US 8124988

US Patent 8124988 — Summary

Bibliographic data

• Patent number: US8124988B2 (application no. US12/153,910)
• Title: Light emitting diode lamp package structure and assembly thereof
• Inventor: Jacky Peng
• Assignee (original/current): Semisilicon Technology Corp. (Taiwan)
• Filing date: May 28, 2008
• Issue date: February 28, 2012
• Legal status: Active; adjusted expiration listed as November 3, 2028. Maintenance fees paid through the 12th year (small entity).
• Earlier publication: US20090294782A1 (published December 3, 2009)
• Cited prior art (examiner-cited): US7473933B2 (Ledengin), US7717596B1 (Bell), US20080087903A1 (Patent-Treuhand)

Abstract (condensed)

A light-emitting diode (LED) lamp package in which multiple LEDs, a control integrated circuit (IC), a circuit board, and four electric-conductivity supports are encapsulated inside a package body. The supports are respectively a positive DC voltage pad (Vdd), a data-input pad (Din), a data-output pad (Dout), and a negative DC voltage pad (Vss). The invention also provides a LED lamp assembly with a LED lamp, lampshade, socket, and a mount formed with a socket having four electrode contacts, such that the four electric-conductivity supports contact the four electrode contacts when plugged together.

Plain-language overview of each original independent claim

• Claim 1 (LED lamp package structure): A package containing (a) plural LEDs sharing a common electrode end; (b) a control IC having data-in, data-out, positive power-in, and negative power-in electrodes, which receives external control data through Din and controls each LED's illumination; (c) four conductors — first = Vdd (tied to the LED common electrode and IC positive power), second = data input, third = data output, fourth = Vss (tied to IC negative power); and (d) a package body encapsulating the LEDs, IC, and conductors, with portions of the four conductors exposed outside the package. In essence: a self-contained, data-addressable LED package with power and serial-data pins.

• Claim 16 (LED lamp assembly): An assembly combining (a) a LED lamp whose package body encapsulates LEDs and four electric-conductivity supports; (b) a lampshade covering the package body; (c) a socket body having a perforation and a plug-in member, the perforation holding the LED lamp while leaving the four supports exposed; and (d) a mount formed with a socket containing four electrode contacts, so that when the socket is plugged into the mount's socket the four supports contact the four electrode contacts. In essence: a pluggable LED-lamp + socket + mount structure for easy lamp-string assembly and replacement.

• Claim 21 (LED lamp assembly — board-insertion variant): A LED lamp made of a package body plus a circuit board; the package body encapsulates LEDs and four electric-conductivity supports wired to the board, and the board carries a control IC and forms Vdd, data-in, data-out, and Vss electrodes. A lampshade covers the package, and a mount's socket has four electrode contacts that contact the board's four electrodes when the circuit board is inserted into the socket.

• Claim 22 (LED lamp assembly — two-mount variant): A LED lamp with a package body and a circuit board carrying LEDs and a control IC on one side; electrodes are placed on two opposing sides of the board, with a data-input electrode and a data-output electrode respectively on those two sides. A lampshade covers the lamp's periphery, and two mounts — each having a socket with matching electrode contacts — couple to the two opposing sides of the board, with the lampshade situated between the two mounts. In essence: a through-board serial connection topology where each LED lamp bridges two mounts.

• Claim 23 (package structure — partial-board encapsulation): A package with plural LEDs (each having a common electrode end) mounted on a circuit board; a package body encapsulating the LEDs and only a part of the board; and a control IC (with data-in and data-out electrodes) electrically connected to the LEDs. The portion of the circuit board exposed outside the package body carries the control IC, and the exposed board provides first, second, third, and fourth conductive regions ("conduct electricity" portions) exposed outside the package. (Dependent claim 24 clarifies those regions as Vdd+, data-out connection, data-out connection [as printed], and Vss−.)

Post-grant events noted (with caveats)

• According to the Google Patents legal-events table included in the supplied record, an inter partes review was filed (IPR2024-00157, petitioner Xu Peicheng (Hong Kong) Optoelectronic Technology Co., Ltd., filed November 8, 2023) and a request for reexamination was filed November 17, 2023. A reexamination certificate dated December 13/17, 2024 reportedly confirmed patentability of claims 1–15 and 22, cancelled claims 16–21 and 23–24, and added new claims 25–43. I could not independently verify this reexamination certificate via USPTO search results in this session, so treat that status as per the Google Patents metadata rather than confirmed USPTO data.
• Litigation references on the Google Patents page list cases in the Delaware District Court (1:25-cv-01566) and Florida Southern District Court (1:26-cv-23238; 0:26-cv-61380), plus a "not instituted (merits)" notation for IPR2024-00157. I did not independently verify these dockets.

CAFC 2026 docket search — result

• No CAFC (Federal Circuit) 2026 appeal involving US patent 8124988 was found in the searches performed. My searches of the USPTO/CAFC docket space for "8124988" plus "CAFC 2026" and related terms returned no matching Federal Circuit matter. The only hits were general patent-database pages (Google Patents, FreePatentsOnline) and a Darts-ip case-family listing that requires a subscription for case details. I therefore have no authoritative evidence of a 2026 CAFC appeal for this specific patent number; if one exists it was not surfaced in these searches, and I cannot confirm or rule it out.

Uncertainty notes

• I did not access the USPTO assignment database, PAIR/Patent Center, or the CAFC docket directly in this session; the bibliographic data above comes from the supplied Google Patents record (which the user directed me to treat as authoritative when conflicts arise) and is consistent with the FreePatentsOnline result.
• The reexamination certificate content, new claims 25–43, and the listed litigation matters come from third-party metadata and were not independently confirmed against official sources here.

I'll search litigation databases for cases specifically asserting US 8,124,988.

Let me check the remaining dockets and confirm which patents-in-suit each case asserts.

Litigation involving US Patent 8,124,988

⚠️ Date flag first. This task header says today is April 26, 2026, but the dockets I pulled contain events dated May 7, 2026, June 30, 2025, August 2026, and September 2026 — i.e., later than the stated "current date." I am reporting the dates as the dockets give them and am not correcting any of them, per the standing instruction. Consistent with the earlier sections of this analysis, the header date appears stale.

Cases in which US 8,124,988 is the (or a) patent-in-suit

Patent-in-suit confirmation. Case 1 is confirmed on the '988 patent: the D. Del. docket entry "Report to the Commissioner of Patents and Trademarks for Patent/Trademark Number(s) 8,124,988" (D.I. 3, 2025-12-29). Cases 2 and 3 are confirmed by the Ex Parte complaint-analysis records and the party docket, which list "US 8,124,988 B2" for Semisilicon v. inMusic filed 05/07/26 in S.D. Fla.

Accused products (cases 1–3). The ION Audio "Uber Boom Ultra" (a Bluetooth speaker with beat-synced multicolor LED lighting) is the exemplary accused product, with the complaints listing 80+ additional ION Audio and other inMusic audio products alleged to contain the same infringing LED package. Assertion is on at least independent claim 1 of the '988 patent.

Notable procedural facts from the record:

• The inMusic complaints plead the ex parte reexamination certificate (recited as issued 2024-12-13) and allege infringement notice as early as 2024-08-28, formal notice 2025-05-15, and failed pre-suit sales-data negotiations (Oct 2025).
• Case 1's dismissal (2026-05-07) and the two S.D. Fla. filings (2026-05-07) are the same day — the D. Del. case was dropped and re-filed in Florida. That is consistent with the "Delaware case dismissed; refiled against inMusic in S.D. Florida" entry in the ownership timeline generated earlier, and it looks like a venue/first-to-file maneuver, not a merits loss.
• No CAFC appeal of any of these cases exists or is possible on this record (no judgment, and the cases ended by voluntary dismissal / are pending). This is consistent with the earlier section's finding of no 2026 Federal Circuit matter involving 8,124,988.

Semisilicon enforcement cases that do NOT involve the '988 patent

(Listed only to avoid mis-attribution — do not cite these as '988 litigation.)

Adjacent administrative proceedings (not "litigation," but part of the same dispute)

• IPR2024-00157 — Xu Peicheng (Hong Kong) Optoelectronic Technology Co., Ltd. v. Semisilicon Technology Corp., US 8,124,988 — petition filed 2023-11-08; institution denied on the merits 2024-05-20; rehearing denied 2024-09-16; post-institution fee refunded 2024-09-20. No FWD; no claim cancelled.
• Ex parte reexamination (control no. 90/015,32_, number truncated in sources) — request filed 2023-11-17; certificate recorded 2024-12-17 (effective 2024-12-13): claims 1–15 and 22 confirmed; claims 16–21 and 23–24 cancelled; new claims 25–43 added. This is the only claim-level adjudication touching the '988 patent, and it came from the CRU, not the PTAB or a district court.

