US 6383668

I'll verify against live sources, then summarize.

Search results and scope note

I searched for US 6,383,668 and for CAFC 2026 docket activity tied to that number. I found no CAFC 2026 appeal, PTAB proceeding, or district-court litigation involving US 6,383,668. The only hits keyed to "6383668" were (a) the patent itself and its family members, and (b) prior-art search strings inside unrelated USPTO office actions/reexamination papers where '668 was cited as a reference. This is consistent with the patent having lapsed in 2010 for non-payment of maintenance fees — an expired, never-litigated patent would not plausibly be on a 2026 CAFC docket. I am stating this rather than implying certainty: absence of a hit in web search is not proof of absence from PACER/CAFC records, which I cannot query directly.

Also note: the requested ID "6383668" (7 digits) maps cleanly and uniquely to US 6,383,668 B1. No similar-but-different number was substituted.

Bibliographic summary — US 6,383,668 B1

Source: https://patents.google.com/patent/US6383668/en (authoritative text supplied); corroborated at https://insight.rpxcorp.com/patent/[US6383668B1](/patent/US6383668B1)

Abstract (as given)

A magnetic recording disk has a magnetic recording layer formed on a special multilayered "host" layer. The host layer is a "synthetic antiferromagnet," i.e., at least two ferromagnetic films exchange-coupled antiferromagnetically (AF) to one another across a nonferromagnetic spacer film so that their magnetic moments are oriented antiparallel. The magnetic recording layer has a different composition from the top ferromagnetic film in the host layer and is ferromagnetically coupled to that top film. The magnetic volume V of the composite structure (recording layer + host layer) that determines thermal stability is approximately the sum of the grain volumes in the recording layer and the AF-coupled host films. However, the magnetic moment of the composite is primarily just that of the recording layer, because the host layer is designed to have essentially no net magnetic moment.

Plain-language claim overview

There are 5 claims: one independent (claim 1) and four dependent (claims 2–5).

Claim 1 — the disk (independent). A magnetic recording disk built up in this order:

1. Substrate, then
2. a nonferromagnetic underlayer, then
3. a "host" layer consisting of:
   • a first cobalt-alloy ferromagnetic film (thickness t1, magnetization M1),
   • a nonferromagnetic spacer film in contact with it, limited to Ru, Cr, Rh, Ir, Cu, or their alloys,
   • a second cobalt-alloy ferromagnetic film in contact with the spacer (thickness t2, magnetization M2),
   • the spacer being thick enough to antiferromagnetically exchange-couple the two films across it, and
   • the two films' magnetic moments per unit area (M1×t1) and (M2×t2) must be different from one another — i.e., the host is not compensated;
4. a magnetic recording layer on and in contact with the second ferromagnetic film, of a different composition from that film and ferromagnetically coupled to it; and
5. a protective overcoat on the recording layer.

Claim 2. Same host-film material for both films, but t1 ≠ t2 — the way to unbalance the moments while keeping the material identical.

Claim 3. Different materials for the two host films, with t1 and t2 substantially equal — the alternative route to differing moments per unit area.

Claim 4. The first (lower) host ferromagnetic film includes an interface film consisting essentially of cobalt in contact with the spacer film (thin Co layer to boost interfacial moment / strengthen AF coupling).

Claim 5. Same as claim 4, but the cobalt interface film is in the second (upper) host ferromagnetic film, contacting the spacer.

Analyst observations worth flagging

• The granted independent claim does not cover the embodiment the specification calls "preferred." The detailed description and abstract emphasize the compensated case (M1t1 ≈ M2t2, essentially zero net moment of the host layer). But claim 1 expressly requires "(M1×t1) and (M2×t2) … are different from one another," which is the specification's alternative embodiment (nonzero net host moment, Mr1t1 ≠ Mr2t2). A literal claim read gives the alternative case, not the compensated case. This is a claim-scope/vs.-disclosure mismatch a litigator or reexam analyst would immediately note.
• "Cobalt alloy" is baked into claim 1 for both host films, even though the specification's general teachings say the host films "include ferromagnetic films made of Co, Fe, Ni, and their alloys" and that films 32/34 "may be a CoCr alloy or a CoPtCr alloy." The claim is narrower than the disclosed genus.
• The spacer list in claim 1 is closed-ended ("selected from the group consisting of Ru, Cr, Rh, Ir, Cu, and their alloys"), mirroring the Parkin et al. material set.
• Layer-sequence limitations matter: claim 1 recites "on and in contact with" for the spacer/film interfaces and for the recording layer/second film interface. The specification also says the recording layer's composition "must be different from the composition of film 32," which claim 1 captures.
• Small internal inconsistency in the specification's figure labels: the description refers to "underlayer 23" in the FIG. 2 discussion (the FIG. 1 discussion uses underlayer 13). Likely a typographical/figure-numbering artifact; no claim depends on it.
• Commercial/prosecution context: the patent is a sibling of the earlier IBM filing described in its own background as Ser. No. 09/416,364, "now U.S. Pat. No. 6,280,813" (AFC films as the recording layer). The '668 patent's stated point of novelty over that sibling is architectural: the AFC "synthetic antiferromagnet" serves as an underlayer/nucleation host beneath a conventional granular recording layer, rather than being the recording layer itself.
• Where it fits historically: this is an early-2000s AFC media family (IBM → Hitachi GST → HGST). Its descendant art appears in the "Cited By" lists (e.g., Seagate's US 6,759,149 "Laminated medium with antiferromagnetic stabilization layers," and later Seagate/HGST AFC filings), confirming its role as foundational rather than litigated art.

Uncertainty / limits of this summary

• The bibliographic data above is taken from the supplied Google Patents text (authoritative per my operating rules) and is corroborated by the RPX Insight record. Where Google Patents and other aggregators can differ (e.g., exact "current assignee" string, legal-status phrasing), treat the reassignment events in the Legal Events section as the more granular record: Mariana HDD B.V. → Hitachi Global Storage Technologies Netherlands B.V., with HGST Netherlands B.V. shown as the present listed owner.
• I have no verified USPTO PatentCenter/PACER/CAFC query result in hand; my "no 2026 CAFC activity found" statement rests on web search only and should be confirmed against PACER/CAFC if this question is load-bearing.
• The claim characterizations above are plain-language paraphrases, not verbatim claim text; for any legal purpose, the literal claim language as issued (5 claims, reproduced in full in the supplied patent text) controls.

