US 5818694

I'll run searches against USPTO and CAFC sources for this specific number.

Let me confirm whether any litigation or appeal activity exists for this specific patent.

US Patent 5,818,694 — Analysis

Source reliability note: My searches were general web searches rather than direct queries of the USPTO PatentCenter or the CAFC docket system; I could not open those databases' internal search interfaces. However, the full authoritative patent text (including claims, front-page data, citation lists, and legal-events history) was supplied in the prompt and is the primary basis for what follows. Where search results are corroborated by that text, I say so.

---

Bibliographic data

Abstract (verbatim substance)

The cooling apparatus has cooling fins attached to heat generating units of electronic devices and provided with a plurality of elemental fins. In one embodiment, a first cooling fin has elemental fins of uniform fin density in the height direction from the fin base. A second cooling fin has elemental fins with a zero or small fin density at the lower portion close to the fin base as compared with the fin density at the upper portion. The first and second cooling fins are positioned adjacent to one another. Various different embodiments of cooling fins are disclosed.

---

Plain-language overview of the independent claims

The patent has six independent claims: 1, 4, 10, 11, 13, and 14. (Claims 2, 3, 5, 6, 7, 8, 9, and 12 depend from earlier claims. Note that claims 10 and 11 are drafted as "In a cooling fin…" and contain no dependency reference, so they are independent claims notwithstanding the preamble style.)

Claim 1 — An apparatus having at least two adjacent cooling fins mounted to heat-generating units. The first fin has elemental fins at uniform density from base to top. The second fin has a higher elemental-fin density at its upper portion and a lower density at its lower portion near the fin base. The two fins must be mounted adjacent to one another.

Claim 4 — An apparatus with two adjacent cooling fins of different orientations. The first fin has many parallel, rectangular elemental fins attached perpendicular to its fin base. The second fin has parallel, rectangular elemental fins attached to a support fin that is erected perpendicular to the fin base (i.e., the blades lie parallel to the base). Claim 4 does not itself require a density gradient.

Claim 10 — Drawn directly to a cooling fin rather than an apparatus: the fin's elemental fins are denser near the top than near the bottom.

Claim 11 — Also drawn to a cooling fin: its parallel, rectangular elemental fins run parallel to the fin base and are carried on a support fin mounted perpendicular to the base; density is lower at the bottom than at the top.

Claim 13 — Same two-fin, adjacent arrangement as claim 1, but with a hard limit: the second fin has elemental fins at its upper portion and no elemental fins at all at the lower portion near the base (zero density, not merely lower density). This corresponds to the fin 90 depicted in FIG. 8.

Claim 14 — Same as claim 4 (perpendicular fins on the first fin; parallel fins on a perpendicular support fin in the second), with the added requirement that the second fin's elemental fins are present only at the upper portion of the support fin and none at the lower portion.

Design intent shared by the independent claims: each is aimed at the same physical problem — in a duct between parallel circuit boards, cooling air moves fastest in the open space above the fins and slowest near the board surface where the chips (heat sources) actually sit. By thinning or removing fins near the base, the upstream fins redirect more high-velocity, low-temperature air down to the fin bases, reducing pressure resistance and improving downstream cooling. Dependent claims add the alternatives of wider fin spacing (claim 6) or thinner fins (claim 7) at the bottom, and impingement-style cooling from above (claims 8, 9).

---

CAFC 2026 docket check — result

No litigation, PTAB, ITC, or Federal Circuit activity involving U.S. Patent 5,818,694 was found. My searches of CAFC 2026 dockets and general litigation sources returned no cases naming this patent. This is unsurprising: the patent's term expired (anticipated expiration 2017-02-14) roughly nine years before the current date, so it is not enforceable and would not normally appear on a 2026 docket.