Bottom line

Yes — US 8,124,988 has been asserted in litigation, and all three '988 suits are Semisilicon-as-plaintiff against inMusic, arising from the same ION Audio accused products.

1. 1:25-cv-01566 (D. Del.) — filed 2025-12-29, voluntarily dismissed 2026-05-07.
2. 1:26-cv-23238 (S.D. Fla.) — filed 2026-05-07, terminated 2026-05-08 (same-day-abandoned filing).
3. 0:26-cv-61380 (S.D. Fla.) — filed 2026-05-07, the live case; status open.

No case has produced a merits ruling, judgment, or appeal. The only substantive adjudication of the '988 patent's scope is the reexamination certificate, which cancelled claims 16–21 and 23–24 (now moot) and left claims 1–15 and 22 plus new claims 25–43.

Verification gaps I am flagging rather than papering over:

1. The patents-in-suit in Aura Labs (W.D. Wash.) and Huzhou Xinju (E.D. Tex.) are unconfirmed — I hit the tool-call limit on the last query. Neither has been shown in the sources reached to involve 8,124,988, and Aura Labs has Semisilicon as the defendant, so it should not be listed as '988 enforcement until the complaint is read.
2. The precise relationship among 1:26-cv-23238, 0:26-cv-61380 and the dismissed Delaware case is inferred from filing/dismissal dates that coincide on 2026-05-07. The dockets support "refiled," but I did not read the transfer/dismissal pleadings to confirm the parties' stated reason.
3. The reexamination control number is truncated ("90/015,32_") in every source I could reach, and I have not read the certificate PDF or the text of new claims 25–43.
4. The live status of 0:26-cv-61380 as of the true current date is unverified — the last docket activity I found is its 2026-05-07 filing; no post-filing docket was surfaced.

Primary sources: D. Del. docket (https://gaeflexstaging-dot-docketupdate.appspot.com/cases/Delaware_District_Court/1--25-cv-01566/Semisilicon_Technology_Corp._v._inMusic_LLC/ and https://cdn.pacermonitor.com/public/case/62156722/Semisilicon_Technology_Corp_v_inMusic,_LLC) · S.D. Fla. (https://gaeflexstaging-dot-docketupdate.appspot.com/cases/Florida_Southern_District_Court/1--26-cv-23238/SEMISILICON_TECHNOLOGY_CORP._v._Inmusic_LLC_et_al/) · https://dockets.justia.com/browse/circuit-11/state-florida/noscat-10?page=2 · N.D. Cal. & N.D. Ill. (https://www.pacermonitor.com/public/case/57337611/Semisilicon_Technology_Corp_v_Changzhou_Jutai_Electronic_Co,_Ltd_et_al ; https://law.justia.com/cases/federal/district-courts/illinois/ilndce/1:2025cv07388/[481651](/patent/481651)/53/) · PTAB IPR2024-00157 (https://ptacts.uspto.gov/ptacts/public-informations).

I'll reconcile the canonical "no PTAB activity on file" structured block against the IPR reference that appears in the patent record itself, then dig for details.

Found the proceeding the ODP block missed. Let me pull the merits reasoning and check for other filings.

PTAB / AIA-trial landscape — US 8,124,988 (Semisilicon Technology Corp.)

⚠️ Discrepancy flag (read first). The structured "PTAB proceedings on file" block in this prompt — the canonical USPTO Open Data Portal list — states there is no AIA trial activity on this patent. That is wrong or stale. Both the supplied Google Patents record (which lists "PTAB case IPR2024-00157 filed (Not Instituted - Merits)") and multiple independent PTAB docket sources show a filed and fully resolved IPR. I have therefore treated the structured block's "none" as superseded by the search results, per the operating rule to prefer live search results, and flag the conflict here rather than silently reconciling it.

Proceedings overview

One AIA trial proceeding has been filed against US 8,124,988 — IPR2024-00157 — and it never got past the institution threshold: institution was denied 2024-05-20, rehearing was denied 2024-09-16, and the case was terminated. No claim of this patent has ever been cancelled by the Board, and no FWD exists. The claim-level damage to this patent came from a different forum: an ex parte reexamination whose certificate cancelled claims 16–21 and 23–24 and added new claims 25–43, confirming only claims 1–15 and 22. So the honest defensive posture is not "the patent survived two IPRs and is hardened," and not "the troll has no case" either: it is "the Board rejected the obviousness challenge outright — meaning no IPR estoppel and no FWD findings to borrow — but the lamp-assembly claims (16–21) and the partial-board claim (23) are dead by reexamination, so the only live assertion targets are the package-structure claims (1–15), claim 22, and the new, unexamined-by-me claims 25–43."

IPR2024-00157 — Xu Peicheng (Hong Kong) Optoelectronic Technology Co., Limited v. Semisilicon Technology Corp.

• Type: Inter Partes Review
• Filed: 2023-11-08 (petition accorded a filing date 2023-11-22)
• Status: Institution Denied (verbatim from the docket: "Institution Denied"; "Trial Outcome: Institution Denied"). Case terminated 2024-05-20.
• Judge panel: Administrative Patent Judges Charles Boudreau, Elizabeth Reagan, and Sheila McShane (Tech Center 2800, Art Unit 2814)
• Petition grounds: Obviousness (§ 103) — the docket's issue-type field records "103"; the decision cites 35 U.S.C. §§ 312 and 314 and 37 C.F.R. § 42.100. The petition's exhibit set shows the art relied on:
  • Ex. 1003 — Declaration of Dr. E. Fred Schubert
  • Ex. 1004/1005 — CN201017882Y to Lai (with certified English translation)
  • Ex. 1006 — US 2008/0087903 (Stoyan) — note this is also the reference the examiner cited during original prosecution (listed in the patent record as US20080087903A1, Patent-Treuhand)
  • Ex. 1007 — US 2004/0240890 (Morgan); Ex. 1008 — US 2006/0012978 (Allsop)
  • Ex. 1009 — US 7,473,933 (Yan) — also examiner-cited
  • Ex. 1010–1013 — dictionary excerpts (claim-construction support); Ex. 1002 — the '988 prosecution file history
  • Caveat: the petition and Paper 9 are not reproduced in the sources I could reach, so I cannot quote the precise claim-by-claim ground mapping (e.g., which of Liang/Stoyan/Morgan was the primary reference for which claim). The exhibit list above is verbatim from the docket record.
• Institution decision: Denied, 2024-05-20 (Paper 9). The patent record characterizes the denial as "Not Instituted - Merits" — i.e., the Board was not persuaded Petitioner established a reasonable likelihood of prevailing under § 314(a) on the merits, rather than exercising a discretionary denial. I could not retrieve the text of Paper 9 and therefore cannot quote the panel's reasoning or say which limitation(s) carried the day. Treat the merits basis as well-supported by the record label but not independently verified.
• Final Written Decision: None. No FWD ever issued, so no claim was cancelled, confirmed, or construed on the merits in this proceeding. Do not let anyone tell you this IPR invalidated anything — it did not reach the merits.
• Settlement / termination: Not a settlement. Post-denial, Petitioner filed a Request for Rehearing on 2024-06-18 (Paper 11), which the Board denied on 2024-09-16 (Paper 12). Petitioner then filed a Request for Refund of the $22,500 post-institution fee on 2024-09-18 (Paper 13) — the Office refunds the entire post-institution fee when review is not instituted — and the Board's refund was approved 2024-09-20. That refund approval is the cleanest documentary proof that this IPR was never instituted.
• Appeal: No Federal Circuit appeal found, and none is realistically available: a denial of institution is generally insulated from judicial review under 35 U.S.C. § 314(d) (Thryv, Inc. v. Click-to-Call Technologies, 590 U.S. 45 (2020)). My searches for a 2026 CAFC matter involving 8,124,988 returned nothing.
• Defensive value: Two-sided. Because there was no final written decision, no § 315(e)(2) estoppel attaches — neither Petitioner nor its privies are barred from re-raising this art in a later IPR, and nothing stops you from asserting the same or better art. But equally, there is no FWD holding to cite and no institution decision reasoning you can borrow; you would be litigating the same obviousness fight from scratch against a panel that already found the Petitioner's showing wanting. Note the panel is now on record as skeptical of the Lai/Stoyan/Morgan-type combination, which cuts against a "copy-paste the petition" strategy in a new IPR.