I'll research litigation involving US patent 6383668 across patent litigation databases.

Let me search more targeted patent litigation resources.

Let me search additional litigation-specific databases.

Litigation Search: U.S. Patent No. 6,383,668

Bottom line: I found no record of any litigation (or PTAB post-grant proceeding) involving U.S. Patent No. 6,383,668. I searched for the exact number (both "6383668" and "6,383,668") in combination with litigation terms and on litigation-oriented sources, and no complaint, case number, docket, or proceeding tied to this specific patent surfaced.

What the patent is

For context, US 6,383,668 ("Magnetic recording media with antiferromagnetically coupled host layer for the magnetic recording layer") is:

• Inventors: Eric Edward Fullerton, David Thomas Margulies, Ernesto Esteban Marinero, Manfred Ernst Schabes
• Original assignee: International Business Machines Corp. (IBM)
• Filing/priority date: March 27, 2000
• Grant date: May 7, 2002
• Current/ultimate assignee of record: HGST Netherlands B.V. (via Mariana HDD B.V. → Hitachi Global Storage Technologies Netherlands B.V.)
• Status: Expired – Fee Related (lapsed 2010 for nonpayment of maintenance fees; anticipated expiration March 27, 2020)

Search results — what I actually observed

• No hit on Unified Patents' litigation/PTAB portals, District Court docket aggregators, or the Federal Circuit/PACER results tied to '668 specifically.
• The only near-hits were different patents that should not be conflated with '668:
  • U.S. 6,557,054 (the "'366 patent" referenced in a CBM2015-00005 prosecution-history exhibit) — an unrelated software-distribution patent, not a Hitachi media patent.
  • U.S. 6,280,813 and related family members (Fullerton media patents) — occasionally appear in patent-family/citation listings, but that is citation/related-art activity, not litigation.
  • The PTAB search-result tables showing "6383668" are prior-art search logs (patentability search strings listing '668 among many other patents), not litigation or IPR records.
• Sibling patents in the same technology family (e.g., 6,725,450; 6,835,476; 7,125,616; 7,555,870) appear in patent-listing and forward-citation databases, but I found no enforcement action against '668.

Caveats

• I was unable to complete additional targeted queries (I hit a tool step limit before finishing third-party docket searches), so this is not an exhaustive clearance-style search.
• A definitive answer requires a paid docket search (e.g., Lex Machina, Docket Navigator, RPX Insight, PACER/ Bloomberg Law). Free aggregators do not reliably index every short-lived or sealed HDD-industry case. Notably, the disk-drive media patents were typically licensed/cross-licensed among IBM/Hitachi, Seagate, WD, and Fujitsu rather than litigated, and this patent expired in 2010 for failure to pay maintenance fees, which strongly reduces any litigation incentive.

Conclusion: Based on the searches performed, there is no known litigation involving US 6,383,668. I cannot certify this as a complete negative without a commercial docket database review.

I'll verify the structured data's "no proceedings" signal against public sources before finalizing.

The results are dominated by a different '668 patent (US 7,224,668, the Cisco v. Arista switch patent) and a US 6,668,xxx image patent — not our magnetic-media patent. Let me run tighter queries to confirm there's no proceeding on US 6,383,668 specifically.

Confirmed the '668 hits are US 7,224,668 (Cisco/Arista) — a different patent. One final check for any reexam or litigation touching US 6,383,668.

PTAB / AIA Trial Report — U.S. Patent No. 6,383,668 B1

"Magnetic recording media with antiferromagnetically coupled host layer for the magnetic recording layer"
Inventors: Fullerton, Margulies, Marinero, Schabes · Original assignee: IBM · Now: HGST Netherlands B.V. · Filed 2000-03-27 · Granted 2002-05-07

Proceedings overview

Total AIA trial proceedings on file: 0 — zero active, zero with claims invalidated, zero with claims sustained, zero settled, and zero institution denials, because no IPR, PGR, or CBM petition has ever been filed against US 6,383,668. The USPTO Open Data Portal returns no AIA trial proceedings for this patent, and my independent search of public PTAB and Federal Circuit material surfaced none. The defensive posture is therefore not "the patent has survived IPRs and is hardened" — it is the opposite and stronger position: the patent lapsed on 2010-05-07 for failure to pay maintenance fees, and in any event its 20-year term from the 2000-03-27 filing date would have run out on 2020-03-27 (Google Patents lists "anticipated expiration 2020-03-27"). There is no PTAB record to fight about, and no live patent right to assert. Claims 1–5 are all UNTESTED — and unenforceable.

Proceedings

None — no AIA trial proceeding exists on US 6,383,668

There are no proceeding numbers to report. Under the operating rule that identifiers are to be taken literally and not auto-corrected, I am not supplying a docket number here: to do so would be fabrication.

• Type: N/A
• Filed: N/A
• Status: N/A (USPTO ODP: no AIA trials; Google Patents "Legal Events" for this patent show only prosecution, three assignments, and a maintenance-fee lapse — no PTAB trial events)
• Judge panel: N/A
• Petition grounds: N/A
• Institution decision: N/A
• Final Written Decision: N/A
• Settlement / termination: N/A
• Appeal: None. No FWD exists to appeal.
• Defensive value: Everything turns on status, not on PTAB outcomes. Per the patent's legal-events record: a 4-year maintenance fee was paid 2005-11-07; a maintenance-fee reminder mailed 2009-12-14; lapse for failure to pay 2010-05-07; status entered as "PATENT EXPIRED DUE TO NONPAYMENT OF MAINTENANCE FEES UNDER 37 CFR 1.362" on 2010-06-07; and the record shows status "Expired - Fee Related." A demand letter invoking claims 1–5 of this patent is asserting an expired, fee-lapsed patent. Absent a grantable petition to revive the lapse (and no revival appears of record), there is no enforceable right to litigate.