Identifier-collision warnings — do not conflate these with US 5,818,694:

• JP5818694B2 ("Magnetic film reinforced inductor," Qualcomm Incorporated, registered 2015-11-18) is an entirely different, unrelated Japanese patent that happens to share the digit string "/5818694."
• US Trademark Registration 5818694 ("ENTRE GENRES," Coty International BV / Orveda Limited) is a trademark registration, not a patent.
• Various unrelated hits (a Brazilian company registration number, a Slovak public-procurement contract ID, an SEC filing accession reference) are coincidental numeric matches.

Uncertainty statement: I have full confidence in the bibliographic data, abstract, and claim characterizations because they come from the authoritative full patent text. I have no authoritative evidence of any 2026 CAFC appeal, IPR, or district court case involving this patent, and I cannot verify a negative from a general web search with certainty. Also note the Google Patents "Current Assignee: Hitachi Ltd" designation carries an express disclaimer that the listed assignees may be inaccurate; the recorded assignment I can confirm is the original one to Hitachi, Ltd.

I'll research litigation involving US patent 5818694 across the sources you mentioned.

Initial results are noisy. Let me query the specific litigation databases directly.

Let me check the specialized litigation portals directly.

Summary of Findings

No litigation involving U.S. Patent No. 5,818,694 was found in the searches conducted. I was unable to identify any district court case, ITC investigation, PTAB proceeding (IPR/PGR/CBM), or Federal Circuit appeal in which U.S. 5,818,694 was asserted or challenged.

I want to be explicit about the limits of this answer: this is a negative finding from an incomplete set of sources, not a certification that no suit ever existed. Please read the caveats below.

---

What U.S. 5,818,694 is (for identification)

• Title: "Cooling apparatus for electronic devices"
• Inventors: Takahiro Daikoku; Shigeyasu Tsubaki
• Assignee: Hitachi, Ltd.
• Priority: 1996-02-16 (JP 8-28938 / JP H09223883A)
• Filed: 1997-02-14 (US 08/800,460)
• Granted: 1998-10-06
• Status: Expired – Lifetime; anticipated expiration 2017-02-14
• Claims: 14 (directed to cooling fins with varying elemental-fin density near the fin base)
• Source: https://patents.google.com/patent/[US5818694](/patent/US5818694)/en

The Google Patents record for this patent contains no litigation section — it lists only prior-art citations (US 5,097,385 to IBM, "Super-position cooling"; US 5,191,511 to Toshiba) and forward citations by later patents (e.g., WO2016192482A1's search report). Those are patentability citations by examiners and third parties, not evidence of litigation, and should not be confused with lawsuits.

Identifier caution (per your strict rule)

A search for "5818694" returns several unrelated documents sharing the same numeric string in other jurisdictions/contexts, which I am flagging rather than treating as the same patent:

• JP 5818694 B2 — this is not a counterpart of the US patent. It is a Qualcomm patent ("Magnetic film reinforced inductor," filed 2010, registered 2015). Same number, different document and jurisdiction.
• The Japanese counterpart of this US patent is JP H09223883 A (published 1997-08-26).
• Various non-patent hits (Brazilian corporate filings, German police reports, Florida probate docket instrument numbers, etc.) also contain the string "5818694" and are irrelevant.

I did not auto-correct or conflate any of these.

Sources searched

• Google Patents (US5818694A) — no litigation entries
• Web searches targeting litigation content tied to "5818694" and "5,818,694" (including case documents surfaced from docketalarm.com, CourtListener/RECAP, and PTAB-related PDFs)
• Searches aimed at Unified Patents' litigation portal and RPX Insight litigation records

The searches surfaced case documents and dockets unrelated to this patent (e.g., Ultravision Technologies v. Holophane/Acuity/Samsung; Intellectual Ventures v. Xilinx; Dynamic Advances v. Apple; Daignean Technologies telecom cases; NuVasive/NMP trade-secret/IPR matters). In each instance the "5818694" string appeared as a page/exhibit artifact, a docket-entry number, or an unrelated identifier — not as the asserted patent number.