Primary sources: PTAB E2E / PTAB Public Information (https://ptacts.uspto.gov/ptacts/public-informations) for case IPR2024-00157; docket mirror with the paper timeline at https://gaeflexstaging-dot-docketupdate.appspot.com/cases/PTAB/IPR2024-00157/Xu_Peicheng_%28Hong_Kong%29_Optoelectronic_Technology_Co._Limited_v._Semisilicon_Technology_Corp/ ; refund paper (Paper 13) at https://gaeflexstaging-dot-docketupdate.appspot.com/cases/PTAB/IPR2024-00157/Xu_Peicheng_%28Hong_Kong%29_Optoelectronic_Technology_Co._Limited_v._Semisilicon_Technology_Corp/09-18-2024-Petitioner/Other__Refund_request-13-Petitioners_Request_for_Refund_of_Post_Institution_Fees/

Ex parte reexamination (Control No. 90/015,32_ — number truncated in sources) — not an AIA trial; included because it is the only claim-level adjudication

• Type: Ex parte reexamination, 35 U.S.C. §§ 301–307 (Central Reexamination Unit — outside PTAB jurisdiction; no APJ panel)
• Filed: request filed 2023-11-17; reexamination ordered 2024-02-15
• Status: Concluded — claims changed and cancelled. Reexamination certificate recorded 2024-12-17 (effective date 2024-12-13 per the record).
• Outcome, verbatim from the certificate entry: "REEXAMINATION CERTIFICATE; THE PATENTABILITY OF CLAIMS 1-15 AND 22 IS CONFIRMED. CLAIMS 16-21 AND 23-24 ARE CANCELLED. NEW CLAIMS 25-43 ARE ADDED AND DETERMINED TO BE PATENTABLE."
• Appeal: No appeal found; the certificate stands.
• Defensive value: This is where the patent actually got wounded. If a demand letter or complaint asserts claims 16, 17, 18, 19, 20, or 21 (the entire pluggable lamp + lampshade + socket + mount assembly family) or claim 23 (the partial-board-encapsulation package), those claims were cancelled — the assertion is baseless as a matter of the patent's current scope, and pressing them invites Rule 11 exposure. The surviving, enforceable-set prior to any new claims is claims 1–15 and 22 only.
• Caveat (important): I could not retrieve the text of new claims 25–43 or the certificate PDF. Those new claims were added after the period covered by the claim set reproduced in the supplied patent text, and their scope is unknown to me. Any infringement analysis must pull the reexamination certificate itself (PTAB/Patent Center "Reexamination Certificate" document) before evaluating the patent's current reach. Do not assume new claims 25–43 are merely the cancelled assembly claims renumbered — that would be guesswork.

Not a proceeding against this patent (context only)

IPR2024-00156 — Xu Peicheng (Hong Kong) Optoelectronic Technology Co., Limited v. Semisilicon Technology Corp. — filed the same day (2023-11-08) by the same petitioner, but against US 8,884,546 ("Driving Apparatus for Light-Emitting Diode Curtain Lamp"), a different Semisilicon patent. Unlike the '988 IPR, this one was instituted on 2024-05-20, went to an oral hearing on 2025-02-24, and a Final Written Decision issued 2025-05-14 (Paper 34). Petitioner identified Shenzhen Yuyixin Lighting Technology Co., Ltd. as a real party in interest — a genuine competitor, not a defensive aggregator. I have not verified what the '546 FWD held and expressly decline to characterize it here, since it does not bind or invalidate anything in the '988 patent. It matters only as a pattern signal: the same petitioner, the same day, the same expert (Dr. Schubert), one win at institution and one loss. Sources: https://www.casemine.com/judgement/us/682897826d97a40d1776b4c9 ; https://gaeflexstaging-dot-docketupdate.appspot.com/cases/PTAB/IPR2024-00156/

Strategic summary

Claim status as of today — this is the whole ballgame.

• CANCELLED by reexamination certificate: claims 16, 17, 18, 19, 20, 21, 23, 24. That removes every independent assembly claim covering the socket/mount pluggable architecture (16 and 21), all their dependents, and the partial-board-encapsulation claim 23.
• SUSTAINED (confirmed patentable) by reexamination: claims 1–15 (the package structure: LEDs with a common electrode, control IC with Din/Dout/Vdd/Vss, four conductors, encapsulating package body, plus the layout/geometry dependents) and claim 22 (the two-mount, board-edge-inserted variant).
• UNTESTED / unknown to me: new claims 25–43, added in the certificate. Their scope is not in any source I could reach, and they have never been construed by any tribunal.
• Never adjudicated by the PTAB: the entire patent. IPR2024-00157 was denied institution, so no claim of the '988 patent has an FWD sitting behind it.

Estoppel landscape. Because IPR2024-00157 ended in a denial of institution, no § 315(e)(2) estoppel arose — estoppel under that provision is triggered by a final written decision, and there was none. Practically:

• The Petitioner (Xu Peicheng) and Shenzhen Yuyixin are not estopped and remain free to file a second IPR on the same or different art (subject only to General Plastic-style follow-on factors and § 315(b) timing if they were served with a complaint).
• You are unconstrained. All of the '988 IPR's art (Lai CN201017882Y, Stoyan, Morgan, Allsop, Yan) and anything better remains available as IPR grounds, IPR-style invalidity contentions, or inter partes reexamination art. Nothing was "spent."
• The flip side: the denied panel (Boudreau, Reagan, McShane) has already seen and rejected the most natural obviousness combination. A fresh IPR that recycles Stoyan + Morgan without addressing why the Board found the showing deficient risks a second denial and a General Plastic discretionary refusal. You need the Paper 9 text before you brief anything — see next steps.
• One asymmetry worth pricing: the ex parte reexamination that killed claims 16–21 and 23–24 is a § 315(e)(1)/(e)(2)-free channel — no estoppel, no Fintiv risk, and, critically, a requester can stay anonymous as to the public record even though the Office knows the identity.

Pattern signals. (1) Not a defensive aggregator — this is a competitor-vs-competitor fight: Xu Peicheng (Hong Kong) Optoelectronic Technology with Shenzhen Yuyixin Lighting as RPI, using Rimon Law and Dr. E. Fred Schubert as expert. (2) Coordinated dual-track attack — an IPR and an ex parte reexamination filed nine days apart (2023-11-08 and 2023-11-17), with the reexam order (2024-02-15) appearing in the IPR record. The reexam delivered the kills the IPR could not. (3) The patent owner pursued rehearing opposition, not appeals — the only appellate-adjacent activity is Petitioner's rehearing request, not any Semisilicon CAFC filing; the patent owner's U.S. enforcement posture lives in district court (Delaware 1:25-cv-01566; Florida S.D. 1:26-cv-23238 and 0:26-cv-61380), per the Google Patents metadata.

Recommended next steps

1. Get Paper 9 (the 2024-05-20 denial) before you do anything else. It is the single most valuable document for anyone attacking this patent: it tells you exactly why the PTAB found the Lai/Stoyan/Morgan/Yan combination insufficient. Order it from PTAB E2E (https://ptacts.uspto.gov/ptacts/public-informations) or via PRPS. Everything below is sequencing around a document I could not read in this session.
2. Notice the cancelled claims in writing, immediately, if they are being asserted. If opposing counsel's letter or complaint pleads claims 16–21 or 23–24, respond citing the reexamination certificate (recorded 2024-12-17, effective 2024-12-13) and its exact disposition text: "THE PATENTABILITY OF CLAIMS 1-15 AND 22 IS CONFIRMED. CLAIMS 16-21 AND 23-24 ARE CANCELLED. NEW CLAIMS 25-43 ARE ADDED." Asserting a cancelled claim is frivolous and supports fees/sanctions and a § 285 motion if the assertion reaches suit.
3. Obtain the reexamination certificate and read new claims 25–43. They are the patent's forward-looking scope and are entirely un-addressed in any public decision I found. Until you have their text, treat the "surviving claims" analysis as provisional. The certificate is available in Patent Center / the PTAB & reexam document set for control no. 90/015,32_.
4. If you file an IPR, treat it as a first IPR, not a follow-on — there is no FWD, so § 315(e)(2) estoppel will later attach to whatever grounds you raise or reasonably could have raised, and you get one clean shot. Budget for a § 314(a) institution decision at ~6 months from filing and, if instituted, a FWD within 12 months of institution (§ 316(a)(11)). Avoid General Plastic exposure by distinguishing Paper 9's reasoning rather than refiling the same combination.
5. Consider the reexamination channel first. It has already produced claim cancellations here, carries no § 315 estoppel and no Fintiv discretionary-denial risk, and lets the attacking art be developed through prosecution-style prosecution at the CRU. Given that the IPR route failed at the threshold in this very case, the reexam track is empirically the stronger one for this patent.
6. Do not cite any FWD for this patent, and do not represent that an IPR invalidated anything. There is one AIA trial, it was denied institution, the post-institution fee was refunded on 2024-09-20, and rehearing was denied on 2024-09-16. If an adversary claims a PTAB "loss" for Semisilicon, the correct response is that the Board never reached the merits — the claim cancellations came from reexamination.