De-confliction note — three "the '668 patent" false positives you must not cite

Search results for "the '668 patent" and "6,383,668" are heavily polluted by unrelated patents. If you are briefing or responding to a demand letter, verify the number before you rely on any of the following:

1. US 7,224,668 B1 (Cisco Systems) — the "'668 patent" in Arista Networks v. Cisco Systems, ITC Inv. No. 337-TA-945. This is the patent actually litigated at the PTAB. Arista filed the petition; the Board instituted 2016-06-11, and the FWD issued 2017-06-01 in IPR2016-00309, finding claims 1–10, 12, 13, 15–28, 30, 33–36, 55–64, 66, 67, and 69–72 unpatentable (per ITC Pub. 4910 and the Commission's 2017-07-20 notice). Cisco's notice of appeal and the Fed. Cir. Appeal No. 17-2384 summary affirmance of 2018-02-14 (mandate 2018-03-23) also belong to that patent. Cisco also filed IPR2016-00303 (US 6,377,577) and IPR2016-00306/00308. None of this touches US 6,383,668.
   • Primary source: USITC Pub. 4910 · FWD-related notice, 83 FR 47351 · IPR2016-00309 Cisco Notice of Appeal
2. US 6,668,xxx ("the '668 patent," image processing) — the MPV/GE indefiniteness litigation (S.D. Cal. 3:18-cv-01158). Different patent, different art unit.
3. US 6,383,668 as prior art only — our patent appears inside later filings solely as a cited reference. Example: it is listed in the prior-art string of petitions in the Dieter Suess "multilayer exchange spring recording media" family, where a search ran ("20070292720") or ("5583727") or ("6280813") or ("6383668") or ("6468670") or ("7550210")).PN, and it is cited as background in US 11,133,031 ("6,280,813 and 6,383,668 addressed the thermal instability problem…"). Being cited as prior art is not a proceeding against the patent.

Bottom line on the false positives: every "IPR" hit for a "'668 patent" in the public record belongs to Cisco's US 7,224,668. Do not attribute IPR2016-00309 or Fed. Cir. 17-2384 to US 6,383,668.

Strategic summary

Claim status: all five claims UNTESTED; none CANCELED, none SUSTAINED. Claims 1–5 of US 6,383,668 have never been before the Board. Claim 1 is the sole independent claim and recites the substrate / nonferromagnetic underlayer / AF-coupled host layer with M₁×t₁ ≠ M₂×t₂ / dissimilar-composition recording layer / overcoat stack; claims 2–5 depend from it (same-material-different-thickness, different-material-same-thickness, and Co interface films on the first or second host ferromagnetic film). Because the patent lapsed in 2010, there is no residual narrowing to trace and no surviving claim set to avoid — the practical answer is that all of claims 1–5 are expired, not narrowed.

Estoppel landscape: essentially clean, and moot. § 315(e)(2) estoppel attaches only to a petitioner that obtains an FWD, and there is no petitioner here — so no party is estopped from any ground on this patent. In principle every § 102/§ 103 ground is "still available"; in practice an IPR is a dead letter because the patent is expired and fee-lapsed, so there is no claim to cancel and no plausible forward-looking exclusionary relief to defeat. Note also that the pre-AIA inter partes reexamination estoppel (§ 315(c)) cited in contemporaneous case law would only matter if someone had filed one — the record shows no reexamination of this patent either (no "Reexamination Certificate," no 90/ control number in the legal events).

Pattern signals: none. No petitioner filed multiple IPRs against this patent (there is no petitioner at all). The patent owner — IBM, then Mariana HDD B.V., then Hitachi Global Storage Technologies Netherlands B.V. (now HGST) — never had occasion to defend a PTAB challenge or pursue a PTAB appeal on this patent. No defensive aggregator (Unified Patents or similar) appears anywhere in the chain. The only "activity" in the public record is (a) the Mrt/thermal-stability patent family around it — notably sibling US 6,280,813 (Ser. No. 09/416,364), the AF-coupled recording layer patent — and (b) decades-later prior-art citations to 6,383,668 in perpendicular/exchange-spring media patents. The absence of IPR activity here is explained by the patent's short commercial enforcement life, not by a hardened claim set.

Recommended next steps

• If you received a demand letter citing US 6,383,668, claims 1–5: lead with status, not validity. The legal-events record shows the patent expired for nonpayment of maintenance fees effective 2010-05-07 (37 CFR 1.362), publication date 2002-05-07, anticipated expiration 2020-03-27. Verify current USPTO Patent Center status before responding, and demand the sender produce any petition to revive and its grant — none appears on the face of the file. Source of record: US6383668B1 on Google Patents.
• Confirm the number on the face of the letter. If the letter actually says 7,224,668, you have a genuine PTAB history to work with and a very different analysis — that patent's claims were held unpatentable in IPR2016-00309 (FWD 2017-06-01), affirmed at Fed. Cir. 17-2384 (2018-02-14). Do not import that outcome into 6,383,668, and do not let opposing counsel import 6,383,668's prior-art citations into a 7,224,668 dispute.
• Run your own docket sweep before relying on this memo. Re-query PTAB E2E / the ODP API directly by patent number (register a free key at https://data.uspto.gov/apis/getting-started), search the PTAB decisions portal at https://www.uspto.gov/patents/ptab/decisions, and pull the patent's full file wrapper and assignment history. My search found no proceeding, but the file is old enough that a stray pre-AIA 90/ reexamination or an unnoticed revival petition would not have surfaced in the web queries I could run here.
• No trial-stage milestones to track. With zero petitions filed, there is no institution-decision deadline, no oral hearing, and no § 316(a)(11) one-year FWD due date on this patent.

Honesty caveat: I found no PTAB proceeding against US 6,383,668 and I have not fabricated one to fill the template. The "no PTAB activity" signal here is the structured-data default and the search result — and in this instance the absence is well explained: the patent lapsed years before post-grant practice matured, so it never attracted a petition.

I'll reconstruct the assignment record. Let me verify the recorded assignments and correspondents via the USPTO Assignment Center and related sources.

Let me attempt to retrieve the specific recorded assignment documents and their correspondents.