Caveats — where this answer could be wrong

1. I could not directly query the primary specialized databases. The searches did not return first-party results from portal.unifiedpatents.com/litigation or live PACER dockets for this number. Those are the authoritative sources you specified, and I was unable to confirm a "no cases" result directly from them.
2. Pre-2010 litigation may be poorly indexed online. The patent was enforceable roughly 1998–2017. Any older suit (especially a 1990s–2000s Hitachi enforcement action) may not appear in modern free litigation databases.
3. Settled or sealed cases may never have generated a public opinion or indexed docket entry.
4. Absence of a litigation record is consistent with the patent's profile: it appears to have been a defensive/product patent for Hitachi's own cooling designs, with no assignment to a non-practicing entity (a common trigger for assertion campaigns). No reissue, no later assignee, and expiration in 2017 reduce the likelihood of assertion activity.

Recommended next steps to get a definitive answer

• Unified Patents litigation search: https://portal.unifiedpatents.com/litigation — search by patent number 5,818,694 (and separately by "Hitachi" as plaintiff/defendant).
• PACER (paid): run a party search for "Hitachi" and a patent-number search; also check Docket Alarm and CourtListener/RECAP full-text.
• Lex Machina / Darts-IP / RPX Insight (subscription): these index patent case-by-patent and would give a definitive negative or a case list.
• USPTO Patent Center / Assignment record (Reel/Frame 009251/0208) to confirm chain of title — useful to see whether the patent was ever assigned to an asserting entity.

Bottom line: Based on the searches performed, I found no known litigation involving U.S. 5,818,694. If you need this represented as an exhaustive clearance opinion, the subscription databases above should be run directly, since I could not confirm their contents for this number.

I'll first verify the structured data against live sources before drafting.

Let me run a couple more targeted checks before concluding.

Proceedings overview

There are zero AIA trial proceedings on file for US 5,818,694. The USPTO Open Data Portal returned no IPR, PGR, or CBM petitions as of the most recent ingest, and independent web searches (PTAB/Docket Alarm docket text, Google Patents' litigation and "Cited By" tables, CourtListener) surfaced no petition, institution decision, Final Written Decision, or Federal Circuit appeal naming this patent.

Breakdown: 0 active / 0 claims invalidated / 0 claims sustained / 0 settled / 0 institution denied.

Defensive posture for a defendant: this is not the "hardened patent" scenario (there is no IPR history to point to at all) — it is the "nobody bothered, because the patent is dead" scenario. US 5,818,694 issued 1998-10-06, its twenty-year term ran from the 1997-02-14 US filing date, and the record shows anticipated expiration 2017-02-14 with legal status "Expired - Lifetime." An expired patent cannot support forward-looking injunctive relief, and § 286's six-year damages lookback from any suit filed today (2026) reaches back only to roughly 2020 — entirely after expiry. If you are being asserted against on this patent today, the first question is not "how do I survive an IPR," it is "how is this patent still enforceable at all."

No proceeding entries to report

Nothing to enumerate. I have deliberately not fabricated proceeding numbers to fill the template. For the record, here is what the negative search covered:

Collision warnings (things that will mislead you in a docket search)

These are the false positives a "581" or "5818694" search throws off. None of them is US 5,818,694:

• "MPH's '581 patent" in MPH Technologies Oy v. Apple Inc., No. 3:18-cv-05935-TLT (N.D. Cal.) — the court construed "the secure connection" claims and reversed itself on indefiniteness under Fed. R. Civ. P. 54(b) (ECF 134-8). That is a secure-messaging patent, not the Hitachi cooling-fin patent. Do not cite it.
• JP 5818694 B2 — a Qualcomm "magnetic film reinforced inductor" (Japanese patent number, different jurisdiction and subject matter).
• Trademark Registration 5,818,694 — COTY International, "ENTRE GENRES."
• US 5,818,694 as prior art — it is cited in the search report of WO 2016/192482 A1 (ZTE plug-in box), where the ISA marks it "A" (background art only). Being cited as prior art is not a validity challenge to it.