Verification gaps I am flagging rather than papering over: the full text of Paper 9 (denial reasoning); the complete reexamination control number (truncated to "90/015,32_" in my sources); the identity of the reexamination requester; the text and scope of new claims 25–43; and the RPI/privy relationships binding Xu Peicheng to Shenzhen Yuyixin. Each of these should be confirmed from primary USPTO records (PTAB E2E and the reexamination file) before being relied on in a brief.

I'll verify the assignment record and the assignee's corporate status before writing this up.

Let me check the assignee's corporate profile and whether any post-issuance assignment exists.

Let me check the recording attorney of record and whether any other assignments were recorded.

Inventors

Pattern note (negative finding): the classic "all inventors departed the original assignee within 12 months, teeing up a fire-sale" pattern is absent. Peng appears as the named inventor across the Semisilicon family well into the 2020s (e.g., US 10,231,303; US 10,462,866; the 2023–2025 applications on the Onscope/IPqwery owner profile for Semisilicon Technology Corp.). He is also a listed inventor on the design patent US D587,841 S ("Lampstand"), filed 2008-04-10, assigned to Semisilicon — same assignee, same period. No inventor-departure signal.

Original assignee

Semisilicon Technology Corp. (矽誠科技股份有限公司), New Taipei City, Taiwan — named as assignee on the issued patent (address as printed on the face: Jhonghe, Taipei County, TW; the county was renamed New Taipei City in 2010, and the registered address today is still 8F.-2, No. 504, Yuanshan Rd., Zhonghe Dist., New Taipei City 23545).

• Primary line of business: fabless IC design / semiconductor manufacturing and sales — cool-light-sheet driver ICs and LED control ICs (registry business items expressly cover "design, manufacture, processing, buying/selling and import/export of integrated circuits and electronic components"). Taiwan unified number 89448545, incorporated 1995-04-14, paid-in capital NT$27,000,000, ~50-person scale. Registry status 核准設立 (active/approved), last change recorded 2025-07-07, with a factory registration in production status (生產中).
• Did it ship a product embodying the claims? Yes — the specification's own framing is a product/package architecture (addressable LED beads/lamp strings), and Semisilicon sells LED control ICs and LED display module products. Its patents are also cited as background art in competitors' later filings (e.g., EP 4 465 766 B1 cites US 8,124,988 B2 as a reference in the description). Independent product-level proof (a Semisilicon-branded part number reading on claim 1) was not obtained in this session; the finding rests on the company's registered business scope, active factory, and continuing product-family filings.
• Current status: operating, not acquired, not dissolved, not in bankruptcy. It paid maintenance fees on this patent at the 4th (2015-07-10), 8th (2019-07-30) and 12th (2023-07-28) years, and is still filing (e.g., a 2023-priority application published as US 2025/0048515 A1, granted as US 12,707,543 B2 on 2026-08-11 per the INPADOC family data).

Assignment timeline

The record I can document contains exactly one recorded assignment on this patent. There are no post-issuance assignments, no security interests, no releases, no mergers, and no change-of-name filings.

• 2008-05-19 (executed) / recorded 2008-05-28 — Reel 021070 / Frame 0362
  • Conveyance: Assignment — recorded description: "ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS)"
  • Assignor: PENG, JACKY (individual inventor, Taipei County, TW)
  • Assignee: SEMISILICON TECHNOLOGY CORP. (Jhonghe, Taipei County, Taiwan)
  • Correspondent: not recoverable in this session. The Google Patents legal-events entry reproduces the reel/frame, parties and effective date but omits the correspondent/recording-attorney field; the USPTO Assignment Center detail view (which is where that field lives) was not reachable before my tool budget was exhausted. I am not supplying a name here — see "Verification gaps."
  • Context: routine inventor-to-employer assignment of rights, filed on the same date the application was recorded, with an execution date nine days earlier. This is the initial chain-of-title instrument, not an acquisition, sale, securitization or reorg.

Nothing after that. The only other entries in this patent's legal-events history are the grant (2012-02-08 STCF), maintenance-fee payments (2015, 2019, 2023 — these are not assignments), the IPR filing (2023-12-19), the reexamination request (2024-01-30) and the reexamination certificate changing the claims (2024-12-17). None of them moves title.

Because the Assignment Center has a record (the 2008 filing) but nothing thereafter, the operative conclusion is: title has been continuously held by Semisilicon Technology Corp. since 2008-05-28, and the patent has never left the original operating assignee.

Timeline diagram

timeline
    title Ownership of US 8124988
    2008 : Peng executes assignment to Semisilicon
         : Recorded reel 021070 frame 0362
         : Application filed 2008-05-28
    2012 : Patent issues to Semisilicon
    2015 : Maintenance fee paid 4th year
    2019 : Maintenance fee paid 8th year
    2023 : Maintenance fee paid 12th year
         : Competitor IPR filed and denied
    2024 : Reexam cancels claims 16 to 21 and 23 to 24
    2025 : Semisilicon sues inMusic in Delaware
    2026 : Delaware case dismissed
         : Refiled against inMusic in SD Florida

NPE / troll-pattern signals

1. Shell-entity transfer — NOT PRESENT. The single recorded instrument (reel 021070/0362, recorded 2008-05-28) runs toward the operating company, not away from it. There is no assignment of this patent to any "IP / Patents / Licensing / Holdings / Ventures"-suffixed entity, no single-member LLC, and no registered-agent address anywhere in the chain. The assignee is a 1995-incorporated Taiwanese corporation with a factory registration listed as in production and paid-in capital of NT$27,000,000 — the opposite of a shell.

2. Known asserter in the chain — NOT PRESENT. Neither entity in the chain — Jacky Peng or Semisilicon Technology Corp. — matches any entity on the enumerated lists (Acacia, Marathon, Intellectual Ventures, IPNav, Wi-LAN, Mosaid/Conversant, Vringo, Pendrell, Innovatio, MPHJ, Lumen View, Round Rock, Document Generation Corp., Spangenberg entities), and neither surfaces on the Unified Patents / RPX high-frequency-plaintiff directories for this chain. Semisilicon appears in Unified Patents' database as the target of an IPR (IPR2024-00157), which is the inverse of the asserter-directory signal. Caveat worth pricing: a federal court opinion in Semisilicon's own N.D. Ill. case (1:25-cv-07388) records that defendants characterized "Semisilicon's business model" as "licensing its patents," and the court noted Semisilicon's entry into "exclusive license agreement[s]" as evidence of market interest. That is aggressive monetization by an operating owner — but it is not an NPE transfer, and it is a litigant's characterization plus a court's passing observation, not a title record.

3. Repeat correspondent across the chain — UNCLEAR / DATA GAP. With only one recorded assignment (reel 021070/0362), a recurrence test is impossible within this chain, and the correspondent field for that single entry was not exposed in the sources I could reach. I will not name a correspondent. Note that a single appearance would not be a finding anyway under your own standard. This signal should be re-run against the Assignment Center detail view; if the same attorney also filed recordings for other Semisilicon patents (the company holds ~60 US patents), that is worth a note — but only if the name is actually read off the records.

4. Cascading transfers — NOT PRESENT. One assignment in ~18 years. No chained LLCs, no sub-24-month sequence, no shared correspondent addresses (there are no other links to share anything with).

5. Pre-litigation transfer — NOT PRESENT. The only recording is 2008-05-28. The first assertion events are the 2023-12-23 cease-and-desist letter and the 2025 filings (D. Del. 1:25-cv-01566, filed 2025-12-29; S.D. Fla. 1:26-cv-23238, filed 2026). That is a gap of ~17.5 years — no assignment was arranged to enable assertion, and no standing/venue-by-transfer issue exists. This is actually a clean-standing finding in the patentee's favour: it has owned the patent since the filing date.

6. Bankruptcy fire-sale — NOT PRESENT. No Chapter 7/11, no receivership, no patent sale in proceedings. Affirmative counter-evidence: three maintenance-fee payments by Semisilicon (2015-07-10, 2019-07-30, 2023-07-28) and continuing prosecution activity into 2025–2026.