Assignment & Ownership Chain — US 6,383,668 B1

Patent: Magnetic recording media with antiferromagnetically coupled host layer for the magnetic recording layer
App. No.: 09/536,868 · Filed: 2000-03-27 · Granted: 2002-05-07
Primary source for the chain below: USPTO assignment records as mirrored in Google Patents Legal Events (https://patents.google.com/patent/US6383668/en#legal-events), verified against the USPTO Patent Assignment Center (https://assignmentcenter.uspto.gov/ ; legacy index https://assignment.uspto.gov/patent/index.html).

Data-quality note up front: the USPTO Assignment Center mirror exposes three recorded events for this patent (reel/frame, conveyance type, assignor, assignee, execution and recording dates). I was unable to retrieve the correspondent-of-record field (the filing attorney/firm) for any of the three records through the sources available to me — the searchable mirrors do not surface it, and the Assignment Center's per-record detail view was not reachable via my tools. Any statement below about correspondents is therefore flagged unclear / not retrieved, and I have not guessed at names.

Inventors

• All four are the assignors named on the first recorded assignment (Reel 010712/0885), which conveys their interest to International Business Machines Corporation. That assignment itself is the documentary evidence of common employment — the inventors assigned to IBM, and IBM is the named original assignee.
• Pattern check — no fire-sale tell. The task calls out "all inventors departing the original assignee within 12 months of filing" as a precursor to a portfolio sale. That is not present here. The inventors stayed with IBM through and past the 2000 filing; the relevant exit is the entire HDD division being divested to Hitachi effective 2002-12-31 — an organizational transfer, not individual inventor attrition. Publicly documented careers are consistent with this (the inventors moved with the storage business; e.g., Fullerton's well-known path IBM → Hitachi Global Storage Technologies → academia later in the decade), but the assignment record itself only confirms the inventors→IBM link.

Original assignee

• Entity on the issued patent: International Business Machines Corporation ("IBM"), New York (Armonk, NY).
• Line of business: diversified computing/hardware/software/services; relevant division was its Hard Disk Drive (HDD) business (Deskstar/Travelstar storage products), based around the Almaden Research Center in San Jose.
• Did it ship a product embodying the claims? Yes — in the substantive sense. IBM commercialized antiferromagnetically-coupled ("AFC," marketed as "Pixie Dust") granular Co-alloy media in its desktop and mobile drives in the early 2000s. This patent is the host-layer variant of that AFC media family, the sibling of the recording-layer patent US 6,280,813 (the '813 patent appears as a family citation/related document here). Note the caveat: the asserted/commercial AFC product line maps most directly onto the AFC recording-layer claims; whether any shipped SKU practiced the specific "host layer beneath a compositionally distinct recording layer" structure of claim 1 is not something the assignment record can establish.
• Current status: IBM is an operating company and remains so. The HDD business, however, was sold to Hitachi, Ltd. — the transaction closed effective 2002-12-31, forming Hitachi Global Storage Technologies (HGST). HGST was later acquired by Western Digital (2012), and the HGST brand was folded into WD in 2018. The named current assignee of record, HGST Netherlands B.V., sits within the Western Digital corporate family. This is a normal corporate chain, not a dissolved or bankrupt entity.

Assignment timeline

Three recorded events. Chronological, with execution vs. recording dates distinguished.

• 2000-03-20 (executed) / recorded 2000-03-27 — Reel 010712/0885

  • Conveyance: Assignment of assignors' interest
  • Assignor: Eric Edward Fullerton; David Thomas Margulies; Ernesto Esteban Marinero; Manfred Ernst Schabes ("and others" per the record)
  • Assignee: International Business Machines Corporation, New York
  • Correspondent: not retrieved — Assignment Center left blank/unreachable for this record; I did not guess.
  • Context: Original employment/obligation assignment of inventors' rights to IBM at filing.

• 2002-12-31 (effective) / recorded 2003-07-28 — Reel 013835/0137

  • Conveyance: Assignment of assignors' interest
  • Assignor: International Business Machines Corporation
  • Assignee: Mariana HDD B.V., Netherlands
  • Correspondent: not retrieved. Flag on recurrence: this record and the next share the same reel (013835) and the same recording date (2003-07-28), which indicates both were recorded in a single batch by one recording agent. I cannot name that agent from available sources, so this is an observation about the recording batch, not a named repeat correspondent.
  • Context: Internal/corporate transfer — the carrier entity for IBM's divestiture of its HDD business to Hitachi. Not a fire-sale of a portfolio and not a bankruptcy sale; IBM was solvent and this was a strategic business-unit sale.

• 2002-12-31 (effective) / recorded 2003-07-28 — Reel 013835/0329

  • Conveyance: Change of Name
  • Assignor: Mariana HDD B.V.
  • Assignee: Hitachi Global Storage Technologies Netherlands B.V.
  • Correspondent: not retrieved; see recurrence flag above (same reel/date as 013835/0137).
  • Context: Change of name only — no new consideration, no change in beneficial ownership; the Mariana HDD entity was renamed into the Hitachi Storage family.

Corroboration for the Mariana → Hitachi → HGST naming pattern (U.S. affiliate): Florida Division of Corporations records for the U.S. counterpart (document no. F02000006067) show Mariana HDD (U.S.), Inc. → Hitachi Global Storage Technologies, Inc. (filed 2003-01-21) → HGST, Inc. (filed 2012-10-04), consistent with the Netherlands-side chain recorded at 013835/0329. Source: https://search.sunbiz.org/Inquiry/CorporationSearch/EventHistory?...entityId=F02000006067

Post-2003 record events (not assignments): maintenance fee paid 2005-11-07 (4-year fee); maintenance-fee reminder 2009-12-14; lapse for nonpayment 2010-05-07; patent expired for failure to pay maintenance fees under 37 CFR 1.362, discontinuation recorded 2010-06-07. Anticipated expiration: 2020-03-27. No assignment of any kind was recorded after 2003.