Strategic summary

Claim status. All 14 claims of US 5,818,694 — independent claims 1, 4, 10, 11, 13, 14 and dependents 2, 3, 5–9, 12 — are UNTESTED at the PTAB. Not one was canceled, and not one was confirmed after adversarial review. There is no IPR-narrowed claim set to work with, but that cuts against the patent owner in a different way: this patent never had to survive a serious validity attack because it expired before the modern IPR wave reached 1990s-era cooling hardware.

Estoppel landscape. There is no § 315(e)(2) estoppel in existence, because no petitioner ever reached a Final Written Decision. Any invalidity theory a defendant wants to run — including well-trodden § 102/§ 103 art on stacked and variable-density fin arrays (see, e.g., IBM's US 5,097,385, "Super-position cooling," cited on the face of the patent) — is unburdened by estoppel. But note the flip side: the patent's 1996 priority date means an enormous body of intervening art is not prior art, and old fin-geometry references are what you would need.

Pattern signals. No serial petitioner, no defensive aggregator (Unified Patents, RPX, OpenSky) involvement, no patent-owner appeal activity, and no PTAB appeal to the Federal Circuit. The complete absence of PTAB activity on a 1998 Hitachi patent is unremarkable: it is an expired, never-asserted component patent. The 59 forward citations are almost entirely modern display-cooling patents (Manufacturing Resources International's portfolio) citing it as background, not as a blocking reference.

Recommended next steps

1. Establish the true legal status before anything else. The source record shows the last maintenance-fee payment at the 12-year window (2010-03-30) and then lists anticipated expiration of 2017-02-14 with status "Expired - Lifetime." Confirm in USPTO PatentCenter whether the 16-year and 20-year maintenance fees were paid, whether the term was extended under § 154(b)/§ 156, or whether it lapsed early. If it lapsed for non-payment before 2017, your exposure window closes even sooner.
2. Serve a Rule 11 / § 285-grade demand for proof of standing and enforceability. If a demand letter cites claims 1–14, ask the sender to identify (a) the current owner of record, (b) the expiration date they concede, and (c) the accrual dates of any asserted damages. If they assert claims 1 or 13 (the uniform-density-plus-lower-void fins) or claims 4/14 (the perpendicular-vs-parallel fin arrangement), quote the specification's own admission at col. 1 that variable-density and orthogonal fins were known — that is your § 103 story, and the patent's own background section is your best exhibit.
3. If you need an IPR for a co-defendant or licensing posture, be candid about the economics. The Board can institute on an expired patent, but with no injunctive value and a damages window that appears to be zero, an IPR petition is almost never the right spend. A motion to dismiss on unenforceability/expiry, or simply declining to take a license, is the better play.
4. No litigation-calendar items to track. There is no institution deadline, no oral hearing, and no statutory one-year FWD date, because there is no trial to schedule.

Caveat on sources: the PTAB proceedings determination rests on the structured ODP block supplied in the prompt (canonical) plus independent searches of PTAB/Docket Alarm text, Google Patents' litigation and citation tables, and CourtListener. I did not locate a Docket Alarm or E2E docket page for this patent because none appears to exist. If you need a belt-and-suspenders confirmation, run the patent number directly against the PTAB E2E "Patent Trial" search and the USPTO PatentCenter "Transactions" tab; I found no evidence of any proceeding, and I am not aware of any beyond those tools' coverage.

I'll verify the assignment record against available sources before reconstructing the chain.

Assignment & Ownership Chain — U.S. Patent No. 5,818,694

Scope note / source limitation. The USPTO Patent Assignment Search interface (assignmentcenter.uspto.gov) could not be queried directly — it is a JavaScript application that does not return records to a general web query, and my searches returned no first-party assignment records for this number. Everything below is therefore reconstructed from the authoritative full patent text supplied in the prompt, whose "Legal Events" table is a Google Patents mirror of the USPTO assignment and fee-payment records, plus corroborating searches. Where a field could not be obtained, I say so rather than filling it in. I found exactly one recorded assignment, and it is the original inventor→employer assignment.