7. Privateering — NOT PRESENT (in the classic sense). No operating company transferred this patent to an NPE to sue on its behalf. The operating company is the plaintiff. Flag the adjacent fact pattern though: Semisilicon runs a broad enforcement program in its own name — 80+ accused ION Audio/inMusic products in the Florida complaint; mass Amazon takedown activity against the "addressable LED bead" category (Chinese investor-relations Q&As from June 2023 referencing "矽诚美国专利8124988"); a cease-and-desist program; and, per the N.D. Ill. opinion, exclusive licensing. That is a manufacturer-led monetization campaign, which is a distinct posture from a shell-NPE campaign but will be argued against it on the eBay/irreparable-harm and § 285 axes.

8. Defensive aggregator (anti-NPE) — NOT PRESENT. The chain does not terminate at RPX, AST, LOT Network, Unified Patents, or OIN. Unified Patents appears in this story only as a data provider and a litigation-portal host; it is not an assignee. The inverse of neutralization has occurred: the patent was attacked (denied IPR; reexamination that cancelled claims 16–21 and 23–24), not purchased and parked.

Verdict

Operating-company assertion.

Justification: Title has never left the original assignee. The only recorded instrument is the founding inventor-to-employer assignment, executed 2008-05-19 and recorded 2008-05-28 at reel 021070/0362, running from Jacky Peng to Semisilicon Technology Corp.; there is no subsequent reel/frame of any kind, so there is no shell-entity transfer, no cascading LLC chain, and no pre-litigation transfer to arrange around the 2025–2026 suits (the gap between the last recording and the first assertion is roughly 17.5 years). The assignee is a Taiwanese IC design/manufacturing company (unified number 89448545, active status, factory registration in production, fees paid through the 12th year in 2023, still filing in 2025) that asserts the patent itself against downstream sellers of competing addressable-LED products — which is operating-company assertion, not NPE conduct. The one honest qualifier is that Semisilicon's enforcement is monetization-weighted (exclusive licensing per the N.D. Ill. record; Amazon takedown program; 80+ accused products), so expect an adversary to press the "licensing business model defeats irreparable harm" argument it already made in Illinois — but that argument does not change the ownership verdict.

Verify at: https://assignment.uspto.gov/patent/index.html (search "8124988") or https://assignmentcenter.uspto.gov/ — the single entry should be reel 021070/0362.

Verification gaps (flagged, not papered over)

1. Correspondent of record on reel 021070/0362 is unknown to me. The Google Patents legal-events block for this patent reproduces the reel/frame and the "ASSIGNMENT OF ASSIGNORS INTEREST" description but not the recording-attorney/correspondent field. I could not open the Assignment Center detail view before exhausting my search budget. This is the one input that would let me run signal #3 properly, and it should be pulled directly.
2. Chain-of-title assumptions. The "no later assignments" conclusion rests on the Google Patents legal-events table plus the absence of any assignment entry in the IPR exhibit set and the litigation records — not on a direct USPTO Assignment Center query. A direct query would also catch any security interest or release that never surfaced on the aggregator.
3. Inventor identity mapping. The equation Jacky Peng = 彭文琦 / Peng Wen-Chi is inferred from matching Taiwan registry data (director of Semisilicon, 矽誠科技), matching Chinese-language filings listing 彭文琦 as inventor for 矽诚科技股份有限公司, and matching US patents (US 10,231,303 "Peng") on the same LED lamp/IP family. It is a strong inference but not a certified identity record.
4. Date inconsistency (carried forward from earlier sections). The task header states today is 2026-09-24, while earlier sections of this same analysis flagged sources dated 2026-05-07 and a header date of 2026-04-26, and the prior session header read 2026-09-13. I have reported dates as the sources give them and have not corrected any of them.
5. No CAFC appeal found — consistent with the earlier section; nothing in the assignment record creates an appellate event.

I'll research each cited reference and the IPR prior art for US 8,124,988, prioritizing official sources.

I hit the tool-call limit before I could pull the full text of the last IPR reference, but I have enough grounded material to answer. Below is the prior-art analysis for US 8,124,988 ("Light emitting diode lamp package structure and assembly thereof," Semisilicon Technology Corp., filed 2008-05-28, issued 2012-02-28).

Two important scope notes before the reference-by-reference analysis

1. Which claims actually matter. Per the post-grant record carried forward from the earlier sections, the December 13/17, 2024 ex parte reexamination certificate is reported to have confirmed claims 1–15 and 22, cancelled claims 16–21 and 23–24, and added new claims 25–43. If that certificate is accurate, the enforceable estate is the package-structure family (claims 1–15), independent claim 22 (the two-mount assembly), and new claims 25–43. That means the cited references should be evaluated primarily against claim 1 (the flagship), its dependents 2–15, and claim 22. I flag again that this certificate content comes from Google Patents legal-events metadata and was not independently verified against the official USPTO certificate.

2. The examiner-cited set is thin, and none of it is a clean §102 anticipation of claim 1. US 8,124,988 has only three examiner-cited references on its face (US7473933B2, US7717596B1, US20080087903A1). All three predate the 2008-05-28 filing, so all are §102(a)/§102(e)-type prior art as to filing/naming. But none discloses the core inventive combination of claim 1 — namely a control IC with a data-input electrode and a data-output electrode, integrated in the same package body as the LEDs, with four conductors exposed as Vdd / Din / Dout / Vss. These are §103 (obviousness) references, not anticipations. The closest thing to a true anticipation is an uncited reference that the PTAB petitioner used — see CN201017882Y below.

Reference-by-reference analysis

A. Examiner-cited references (on the face of US 8,124,988)

A1. US 7,473,933 B2 — Yan (Ledengin, Inc.)

• Full citation: US 7,473,933 B2, "High power LED package with universal bonding pads and interconnect arrangement," inventor Xiantao Yan, assignee Ledengin, Inc. (Cayman).
• Dates: priority 2004-10-29 (provisionals 60/623,266 / 60/623,171 / 60/623,260); application 11/259,842 filed 2005-10-26; published as US 2006/0091416 A1 (2006-05-04); granted 2009-01-06.
• Brief description: A ceramic/thermally-conductive LED package with at least three bonding pads arranged so the package can accept more than one LED type; electrical contacts on the top or bottom surface coupled to the pads by metallized (full or partial) vias; packages for multiple LEDs that can be selectively operable; pads can substantially cover the cavity floor. (Sources: PubChem patent page; Google Patents US7473933; Sumobrain.) It is cited as a primary reference.
• Anticipation assessment under §102: No true anticipation of any claim. The reference is a package/substrate disclosure — pads, vias, exterior contacts, multiple selectively-operable LED dies. It contains no control IC at all, and therefore cannot disclose the "control integrated circuit, having a data input electrode and a data output electrode" limitation that is the heart of claim 1 (and of dependent claim 17 / original claims 16–21). It is nonetheless the most structurally relevant of the three for:
  • Claim 6 ("conductors are an electric conductivity support or a conductive wire") — its exterior contacts / pads / vias are the same genus of conductor structure.
  • Claim 1 in a §103 combination — its multi-LED-in-one-package, selectively-operable, externally-contacted arrangement supplies the "a plurality of light emitting diodes … a package body encapsulating … a portion of the [conductors] exposed outside the package body" elements.
  • Claim 3 (circuit board carrying LEDs) — arguably, via its substrate-with-traces teaching.

A2. US 7,717,596 B1 — Bell

• Full citation: US 7,717,596 B1, "Rearview mirror assembly with running lights," inventor Alan Bell.
• Dates: filed 2005-07-15; granted 2010-05-18.
• Brief description: A vehicle rearview-mirror assembly with an etched graphical image back-lit by a plurality of LEDs attached to a printed circuit board (PCB) along with electronic circuitry. The light sources "may be of more than a single color, intensity and technology," and "electronic circuitry may be used to program how and when the light sources operate to provide effects such as the sequencing of the lights, flashing lights or dimming lights." Circuitry is internal so an external flashing circuit is not required. (Source: Google Patents US7717596; the granted patent PDF at patentimages.)
• Anticipation assessment under §102: No anticipation. Despite touching several claim-1 concepts (LEDs on a PCB, on-board programmable control circuitry, multiple colors), it discloses nothing about a control IC with data-in/data-out electrodes, and nothing about a four-conductor Vdd/Din/Dout/Vss package interface. Its "programmable" circuitry drives sequencing/flashing of its own LEDs but is not a serial-data-addressable control IC having a data output electrode for downstream devices. It is at best §103 art for the "control … to control the illumination state of each light emitting diode" and "circuit board carrying the light emitting diodes" concepts, and for claim 13's multi-color aspect (its "more than a single color" teaching). Its real value to the examiner was almost certainly as an obviousness reference showing on-board programmable lighting control was known.