Timeline diagram

timeline
    title Ownership of US 6383668
    2000 : Filed by IBM inventors
         : Assigned to IBM
    2002 : IBM HDD unit sold to Hitachi
    2003 : Assignment to Mariana HDD recorded
         : Renamed HGST Netherlands
    2005 : Maintenance fee paid
    2010 : Patent lapsed for nonpayment
    2020 : Anticipated expiration

NPE / troll-pattern signals

Verdict

Defensive / non-asserting — with an explicit definitional caveat.

Justification. Every recorded transfer is an internal operating-company chain: Reel 010712/0885 (inventors → IBM, 2000), Reel 013835/0137 (IBM → Mariana HDD B.V., effective 2002-12-31), and Reel 013835/0329 (Change of Name → Hitachi Global Storage Technologies Netherlands B.V., effective 2002-12-31). Zero NPE signals are present — no shell transfer, no known asserter, no cascading LLC chain, no bankruptcy sale, no privateering, no defensive-aggregator endpoint — and the patent lapsed in 2010 for nonpayment of maintenance fees, so it is now fully neutralized and carries no assertion value. The caveat: the menu's "Defensive / non-asserting" label is defined as a chain terminating at a defensive aggregator, which this chain does not do — it terminates at an operating-company subsidiary (HGST Netherlands B.V.). I select this label because the practical outcome it describes (never asserted, now dead) is the only one that fits; none of the provided options captures "clean operating-company chain with no assertion activity." The NPE verdicts are affirmatively ruled out, and Insufficient data is not applicable because the ownership chain is complete (three recorded events, none missing).

Correspondent note for follow-up: the one field I could not retrieve — the attorney/firm of record on Reels 010712/0885, 013835/0137, and 013835/0329 — should be pulled directly from the Assignment Center detail view for completeness. Based on everything else in the chain, I have no reason to expect a repeat NPE-litigation correspondent, since no NPE ever appears in this chain.

Verify at: https://assignmentcenter.uspto.gov/ (search patent number 6383668) — legacy mirror https://assignment.uspto.gov/patent/index.html. Google Patents legal-events corroboration: https://patents.google.com/patent/US6383668/en#legal-events.

Cross-reference flag (not a contradiction): The previously generated Litigation summary concluded "no known litigation" for this patent and noted the 2010 fee lapse. That is fully consistent with this section — the ownership chain contains no asserter, no litigation-enabling transfer, and the patent expired for nonpayment. Neither section contradicts the other; both independently point to a never-weaponized operating-company patent.

I'll research the USPTO record for US 6,383,668 and each of its cited references. Let me run targeted searches.

I have solid detail on all three examiner-cited patents. Now let me verify the closest sibling/related art (US 6,280,813, the co-pending application referenced in the '668 background) and the family-cited references.

Prior Art Analysis — U.S. Patent No. 6,383,668

Note on sources: The bibliographic record below is drawn from the authoritative Google Patents full text of US 6,383,668 (fetched for this analysis) and from Google Patents / Espacenet records for the individual references. Cross-checks were performed against the patent's own "Patent Citations," "Family Cites Families," and "Non-Patent Citations" tables. I did not retrieve separate databases (Lex Machina, etc.) for this section.

1. The patent under review (confirmed record)

Because this application was filed in 2000 (pre‑AIA), the operative statute is pre‑AIA 35 U.S.C. § 102. For a reference to anticipate, it must disclose every element of a claim, arranged as in the claim. The four items the '668 claims hinge on are: (i) a nonferromagnetic underlayer; (ii) a two‑film synthetic‑antiferromagnet "host" layer (two cobalt‑alloy films AF‑coupled across a Ru/Cr/Rh/Ir/Cu spacer) in which M₁×t₁ ≠ M₂×t₂; (iii) a separate magnetic recording layer of different composition in contact with, and ferromagnetically coupled to, the top host film; and (iv) a protective overcoat.

This distinction matters for everything below. US 6,383,668 is not claimed as an AFC recording layer (that is the sibling patent US 6,280,813). It claims an AFC non‑recording host layer beneath a conventional recording layer.

2. Examiner-cited patent references (the "Patent Citations" of record)

The USPTO record lists exactly three U.S. patent references cited against '668:

2.1 US 5,815,342 A

• Full citation: Akiyama, J.; Tanaka, Y. Perpendicular magnetic recording/reproducing apparatus. U.S. Patent 5,815,342 A; assignee Kabushiki Kaisha Toshiba.
• Dates: JP priority 1992‑07‑13 (JP 4‑184254); earliest US filing 1993‑07‑13 (US 09/090,312); later US filing 1995‑11‑13 (US 08/558,971); granted 1998‑09‑29.
• Description: A perpendicular magnetic recording medium having a substrate, a soft magnetic back layer (high‑permeability layer) and a perpendicular‑anisotropy magnetic recording layer, plus magnetic heads with a field‑generating element that applies a field to the soft back layer during reproduction to reduce Barkhausen noise. Optionally a bias‑field‑applying layer on the substrate side of the soft back layer.
• § 102 analysis: Does not anticipate any of claims 1–5.
  • Granted 1998‑09‑29, so it is available as § 102(b) art (more than one year before the 2000‑03‑27 filing).
  • It discloses a substrate and a recording layer, and (arguably) a protective/overcoat layer — the peripheral elements of claim 1.
  • It fails on the core element (c)/limitation (c): the "host layer." The Toshiba back layer is a single soft ferromagnetic film, not two cobalt‑alloy films antiferromagnetically exchange‑coupled across a Ru/Cr/Rh/Ir/Cu spacer. There is no spacer film, no antiparallel moment pair, and no M₁t₁ ≠ M₂t₂ relationship. It also lacks the separate, different‑composition recording layer ferromagnetically coupled to the top host film.
  • Best characterization: background art for perpendicular two‑layer media; potential § 103 combination material at most.