---

Inventors

Pattern assessment — no departure signal. The red flag you asked about (all inventors leaving the original assignee within 12 months of filing) is not present. Both inventors remained Hitachi employees/researchers for years after the 1996 priority filing; Tsubaki is still publishing under a Hitachi Research Laboratory affiliation and appearing on Hitachi-assigned cooling patents into the 2010s. This is the profile of a corporate R&D team filing in the ordinary course, not a team preparing to spin out or cash out.

---

Original assignee

Hitachi, Ltd. (Chiyoda-ku, Tokyo) — named as assignee on the issued patent, and the assignee on the sole recorded assignment below.

• Primary line of business: diversified industrial/electronics conglomerate; the relevant business unit at the time was Hitachi's computer/mainframe and information-systems division, which used multi-chip modules (MCMs) with multi-layer ceramic substrates — precisely the hardware described in the specification (multi-layer ceramic board 2, socket 5, printed circuit boards 61/62).
• Product embodiment: The specification and the FIG. 10 experiment (a physical module placed in a 34 mm × 93 mm aeration passage at 0.8 m³/min, with measured outlet air-speed and temperature-rise distributions) show this was a built and tested hardware design, not a paper filing. So yes — Hitachi plausibly shipped products embodying the claims as part of its MCM/mainframe cooling design.
• Current status: Operating. Hitachi, Ltd. continues in business today (reorganized into a holding structure with listed subsidiaries), and no bankruptcy, receivership, or liquidation affecting the assignee appears in the record. This is not a Kodak/Nortel/Polaroid-style bankruptcy estate.
• Caution carried forward from the prior section: Google Patents displays "Current Assignee: Hitachi Ltd" but attaches an express disclaimer that listed assignees may be inaccurate. The only assignment I can actually confirm is the original one to Hitachi, Ltd.

---

Assignment timeline

One record only.

• 1997-02-13 (executed) / recorded 1998-06-17 — Reel 009251 / Frame 0208
  • Conveyance: Assignment of Assignors' Interest (USPTO event code AS; "ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS)")
  • Assignor: DAIKOKU, TAKAHIRO; TSUBAKI, SHIGEYASU (jointly)
  • Assignee: HITACHI, LTD. (Japan)
  • Correspondent: Not determinable from the sources available to me. The Google Patents legal-events mirror does not expose the correspondent field, and Assignment Center was not directly queryable. I will not guess an attorney or firm name. This field should be pulled from the Assignment Center detail page for Reel 009251 / Frame 0208 before any recurrence analysis is attempted. Consequently I can neither confirm nor rule out the "repeat correspondent" signal for this chain.
  • Context: Ordinary pre-issuance inventor-to-employer assignment, executed the day before the U.S. filing date (1997-02-14) and recorded after allowance, in preparation for issue. Standard corporate practice; no third party is involved.

There are no post-issuance assignments: no transfer to an IP holding entity, no security agreement, no merger, no change of name, no license recorded. The remaining legal events in the file are maintenance-fee payments only — 4-year (2002-04-05), 8-year (2006-03-31), and 12-year (2010-03-30) — all paid by the original owner, consistent with Hitachi retaining the patent to expiration.

Finding: The chain terminates at the original operating-company assignee. Hitachi, Ltd. was the owner of record from 1997 until the patent's term ended on 2017-02-14.

---

Timeline diagram

timeline
    title Ownership of US 5818694
    1996 : Priority filing in Japan
    1997 : US application filed 14 Feb
         : Inventors execute assignment to Hitachi 13 Feb
    1998 : Assignment recorded 17 Jun
         : Patent issued 06 Oct
    2002 : 4 year maintenance fee paid
    2006 : 8 year maintenance fee paid
    2010 : 12 year maintenance fee paid
    2017 : Term expires 14 Feb

---

NPE / troll-pattern signals

---

Verdict

Insufficient data — but only in the narrow, rubric-defined sense that the file contains no records beyond the original assignment. Applying your exact definition ("no records, or only the original assignment"), that is the correct label: the entire ownership record is the single inventor→Hitachi instrument at Reel 009251/Frame 0208, executed 1997-02-13, recorded 1998-06-17.