A3. US 2008/0087903 A1 — Stoyan (Patent-Treuhand-Gesellschaft für elektrische Glühlampen mbH)

• Full citation: US 2008/0087903 A1, "Method for producing a light emitting diode arrangement, and light emitting diode arrangement," inventor Harald Stoyan, applicant Patent-Treuhand-Gesellschaft für elektrische Glühlampen mbH (parent company Optotronic GmbH).
• Dates: priority 2006-09-26/27; application 11/904,170 filed 2007-09-25; published 2008-04-16.
• Brief description: LED modules each comprising at least one radiation-emitting semiconductor component on a carrier body, joined by a separately fabricated connection carrier that makes a mechanically stable and electrically conductive connection between the carrier bodies of two adjacent modules — i.e., a modular, serially-connectable linear LED chain whose spacing is set by the connection-carrier length. (Sources: Unified Patents patent page; Google Patents US20080087903A1; patents-review.com.)
• Anticipation assessment under §102: No anticipation. The reference is about inter-module connection and modular lamp-string assembly — analogous to the lamp-string wiring layout problem the '988 patent's background section describes — but it discloses no control IC inside the module and no data input/output electrodes. It maps most closely onto the '988 patent's string-level claims (original claims 16–22, now largely cancelled), not onto claim 1. It would support §103 argument on the "conductive wire"/serial-interconnect aspect (claim 6) and on the lamp-string assembly concept, but cannot anticipate.

B. IPR2024-00157 references (petitioner Xu Peicheng (Hong Kong) Optoelectronic Technology Co., Ltd.)

These are the references the petitioner relied on. The Board denied institution on the merits (May 20, 2024), which is itself evidence that, even combined, the petitioner's set was not found reasonably likely to render the '988 claims unpatentable. (Sources: Unified Patents IPR2024-00157 portal; the PTAB docket mirror at gaeflexstaging-dot-docketupdate.)

B1. CN 201017882 Y — Lai (赖金鸿 / Lai Chin-Hung) — the single most relevant reference found

• Full citation: CN 201017882 Y, "可程序化全彩LED灯" (Programmable full-color LED lamp), inventor 赖金鸿 (Lai Chin-Hung), application no. CN 200720117937.2.
• Dates: filed 2007-01-08; published/granted 2008-02-06. (Utility model; term later terminated 2013-01-08 for non-payment.)
• Brief description (from the Chinese record): The lamp is built of LED chips, a control IC board, and pins, all encapsulated in a housing (壳体内). The pins include: a power pin (电源脚), a ground pin (接地脚), a data input pin (数据输入脚), and a data output pin (数据输出脚). Claim 2 expressly recites those four pin types. The abstract states the lamp is "structurally compact and can by itself achieve display control," with display color and timing self-controllable, or controllable via an external control device, and can have red, blue and green LED chips packaged together with an internal control circuit for full-color display. (Source: tianyancha patent record, CN201017882Y.)
• Anticipation assessment under §102: This is the strongest §102 candidate for original claim 1 and original claim 23 — but note two things. (i) It was not cited by the examiner; it surfaced only in the IPR. (ii) Even on the petitioner's own ground, the Board did not institute, so the PTAB did not find it reasonably likely to anticipate or render obvious the claims as a whole.
  • Mapping to claim 1: LED chips + control IC in one package body ⇒ "package body encapsulating the light emitting diodes, the control integrated circuit"; power pin / ground pin / data-in pin / data-out pin ⇒ "the first conductor being a positive DC voltage pad … the second conductor being connected to the data input electrode, the third conductor being connected to the data output electrode, the fourth conductor being a negative DC voltage pad." It is a remarkably close textual match and is the reference a defendant is most likely to lead with. Whether it discloses the claim-1 limitation "one of the electrode ends being a common electrode end" for the LEDs (i.e., a common-anode/common-cathode arrangement) and the positive-power-in electrode of the IC being tied to the first conductor is the disputed ground — and likely why institution was denied.
  • Mapping to original claim 23 (partial-board encapsulation): a weaker fit, since Lai's LED chips and IC appear both fully housed.
  • It is prior art under pre-AIA §102(a) as a patent/publication "in this or a foreign country" dated before the 2008-05-28 filing (and would be §102(b)/§102(a) art if relied on for the invention date; its 2008-02-06 publication is under one year before filing).

B2. US 2004/0240890 A1 — Morgan (Color Kinetics)

• Full citation: US 2004/0240890 A1, "Methods and apparatus for controlling devices in a networked lighting system."
• Dates: divisional of Ser. No. 10/158,579 filed 2002-05-30; earliest provisional priority 2001-06-28 (60/301,692); published 2004-12-02.
• Brief description: A networked, computer-controlled lighting system of independently addressable controllers each coupled to at least one LED light source, communicating data serially (see claims 96–109: controllers "arranged to communicate in a serial fashion"; the apparatus "formed as an integrated circuit mounted on a platform," with the platform having a first side carrying the IC and a second side carrying the LED(s); "the illumination control circuit is adapted to independently control the plurality of LEDs"; LEDs adapted to produce red, green and blue). (Sources: Sumobrain US20040240890; Google Patents.)
• Anticipation assessment under §102: No anticipation, but strong §103 art on several limitations:
  • Claim 1 — supplies the "control … receives external control data via the data input electrode to control the illumination state of each light emitting diode" and the serial-data-input/data-output architecture; its claim 106–109 teaching of an IC on one platform side with LEDs on the other side maps directly onto dependent claim 5 ("light emitting diodes mounted on one surface of the circuit board and the control integrated circuit … on the other surface") and aspects of claim 4.
  • Claim 13 (red/green/blue LEDs on opposing board surfaces) — its RGB "at least two different colors … red, green and blue" teaching.
  • It does not disclose the "single package body encapsulating the LEDs and the control IC with four exposed conductors" — that omission is why it cannot anticipate claim 1.

B3. US 2006/0012978 A1 — Allsop

• Full citation: US 2006/0012978 A1 (Allsop et al.), cited in IPR2024-00157 as a ground reference.
• Dates: as an 2006 publication it predates the 2008-05-28 filing and would be §102(a)/(b) art as to the filing date.
• Brief description / assessment: I could not retrieve and verify this reference's text in this session — my search was cut off. I will not characterize its disclosure or name a claim mapping for it, because doing so without the text would be speculation. What I can say from the IPR record is only that the petitioner used it as part of a multi-reference §103 ground, and the Board did not institute. This should be pulled and read directly before relying on it.

C. Reference named in the patent body but not as a citation

The specification's background section itself describes prior-art LED packages (two-conductor single-LED packages; four-conductor red/blue/green packages with a common electrode). These are admissions about the state of the art and are usable as §102(a)/(b) "admitted prior art," particularly against claim 1's "plurality of light emitting diodes … one of the electrode ends being a common electrode end" and "the fourth conductor … negative DC voltage pad" elements. No specific patent number is identified for these admissions in the '988 text.

§102 anticipation — consolidated conclusion

Bottom line: Of everything on the record, CN201017882Y (Lai) is the most relevant prior art to US 8,124,988, because it alone discloses the package-level combination that claim 1 is built around — LED chips and a control IC co-encapsulated in a housing with power, ground, data-input and data-output pins. Nothing in the examiner-cited set (Yan/Ledengin, Bell, Stoyan) comes close to a §102 anticipation; all three are best characterized as §103 obviousness references supplying individual elements (multi-LED packaging, on-board programmable lighting control, modular serial interconnection). US 2004/0240890 (Morgan) adds the addressable-serial-control and opposite-side IC/LED concepts. The remaining IPR reference, US 2006/0012978 (Allsop), I could not verify and it should be read before any mapping is asserted.

Caveats: (1) The claim set I analyzed against (1–15, 22 surviving; 16–21, 23–24 cancelled; 25–43 added) rests on Google Patents metadata for the reexamination certificate, not an independently verified USPTO certificate — verify before relying on it, since it determines which claims are even worth mapping. (2) I did not retrieve the Allsop reference text, and my provenance for the IPR ground list is the Unified Patents portal/PTAB docket mirror, not the PTAB's own Institution Decision PDF, which should be pulled for the exact ground-by-ground claim mapping and the Board's stated reasons for denying institution.

I'll ground the obviousness analysis in the actual prior-art references of record, then build the combinations. Let me pull the key references.

Now let me pull the remaining IPR-grounds references and the family-cited art.