2.2 US 5,569,533 A

• Full citation: Lal, B. B.; Bourez, A. J.; Shinohara, T. Thin‑film medium with sublayer. U.S. Patent 5,569,533 A; assignee HMT Technology Corporation.
• Dates: filed 1994‑03‑14 (US 08/212,151); granted 1996‑10‑29.
• Description: A longitudinal magnetic recording medium on a non‑metallic substrate. A chromium‑alloy sublayer (≈3–50 Å) is deposited on the substrate, followed by a chromium‑based underlayer, then a cobalt‑based magnetic recording layer sputtered over the underlayer; noted for improved coercivity, HF signal amplitude and SNR.
• § 102 analysis: Does not anticipate any of claims 1–5.
  • Granted 1996‑10‑29, so available as § 102(b) art.
  • It does disclose elements (a) a substrate and (b) a nonferromagnetic underlayer, and (d‑partial) a Co‑based magnetic recording layer.
  • It fails on limitation (c) entirely: the sublayer is a single Cr‑alloy film, not a two‑film synthetic antiferromagnet, and there is no spacer film, no antiparallel coupling, and no M₁t₁ ≠ M₂t₂ limitation. It likewise lacks the requirement that the recording layer be of different composition from, and ferromagnetically coupled to, the top film of a host layer (there is no host layer).
  • Best characterization: relevant to the substrate/underlayer/Co‑alloy‑recording‑layer sub‑combination; not anticipatory.

2.3 US 6,294,490 B1

• Full citation: Zou, X.; Azegami, K. Crystallized glass for information recording medium, crystallized glass substrate, and information recording medium using the crystallized glass substrate. U.S. Patent 6,294,490 B1; assignee Hoya Corporation.
• Dates: filed 1999‑03‑12 (US 09/266,753); JP priorities from 1998‑03‑13; granted 2001‑09‑25.
• Description: A crystallized glass (SiO₂‑Al₂O₃‑MgO‑TiO₂, Y₂O₃‑containing) having high Young's modulus, high strength and fine surface smoothness, and a substrate and information‑recording medium (e.g., magnetic disk) made from it.
• § 102 analysis: Does not anticipate any of claims 1–5.
  • Granted 2001‑09‑25 (after '668's 2000‑03‑27 filing), so it is not § 102(b) art. It is potentially § 102(e) art because its U.S. application (09/266,753) was filed 1999‑03‑12, before '668's filing/invention date, but only as to subject matter actually disclosed.
  • The sole claim element it touches is limitation (a) "a substrate." It discloses no underlayer, no host layer, no AF‑coupled films, no spacer, no M₁t₁ ≠ M₂t₂ relationship, and no recording layer. It cannot anticipate any claim, and adds nothing to the other limitations.
  • Best characterization: a § 112‑enablement/§ 102(a)‑type reference for the substrate material only ("glass" substrates are expressly recited in '668's specification). Not anticipatory of any claim.

3. Non‑patent literature cited (the "Non‑Patent Citations" of record)

• Full citation: Parkin, S. S. P.; More, N.; Roche, K. P. "Oscillations in Exchange Coupling and Magnetoresistance in Metallic Superlattice Structures: Co/Ru, Co/Cr, and Fe/Cr," Physical Review Letters, Vol. 64, No. 19, May 1990, pp. 2304–2307.
• Date: May 1990 (more than one year before the 2000‑03‑27 filing → § 102(b) printed publication).
• Description: Demonstrates that interlayer exchange coupling between ferromagnetic films separated by a nonferromagnetic transition‑metal spacer (Ru, Cr) oscillates between ferromagnetic and antiferromagnetic as spacer thickness increases, and identifies the first AF peak. This is the physical phenomenon '668 relies on for its host layer (the specification cites it and selects 0.6 nm Ru as the first AF peak).
• § 102 analysis: Does not anticipate any claim — it is a physics paper about coupling, not a magnetic recording disk. It discloses no substrate, underlayer, host layer, recording layer, or overcoat. Its role is as § 103 evidence that AF‑coupling of Co films across a Ru spacer was known, which is why the examiner paired it with the structural references.

4. The closest art is not in the "Patent Citations" list — it is the sibling patent

US 6,280,813 B1 — the most relevant prior art

• Full citation: Carey, K. et al. Magnetic recording media with antiferromagnetically coupled ferromagnetic films as the recording layer. U.S. Patent 6,280,813 B1; assignee International Business Machines Corp.
• Dates: filed 1999‑10‑08 (Ser. No. 09/416,364 — the very application named in '668's Background section); granted 2001‑08‑28.
• Description: A magnetic recording medium whose recording layer itself is at least two granular CoPtCrB films AF‑coupled across a 0.6 nm Ru spacer at the first AF peak. One film is made thicker than the other so the net moment is low but nonzero (i.e., Mrt₁ − Mrt₂ ≠ 0), on a Cr underlayer over a substrate, with a protective overcoat. It even includes the 0.5 nm Co interface films at the Ru interface.
• Relevance / § 102 analysis: This is materially closer to '668 than any examiner‑cited reference, and it is commonly owned IBM art.
  • It discloses elements (a) substrate, (b) nonferromagnetic (Cr) underlayer, the two AF‑coupled CoPtCrB films with unequal magnetic moments across a Ru spacer, and a protective overcoat — i.e., most of limitations (a)–(c) and (e).
  • It does not anticipate claim 1 because it makes the AF‑coupled pair the recording layer itself. It does not disclose a separate magnetic recording layer of different composition deposited on and ferromagnetically coupled to the top film of the AFC pair. That "separate host layer beneath a conventional recording layer" is exactly '668's point of novelty (Background: "takes advantage of the well‑known magnetic and recording/reproducing properties of the conventional single layer granular Co alloy magnetic material").
  • So US 6,280,813 is the strongest § 103 reference and the reason '668 was allowed, but it is not anticipatory and, being commonly owned, would have been disqualified as § 103 art under pre‑AIA § 103(c).

Additional family‑cited references (relevant background, none anticipatory)

From the "Family Cites Families (8)" table of record:

Caveat: I was cut off by a tool‑step limit before independently re‑verifying the full text of US 5,583,725 and US 5,843,589 beyond the dates/titles/assignees given in the '668 record's "Family Cites Families" table; the characterizations of those two are based on that table plus the '668 specification's own discussion of spin‑valve AFC structures (which cites IBM U.S. Pat. Nos. 5,408,377 and 5,465,185 for the same principle).