That said, the substance should not be misread as a dead end. All eight NPE signals are not present (with signal 3 merely unclear because the correspondent field was unobtainable, not because anything suspicious exists). The chain never left the original operating assignee, Hitachi, Ltd. retained ownership through the 4-, 8-, and 12-year maintenance-fee payments, the patent expired by its own terms on 2017-02-14, and the prior Litigation section found no assertion of this patent anywhere. On the evidence, this is plain operating-company retention of a product-embodying cooling patent — the opposite of a troll profile — and the only reason it is not scored "Operating-company assertion" is that the rubric reserves that verdict for a current assignee actually suing competitors, and no suit exists.

Verify at: https://assignmentcenter.uspto.gov (search by patent number 5,818,694, or by Reel/Frame 009251/0208) — this will also yield the correspondent of record, the one field I could not retrieve.

Identifier caution, carried forward: do not conflate this record with JP 5818694 B2 (Qualcomm, "Magnetic film reinforced inductor," registered 2015-11-18) — same digit string, unrelated document and jurisdiction. The true Japanese counterpart of this patent is JP H09223883 A (published 1997-08-26), which shares the same priority (JP 8-28938) and the same two inventors.

I'll research the cited prior-art references for US 5,818,694. Let me verify each reference.

Let me verify additional details on these references.

Let me confirm the inventor data for US 5,097,385 to avoid misattribution.

Prior-Art Analysis — U.S. Patent No. 5,818,694

0. Identification of the patent (literal, per strict-ID rule)

I did not conflate this with JP 5818694 B2 (Qualcomm "Magnetic film reinforced inductor"), which shares only the digit string.

---

1. The complete prior art of record — a critical scoping point

This is the single most important finding for your question: US 5,818,694 has an unusually thin prior-art record. Per the authoritative Google Patents front matter for US5818694A (https://patents.google.com/patent/US5818694/en), the "Citations" field contains only two patent documents, and the entire "Non-Patent Citations" field contains only one work (listed twice):

Patent citations (2):

Non-patent citation (1 work, duplicated in the record):

• Japan Society of Mechanical Engineers (JSME) Data Book, Heat Transfer (4th Edition) (1989), p. 206.

Both listed dates above are the dates Google Patents prints in the citation table under the ★/"cited by examiner" column. Both references published more than one year before the 1997-02-14 U.S. filing, so both are §102(b) statutory-bar prior art with respect to that filing date (and predate the 1996-02-16 foreign priority date as well).

Three warnings about what is not prior art in this record:

1. The "Cited By" lists are not prior art. Google Patents shows 59 (short list) / 118 (full list) forward citations — including the entire Manufacturing Resources International family (US9797588, US9801305, US9894800, etc.), US5947192A (Chaun-Choung "Stack-fin radiator"), US6181556B1, US6244331B1, and so on. Every one of those post-dates 5,818,694 and cites it; none can be §102 art against it. Do not treat them as anticipatory.
2. JP H09223883A is the Japanese counterpart of the same invention (same priority/assignee/claims family) — it is a family member, not prior art.
3. The examiner's citation of US 5,191,511 (a Toshiba package/bump patent) is, on its face, a "state of the art / background" citation about semiconductor packages mounted on a board. It is not a cooling-fin reference.

---

2. Reference-by-reference analysis

Reference 1 — US 5,097,385 A ("Super-position cooling")

Full citation: US 5,097,385 A, Super-position cooling, International Business Machines Corporation, filed/priority 1990-04-18, published 1992-03-17.

