line

• Governing provision: pre-AIA § 103(a) (effective filing date March 6, 2012).
• Best combination: US 2007/0040952 A1 (R/G column alternation with column-repeated blue enclosed by four R/G pixels from two adjacent rows; PenTile discussed) in view of EP 2270579 A3 / Brown Elliott's octal grouping (one color given twice the positions; larger areas for the other two), further in view of the '578's own admitted aperture-vs-FMM-reliability trade-off and the recognized OLED rule that the blue, shortest-lived emitter receives the larger area. Motivation: same field and same problem; express teachings in both references; mask-web geometry forcing differentially larger same-color spacing; a finite, predictable set of geometric solutions; near-certain expectation of success. This combination reaches claims 1, 42 and 45 with reasonable force and claims 53 and 67 with somewhat less.
• Weakest flank: claims 85 and 89 and the shape dependents (40, 43, 51, 63, 81), because I could not verify any pre‑2012 reference teaching the "elongated first pixel / end-cap connection portion longer than the inter-side width" geometry. Only the specification's own shape-flexibility admission and KSR's predictable-alternatives reasoning are available there.
• Net assessment: the structural claims rest on well-trodden pentile ground and are, in my view, reasonably vulnerable on a properly evidenced § 103 theory; the shape and negative-spatial limitations are where I would expect the case to be won or lost, and they are also the limitations that make a non-infringement case plausible. That dovetails with the earlier strategic finding: after IPR2023-01075, the productive attack surface is claim construction and non-infringement, not a straight § 103 kill.

9. Uncertainty ledger (do not treat as fact)

1. IPR2023-01075's petition, grounds, references and challenged-claim list were not retrievable; my combinations are independent proposals. No inference should be drawn that the petitioners used any reference named above.
2. KR101328979B1 is record-listed with an examiner-citation marker but its content/publication date is unverified, and its presence in a "citing" table with a 2011‑06‑30 priority is anomalous.
3. Exact publication dates for US 2007/0040952 A1, US 2011/0057950 A1 and EP 2270579 A3 were inferred from their publication-number series and retrieved text, not read off a front page; the disclosures were verified verbatim from retrieved documents, and the underlying Clairvoyance/Brown Elliott material long predates 2012.
4. The POSITA definition is my proposal, not the Board's or the parties' adopted standard.
5. Item 3 in § 7 relies on a secondary Chinese-language investor-relations page and the KED settlement report for the commercial-success/copying narrative; neither is a merits finding.
6. The date anomaly flagged in earlier sections persists: search results include documents dated after the stated current date. Those items are labeled by their own document dates and are not relied on for any § 103 conclusion.