5. Bottom line

• No examiner‑cited reference anticipates any claim of US 6,383,668. Each fails on the same dispositive limitation — the antiferromagnetically coupled two‑film "host" layer (claim 1 element (c)) — and none discloses the separate, different‑composition recording layer ferromagnetically coupled to the top host film (element (d)):
  • US 5,815,342 → only a single soft‑magnetic back layer; not AF‑coupled.
  • US 5,569,533 → single Cr‑alloy sublayer + Cr underlayer + Co‑alloy layer; no AF coupling.
  • US 6,294,490 → a crystallized‑glass substrate only.
  • Parkin et al. 1990 (NPL) → the coupling phenomenon, not a disk structure.
• The most relevant art overall is US 6,280,813 B1 (Carey et al., IBM; filed 1999‑10‑08, granted 2001‑08‑28), the sibling application expressly named in '668's Background. It is extremely close (two unequal‑moment CoPtCrB films AF‑coupled across a Ru spacer) but is not anticipatory, because in '628,0813 the AFC pair is the recording layer, whereas '668 requires a separate, differently‑composed recording layer on top of a non‑recording host layer. It is the best § 103 reference but would be disqualified as commonly owned under pre‑AIA § 103(c).
• Statutory framing: Pre‑AIA § 102 governs. US 5,815,342 (1998) and US 5,569,533 (1996) are § 102(b) art; US 6,294,490 (filed 1999, granted 2001) is § 102(e) art as to its disclosure; the Parkin paper is § 102(b) art; US 6,280,813 (filed 1999, granted 2001) is § 102(e) art but non‑anticipatory and § 103(c)‑disqualified.

Obviousness Analysis — U.S. Patent No. 6,383,668 under 35 U.S.C. § 103

Builds on the prior litigation section (no known litigation; expired 2010 for fee non-payment). Because the patent lapsed in 2010 and its 20-year term ran to 2020, this is a retrospective/hypothetical validity analysis — but it is the analysis a challenger would have run.

1. Governing framework

Because application 09/536,868 was filed March 27, 2000, pre-AIA § 103 governs. The controlling test is Graham v. John Deere: (1) scope/content of the prior art; (2) differences between the prior art and the claims; (3) level of ordinary skill; (4) secondary considerations. Under KSR Int'l v. Teleflex, a combination is obvious where the references teach or suggest the claimed elements and there is a "reason" to combine — including predictable variation, design incentives, and "obvious to try" within a finite set of identified, predictable solutions.

2. Level of ordinary skill

A POSITA here holds an M.S./Ph.D. in materials science, applied physics, or EE (or equivalent) plus ~2–5 years in thin-film magnetic recording media, and is familiar with: sputter-deposited Co-alloy granular media, Cr/Cr-alloy underlayers and Co-alloy onset/nucleation layers, and oscillatory (RKKY-type) interlayer coupling as used in spin-valve GMR heads. This is a mature, well-documented field by 2000 — a fact that matters for the § 103 analysis.

3. Claim 1 — limitation-by-limitation

Claim 1 (like claims 2–5) is directed to the uncompensated embodiment: it requires (M₁×t₁) ≠ (M₂×t₂), i.e., the host layer has a net moment. Notably, the patent's stated "preferred embodiment" (M₁t₁ = M₂t₂, zero net moment) is described but not claimed. (I flag this as an important analytical point, § 8 below.)

4. Prior-art references available on the page

Note on the "Cited By" list (Seagate US 6,759,149; US 6,852,426; US 6,794,057; Hitachi US 6,773,834; etc.): with priorities from 1999-10-08 (US 6,773,834) through 2001–2003, these are not available as § 102/103 prior art against '668 (filed 2000-03-27). They are useful only as evidence of the trajectory of the art, not as references. I flag this because it is a common error to cite a forward citation as prior art.

5. The prima facie § 103 combinations

Combination A (primary): US 6,280,813 + Parkin 1990 + US 5,569,533 (HMT)

• US 6,280,813 discloses the core architecture: a magnetic recording layer formed of two ferromagnetic Co-alloy films antiferromagnetically coupled across a nonferromagnetic spacer (Ru at/near the first AF peak), with a Cr-based underlayer beneath and a carbon overcoat above, expressly for the same purpose claimed here — raising KᵤV without a corresponding Mrt penalty.
• Parkin 1990 supplies the coupling physics and the specific spacer materials recited in claim 1 (Co/Ru, Co/Cr, Fe/Cr), and the fact that AF coupling strength oscillates with spacer thickness (the first AF peak — the exact design rule in '668's 0.6 nm Ru example).
• US 5,569,533 supplies the nonferromagnetic underlayer/sublayer beneath the magnetic layer, and conventional sputtering/overcoat processing.

Difference vs. claim 1: only (i) that the AFC pair is a host beneath a separate recording layer of different composition, and (ii) that it is uncompensated.

Motivation: '668's own Background concedes that the '813 approach of making the AFC pair the recording layer "introduces a new set of unknowns" and that what was needed was media that retained "the well-known magnetic and recording/reproducing properties of the conventional single layer granular Co alloy magnetic material." That statement is an admission of the precise design driver to keep a conventional recording layer while importing the AFC stack as a stabilizing underlayer. The ontogeny of a nucleation/onset layer (Fig. 1: a thin Co-alloy layer between underlayer and recording layer, optionally ferromagnetic and of different composition) supplies (d). Uncompensating the pair (c5) is conventional thickness tuning, explained by the patent itself as tuning Mrt and coercivity.

Combination B (avoids the common-ownership problem): Parkin + US 5,583,725 + US 5,843,589 + US 5,569,533

• US 5,583,725 (IBM spin-valve with a self-pinned laminated layer) teaches a two-film laminated AF-coupled pair, including deliberately unequal film thicknesses to establish a controlled net moment — directly teaching the (c5) limitation — and Co interface layers to enhance coupling.
• US 5,843,589 (Hitachi) teaches magnetic layered materials and storage/read systems using AF-coupled multilayers.
• Parkin + US 5,569,533 + US 5,815,342/JPH06103553A supply coupling, underlayer, and disk structure.

This combination does not require US 6,280,813 and therefore sidesteps the § 103(c) issue (§ 9).