⚠️ Two flags before the analysis itself

1. Date inconsistency (carried forward). This task header says the current date is April 26, 2026; the session header says 2026-09-13; and sources cited in the earlier sections contain events dated May 2026. I have not "corrected" any date; I flag it because the § 103 analysis below is date-sensitive (what qualifies as prior art, and what is still live).
2. The claim set is not the one printed in the patent. Per the reexamination certificate recited in the Google Patents legal-events record (recorded 2024-12-17, effective 2024-12-13), claims 1–15 and 22 were confirmed, claims 16–21 and 23–24 were cancelled, and new claims 25–43 were added. My analysis therefore treats claims 1–15 and 22 as the live § 103 targets, notes 16–21 and 23–24 only as historical/moot, and cannot analyze new claims 25–43 — their text is not in any source I could reach, and I will not guess at their scope.

Obviousness analysis — US 8,124,988 under 35 U.S.C. § 103

1. Governing law and the critical date

The application was filed 2008-05-28 with no earlier priority claim (Google Patents lists the priority date as 2008-05-28). The patent is therefore governed by pre-AIA § 103 and pre-AIA §§ 102(a)/(b)/(e). Pre-AIA § 104 permits proof of inventive acts performed in a WTO member country, so conduct in Taiwan (a WTO member since 2002-01-01) is available to the patentee for antedating purposes — this matters, and I return to it in § 9.

The KSR framework and MPEP 2141–2144 govern: the question is whether the differences between the claimed subject matter and the prior art are such that the subject matter as a whole would have been obvious to a person of ordinary skill, considering (i) the scope and content of the art, (ii) the differences, (iii) the level of ordinary skill, and (iv) objective indicia. KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007).

2. Level of ordinary skill in the art (POSITA)

No more than a bachelor's degree in electrical/mechanical engineering (or equivalent) plus 2–4 years of experience in LED packaging or LED driver design, or a master's degree plus ~1–2 years. The POSITA would be familiar with: RGB/white LED die arrangement and common-anode/common-cathode multi-chip packages; leadframe/lead-pad ("die paddle") and ceramic/PCB substrate packaging; 5050/3528-class SMD LED packages; wire bonding and surface-mount soldering; serial daisy-chain LED control protocols (Din/Dout shift-register or register-based addressable LED driver ICs); and lamp-string/socket hardware for decorative lighting. This is a low-skill, highly incremental, crowded art — the classic environment in which KSR's "predictable variation" rationales do the heavy lifting.

3. Prior art available against the '988 claims, with § 102 status

References that are NOT prior art and must not be used (they appear in the "Families Citing" table and are later than 2008-05-28): GB2467797B (Huang, priority 2009-02-17), US 2012/0187856 A1 (2010), US 9,635,721 B2 (2015), US 8,727,789 B2, JP 5465898 B2 (2009-03-11). They are useful only as evidence of the art's subsequent trajectory, not as § 102/§ 103 art.

4. Ground 1 — Lai as the primary reference (claim 1 and the package-structure family)

This is the strongest single ground, and it is nearly an anticipation.

Claim 1 requires only: (a) plural LEDs each having a common electrode end; (b) a control IC with Din, Dout, positive power-in, negative power-in, electrically connected to the LEDs and receiving external control data via Din to control illumination; (c) four conductors — Vdd tied to both the LED common electrode and the IC's positive power-in, Din, Dout, and Vss tied to the IC's negative power-in; (d) a package body encapsulating the LEDs, IC and conductors with portions exposed. Notably, claim 1 does not require a circuit board (that is dependent claim 3).

Lai's claim 1, verbatim translated, discloses LED chips (2), a control IC board (3) bearing a control IC that drives the LED chips, the LED chips fixed to and electrically connected to that board, all encapsulated in a housing (5), and pins (4) led directly or indirectly from the control IC board out of the housing. Lai's claim 2 expressly recites that the pins comprise a power pin (41), a ground pin (42), a data input pin (43) and a data output pin (44). Lai's abstract adds that R, B and G chips are packaged together and the internal control circuit drives full-color display, and that the lamp can be externally controlled or have control programs loaded into its own control IC.

Lai therefore teaches the whole inventive concept the '988 patent claims as its contribution. Two elements warrant scrutiny:

• "Common electrode end." Lai does not use that phrase. But a package with exactly four external pins (power, ground, data-in, data-out) that independently drives three color channels must return the LED dies through a shared node; the 4-pin budget is exhausted by power/ground/data-in/data-out. That is at minimum an inherency argument, and at worst a § 103 argument: given a fixed four-pin budget, tying the three dies' common electrode to the power (or ground) pin is a design necessity, not an inventive act. The '988 patent's own Background removes any doubt: it admits as prior art an LED package that "encapsulates simultaneously red, blue and green LED chips and four electric conductivity supports," where "[o]ne of the electric conductivity supports serves as a common electrode end for the red, blue and green LED chips." The patentee admitted the four-support common-electrode RGB package was old; the only delta over the admitted art is the on-board controller with data input/output pins — which is precisely Lai's claim 2.
• "Vdd connected to the common electrode and the positive power-in." Lai's power pin (41) is drawn from the control IC board that both powers the IC and (necessarily) the dies. Claim 2 of the '988 patent itself says the common electrode "is a positive DC voltage pad or a negative DC voltage pad," confirming the patentee regarded the polarity as a mere design choice — which is itself a KSR admission of obviousness ("simple substitution of one known element for another").

Dependent claims 2–15 add only routine geometries and packaging options, all of which Lai, Yan and Stoyan supply or which are predictable variations:

Motivation to combine (Lai as base + Yan/Stoyan as secondary). Same field of endeavor (LED lighting and LED packaging), same problem (delivering addressable, programmable multi-color LED packages that are manufacturable in standard SMD/through-hole form factors). Lai supplies the addressable controller, the four-pin set, and the encapsulation; Yan supplies the multi-die package architecture that routes multiple dies and pads to external contacts through vias at edges/corners with the stated Lumens/area objective; Stoyan supplies the modular carrier/edge-terminal construction that makes the packages joinable into strings with freely chosen spacing. Combining them is "the mere arrangement of old elements with each performing the same function it had been known to perform" and yields nothing more than predictable, expected results — the KSR formulation.

5. Ground 2 — Lai + Allsop (+ Yan) for the assembly claims (16–21, 22)

The IPR2024-00157 petition charted Allsop (US 2006/0012978) element-by-element against claim 16, and the excerpt from the petition's slip opinion is unusually explicit: Allsop's mount plate 106 with apertures 104 was mapped to the "socket … having a perforation," the lighting element seat 80 to the "plug-in member," the support receiver sleeve 78 to the "mount formed with a socket," the diffuser 70 to the "lampshade," and the leads 102 to the "four electric conductivity supports." The petition also argued inherency for the four electrode contacts (the leads must connect to conductors that deliver current).

Put together:

• Allsop = the pluggable lighting element + socket/mount hardware architecture: an element with four leads passing through apertures in a mount plate that seats into a receiving sleeve, with a removable diffuser over it. That is claims 16–21's mechanical core almost entirely, and Allsop is § 102(b) art (published 2006-01-19) — not antedatable.
• Lai = the self-controlled four-pin programmable LED element that Allsop's socket was not designed for; combining gives the "addressable lamp in a serviceable socket" of claim 16/17/21.
• Yan = the multi-die package with selectively operable dies and external contacts (support for claim 17's Vdd/Din/Dout/Vss framing).

Motivation. Both references address the same practical problem — building, wiring and servicing strings/arrays of LED light elements. Allsop expressly emphasizes a replaceable, seated lighting element; Stoyan expressly emphasizes modularity ("different light emitting diode arrangements … constructed in a simple and cost-effective manner from components which are preferably standardized"). The '988 patent's own stated object ("a LED lamp assembly that facilitates the mount/dismount of LED lamps and the wiring layout of LED lamp string") is the same objective the references already pursue, which is textbook articulated-motivation territory (MPEP 2144.01; KSR "design incentives … market pressures" and "obvious to try").

Notably, the claims the reexamination cancelled (16–21) are exactly the family the Allsop-based charting targeted, and the claim the reexamination confirmed (22) is the two-mount/edge-insertion variant that the Allsop "single socket + mount plate" mapping fits less cleanly. That correlation is consistent with an Allsop-driven rejection of 16–21 — but it is an inference, not something I verified against the reexamination file.

Claim 22 (confirmed, still live) is the harder one: it requires two mounts whose sockets couple to two opposing sides of the LED lamp's circuit board, with the lampshade between the mounts. Allsop gives one mount/socket on leads, not a board-edge card in two sockets. The clean combination is Stoyan + Allsop + Lai:

• Stoyan teaches joining adjacent modules by a separately fabricated connection carrier with terminals formed in opposing edge regions of the carrier bodies, expressly to allow arbitrary spacing and simple reconfiguration, and expressly contemplates series connection of modules.
• Allsop teaches the socketed, replaceable element with a shroud (diffuser) over it.
• Lai teaches the self-controlled four-pin LED lamp whose controller lives on the board.

A POSITA seeking a serviceable addressable lamp string would (i) take Lai's addressable 4-pin lamp, (ii) mount it as an edge-insertable card with Din on one edge and Dout on the other (the natural way to make the data line run through the lamp rather than around it), and (iii) socket it at each edge using Stoyan's edge-terminal/connection-carrier teaching and Allsop's socket geometry — putting the lampshade between the two mounts, exactly as the string architecture demands. Motivation: bidirectional current flow through the card and the ability to replace a single lamp without dismantling the string (the patent's own stated advantage). Reasonable expectation of success: high — soldering/edge-connector art for daughtercards is mature and the electrical function is unchanged.

6. Ground 3 — Lai + Stoyan + Yan for cancelled claims 23–24 (historical)

Claim 23 (encapsulate the LEDs and only part of the board; the exposed board portion carries the IC and the four "conduct electricity" regions) is the inverse-partition of Stoyan's own architecture — a carrier body with a component encapsulated while edge-region terminals remain exposed for interconnection — plus Yan's "external electrical contacts disposed on an exterior surface of the substrate" connected by electrical paths/vias. Lai supplies the embedded controller and four pins. Motivation: to protect the wire bonds and dies while keeping the connector footprint available for SMD/reflow or edge-socket mating.

Drafting note (relevant to scope, though the claim is dead): claim 23's "conduct electricity" language and claim 24's recitation that "said second conduct electricity [connects] said data output electrode, [and] said third conduct electricity [connects] said data output electrode" duplicate Din/Dout — almost certainly a drafting error, but I am not auto-correcting it. Claim 19 has a parallel antecedent defect ("the upper half of the socket accommodates the body, while the lower half of the body is coupled to the plug-in member"). Likewise, the specification's summary uses "electric conductivity supports" while claim 1 uses "conductors" — a § 112 ¶ 2 / ¶ 1 vulnerability worth pressing if claim 1 is asserted.

7. Teaching away, and the counterarguments to expect

• No teaching away. No reference disparages an encapsulated controller with data-in/data-out, the four-pin scheme, or a socketed lamp string. The '988 Background's own narrative — that prior-art lamps "lack data input and output pins" and that "the design of such controller becomes more complicated" as string length grows — is an articulated problem statement, and Lai's January 2007 filing is a disclosed solution to that exact problem. A problem identified in the specification is fair game as motivation art (MPEP 2144, In re Kahn / KSR).
• "But the controller is inside the package" — Lai's control IC is on a board sealed inside the housing, and its pins are the only electrical interface. That is the claimed structure.
• "Secondary references are non-analogous" — weak. Lai, Stoyan, Yan and Allsop are all LED lighting/packaging art; Allsop is a lighting device and is reasonably pertinent to the socketing/replacement problem, which is the same problem the '988 patent addresses. Bell (vehicle mirror lighting) is the most peripheral, and it adds little.
• "Hindsight" — countered by the fact that both structurally critical references are § 102(b) art (Yan, published 2006-05-04; Allsop, published 2006-01-19) and Lai was on file in January 2007, and by the explicit, pre-existing design incentives quoted from Yan (Lumens/area) and Stoyan (modularity/standardized components).

8. Objective indicia (secondary considerations)

I found no evidence in the available record of long-felt-but-unsolved need, unexpected results, industry praise, copying, or licensing "because of" validity. To the contrary: (a) the '988 filing is only ~16 months after Lai's January 2007 filing, which undercuts a long-felt-need narrative; (b) the patentee's own Background frames the problem as an incremental unmet need already being addressed by others; and (c) any commercial-success argument resting on the enormous volume of addressable LED beads sold bearing this patent family's designs would face a nexus problem, because the commercial success is at least equally attributable to the general market for addressable LEDs. The reexamination confirmation of claims 1–15 and 22 is not a secondary consideration — it is the outcome of an examination over a closed art record; it carries no weight under § 103 in litigation, and it cannot be used to bootstrap validity.

9. The honest weaknesses in these grounds

1. Lai's § 102(a) status depends on antedating. Lai published 2008-02-06, only 3 months and 22 days before the '988 filing date. It is not § 102(b) art. Semisilicon is a Taiwan entity, and pre-AIA § 104 allows proof of inventive acts in a WTO member country (Taiwan since 2002), so the patentee could attempt to sweep behind Lai with documented prior conception/reduction to practice (corroboration required). If the patentee can, the entire Ground 1 chain collapses and Lai must be dropped as the primary reference — leaving Yan, Allsop, Stoyan and Bell, none of which discloses an in-package addressable controller with Din/Dout. Everything hinges on this.
2. The Board denied institution in IPR2024-00157 on the merits (2024-05-20; rehearing denied 2024-09-16; the post-institution fee was refunded 2024-09-20). I could not obtain Paper 9 and therefore cannot say which limitation(s) defeated the petition or which reference was the weak link. Possible explanations include: (a) the panel accepted an antedating/inherency argument on Lai; (b) the panel found the translation did not clearly show the "common electrode" coupling; (c) the panel found the obviousness rationale conclusory. Do not brief a new challenge built on recycled petition charts without reading Paper 9 — the panel is now on record as unpersuaded by the Lai/Stoyan/Morgan/Allsop/Yan combination as presented.
3. Morgan (US 2004/0240890) is unverified. I could not retrieve its disclosure and do not rely on it.
4. Claim 8 (fourth conductor forming a platform carrying the control IC) is the thinnest of the dependent claims in my verified set; a dedicated leadframe die-paddle reference should be added.
5. No estoppel, but a real General Plastic problem. Because the IPR ended in denial of institution, no § 315(e)(2) estoppel arose — the art was not "spent." But a follow-on IPR recycling the same combination invites discretionary denial.
6. New claims 25–43 are a black box. They were drafted after the art of record did its work, and their scope is unknown to me. Anyone evaluating validity must first obtain the reexamination certificate.

10. Bottom line

Claim 1 is obvious, and very nearly anticipated, over CN201017882Y (Lai) alone; claims 3–7 and 9–15 are obvious over Lai + Yan (US 7,473,933), with Stoyan (US 2008/0087903) supplying the edge-terminal modular-string teachings for claims 9–13; and claim 22 is obvious over Lai + Stoyan + Allsop (US 2006/0012978). The cancelled assembly claims 16–21 rest on Allsop + Lai (+ Yan) — the ground the IPR petition actually charted — and cancelled claims 23–24 on Lai + Stoyan + Yan.

The two decisive facts cutting for these grounds are (i) the examiner-cited, incontestably § 102(b) Yan reference, with its verbatim Lumens/area design incentive, and (ii) Allsop § 102(b), whose four-lead/apertured-mount-plate/receiving-sleeve architecture the IPR petition mapped onto claim 16 nearly element-for-element. The decisive fact cutting against them is that the single reference that supplies claim 1's actual point of novelty — an in-package control IC with data-in and data-out pins — is Lai (CN201017882Y), which is antedatable § 102(a) art, and the PTAB already declined to institute on the combination that used it. Get Paper 9 from PTAB E2E (https://ptacts.uspto.gov/ptacts/public-informations, case IPR2024-00157) and the reexamination certificate (control no. 90/015,32_; recorded 2024-12-17) before relying on any of the above in a brief — those two documents would confirm, refute, or narrow every conclusion in §§ 4–6.

Sources: https://patents.google.com/patent/US8124988/en · https://m.tianyancha.com/zhuanli/8cafa47bfd6c489282d53b0e3cf814a6 (CN201017882Y) · https://patents.google.com/patent/US20080087903A1 · https://portal.unifiedpatents.com/patents/patent/US-20080087903-A1 · https://patents.google.com/patent/US7473933 · https://pubchem.ncbi.nlm.nih.gov/patent/US-[7473933](/patent/7473933)-B2 · https://www.patents-review.com/a/20060091416-high-power-led-package-universal-bonding-pads-interconnect.html · https://www.sumobrain.com/patents/usapp/Offset-solar-powered-outdoor-lighting/20060012978.html · https://ptacts.uspto.gov/ptacts/public-informations/petitions/[1554859](/patent/1554859)/download-documents (IPR2024-00157 petition, Allsop charting) · https://portal.unifiedpatents.com/ptab/case/IPR2024-00157