Combination C (structural analogue): US 6,280,813 + US 5,843,589 + US 5,569,533

Same structure, using Hitachi's layered material as the enabling AFC teaching instead of Parkin. Cumulative of A/B.

6. Why a POSITA would have combined — KSR motivations

1. Same field, same problem, same solution space. All references are magnetic recording media/sensors addressing thermal stability and media noise in sputter-deposited Co alloys. The field is predictable.
2. Design incentive. '813 (closest art) had a known deficiency (unpredictable recording properties); keeping a conventional granular recording layer on top is the natural, low-risk fix.
3. "Obvious to try." There are but a handful of identified options — AFC as recording layer vs. AFC as a underlayer/host; compensated vs. uncompensated. KSR holds such predictable variations obvious.
4. Predictable physics. The claimed benefit (V increases by the sum of grain volumes while Mrt stays governed by the top layer) follows directly from the coupling physics Parkin and '813 already disclose; no new mechanism is asserted.
5. Later-industry conduct corroborates. Seagate's "Laminated medium with antiferromagnetic stabilization layers" (US 6,759,149) and "Hybrid anti-ferromagnetically coupled and laminated magnetic media" (US 6,852,426), plus Hitachi/Fujitsu/DSI filings, show the industry independently converged on an AFC layer used as a stabilizer beneath/with a recording layer within ~1–3 years — consistent with the solution being within the ordinary skill level (though see § 10 on nexus).

7. Dependent claims

• Claim 2 (same material, t₁ ≠ t₂): obvious — selecting relative thicknesses to tune (M₁×t₁) vs. (M₂×t₂) is a routine optimization; In re Aller / In re Boesch.
• Claim 3 (different materials, t₁ ≈ t₂): obvious — Parkin expressly teaches that coupling strength/phase depends on the ferromagnetic material; choosing different Ms and equal thickness is a simple substitution and known equivalent.
• Claims 4 & 5 (0.5 nm Co interface film against the spacer): obvious — Parkin's Co/Ru system and US 5,583,725 teach ultrathin Co at FM/spacer interfaces to enhance coupling; the patent itself frames the Co interface film as optional and merely "enhanc[ing]" coupling.

8. The (M₁×t₁) ≠ (M₂×t₂) limitation — weakness for the patentee

Claim 1 requires a net-moment (uncompensated) host. The closest art ('813; the spin-valve synthetic antiferromagnet) teaches that equality of moments gives a compensated structure, and US 5,583,725 teaches that inequality is a deliberate, known option. The patent's own specification supplies the reasons for inequality (tune Mrt, adjust coercivity, oppose the demagnetizing field, avoid top-film noise), which are ordinary design objectives. Selecting the uncompensated variant is therefore the least inventive option in the disclosure — and it is the only option claimed.

9. § 103(c) wrinkle on US 6,280,813

US 6,280,813 issued 2001-08-28, after the '668 filing date. It is therefore available only as § 102(e) art (its 1999-10-08 filing date). Because '813 and '668 were both owned by IBM at the time the '668 invention was made, pre-AIA § 103(c) would disqualify '813 for use in a § 103 combination. A careful challenger would therefore lead with Combination B (Parkin + US 5,583,725 + Hitachi/Toshiba/HMT art) and use '813 only for background/admissions. This is very likely part of why the examiner (who cited only Parkin, US 5,569,533, US 5,815,342, and US 6,294,490 on the face) did not reject over '813 — though '813 appears in the family-citation list.

10. Counterarguments the patentee would raise (and their strength)

• No express suggestion to put a separate recording layer on an AFC "host." The strongest non-obviousness point. All the AFC art uses the AFC pair as the recording layer or as a pinned layer in a sensor. The "host" architecture is the genuine insight. Strength: moderate — KSR's "obvious to try" and the '668 Background's own statement of the problem largely neutralize it.
• Teaching away? '668's Background notes the AFC-recording-layer approach has "unknowns," but that is a statement of inconvenience, not a teaching away; if anything it motivates the host variation, which cuts against the patentee.
• Secondary considerations. Long-felt need for thermally stable high-density media existed and is documented. However: (i) there is no asserted unexpected result — the benefit flows predictably from acknowledged AFC physics; (ii) nexus is weak because the industry's AFC media were predominantly the compensated variety, which the claims do not cover; (iii) the rapid, widespread independent adoption by Seagate, Hitachi, Fujitsu, and DSI, while arguably a compliment to the technology, more naturally reflects a predictable state-of-the-art solution than a breakthrough — and in any event the '668 claims cover the variant industry largely did not use. Licensing/cross-licensing within the HDD industry (noted in the earlier section) provides little probative weight.
• Examiner allowance. Not dispositive of obviousness, and no adversarial validity proceeding ever tested these claims (the patent lapsed for non-payment in 2010 — consistent with low asserted value).

11. Conclusion

On the record of the page's prior art, there is a strong prima facie case of obviousness under § 103 against every claim:

• Claim 1 would be obvious over US 6,280,813 in view of Parkin 1990 and US 5,569,533, and independently over Parkin 1990 in view of US 5,583,725 and US 5,569,533 (the latter avoiding the § 103(c) obstacle).
• Claims 2–3 are routine optimization of thickness/materials ratios.
• Claims 4–5 are obvious over Parkin's Co/Ru interface teaching and US 5,583,725's interface layers.

The single most vulnerable feature for the patentee is that claim 1 requires the uncompensated host, i.e., the very variant the specification presents as an alternative and whose design rationale is ordinary Mrt/Hc tuning. The patent's allegedly distinctive "compensated host" architecture is not claimed. The strongest defense is the absence of an express prior-art suggestion to interpose a separate conventional recording layer atop an AFC pair — a defense that KSR's "obvious to try" and the patent's own Background admissions substantially undercut.

Caveats (stated honestly): For most references only title, assignee, and dates are available on the page; the element mappings above rely on those plus the '668 specification's admissions and the general state of the art, and would require full-text verification. I also flag a record inconsistency: the prosecution cites Parkin 1990 at "p. 2304-2307," while the '668 body text cites the same paper at "p. 2034" — the page number in the body appears erroneous, though the reference itself is unambiguous.