*(Inventor note/uncertainty: my searches did not return a confirmed inventor name for US 5,097,385. I found an IBM European document, EP 0 447 819 A3 — "Multileveled electronic assembly with cooling means," inventor Balan, Albert Lawrence, priority US 495,563 filed 1990-03-19 — but that document's abstract matches US 5,063,475, a different IBM patent, not US 5,097,385. I therefore decline to attribute any inventor to US 5,097,385 and flag this as unverified rather than guess.)*

Description (from the patent's own text, retrieved via Google Patents): A cooling system and method using both jet impingement and conduction cooling for a thermal conduction module. Integrated-circuit chips are mounted by solder balls to a substrate; a cluster of spring-loaded metal pistons (aluminum/copper) contacts each chip, conducting heat to a cap and cold plate (water channel). A dielectric coolant is delivered to each piston cluster from a central/per-piston channel so that coolant "flows over the top of the feet, and in the channels between the feet," giving "jet enhanced convective cooling at the lower portion of the pistons." The pistons act as extended surfaces; microfins are added to the header and the upper piston sections (longitudinal, diamond-shaped, and brick-shaped arrays on the upper piston portion) to increase wetted area. Claim 12–14 of that patent recite "the upper portion of at least one of the pistons comprises [longitudinal / diamond-shaped / brick-shaped] fins disposed thereon."

§102 anticipation analysis against US 5,818,694's claims:

Bottom line on US 5,097,385: It is a jet-impingement + conduction-cooling reference for liquid-cooled thermal conduction modules. It does not anticipate any of claims 1, 4, 10, 11, 13, or 14 — or, derivatively, any dependent claim. Its genuine value to the examiner was almost certainly as general background showing that (i) impingement cooling from above is known and (ii) fins/microfins are added to heat-transfer surfaces to increase area — i.e., §103 material for the impingement-dependent claims (8, 9) if they were ever separately challenged.

---

Reference 2 — US 5,191,511 A ("Semiconductor device including a package having a plurality of bumps arranged in a grid form as external terminals")

Full citation: US 5,191,511 A, Kabushiki Kaisha Toshiba, application US 07/830,892; priority 1991-02-08; published 1993-03-02.

Citation/dating discrepancies I am flagging rather than auto-correcting (per your strict rule):

• Google Patents prints 1991-02-08 / 1993-03-02 (as recorded in the prompt front matter).
• The Unified Patents portal record for US-5191511-A prints priority 1991-02-07, application date 1992-02-03, and grant date 1993-03-01.
• A third source (zhangqiaokeyan) prints application date 1992-02-04.
  These are one-day variances in indexing; I report each literally and do not reconcile them.

Description: A semiconductor package / packaging invention — not a thermal invention. A package body is formed of three laminated ceramic (or plastic) insulation plates with central windows; metallized Cu patterns and a high-thermal-conductivity heat-spreading island (Cu-alloy) carry a chip; wire bonds connect chip to patterns; and solder bumps are arranged in a grid form on the rear surface of the package as external terminals, connected to the metallized patterns through via holes. Its stated objects are suppressing package-size growth from terminal-count increases, lowering cost/weight, avoiding bent PGA pins, and easing visual inspection/mass production. (Notably, the only cooling-adjacent element is the passive "heat radiation plate.")

§102 anticipation analysis: It anticipates no claim of US 5,818,694.

Why it was cited: Its role is background/§102(b) context — it shows the ordinary state of the art in mounting many semiconductor devices on a multi-layer ceramic/organic board with high-density terminals. That is the very environment described in 5,818,694's "multi-chip module" spec (semiconductor devices 31–33 on multi-layer ceramic board 2, socket 5, micropins 4, printed circuit boards 61/62). It supplies none of the claimed fin geometry.

---

Non-patent citation — JSME Data Book, Heat Transfer (4th ed., 1989), p. 206

Full citation (as printed in the record): Japan Society of Mechanical Engineers (JSME) Data Book, "Heat Transfer" (4th Edition) (1989), p. 206.

§102(b) status: Published 1989 — more than one year before both the 1996-02-16 priority date and the 1997-02-14 filing. Properly §102(b) art.

Description and how the patent itself uses it: US 5,818,694 expressly cites this page in its Background as describing the conventional practice — "cooling fins usually employed include elemental fins arranged to have a uniform fin density as measured in a vertical direction of the fin … uniformly provided along the entire height of the fin."

§102 analysis: This is the reference that discloses the "first cooling fin … uniform density from fin base … to top" element of claims 1 and 13 (and the uniform-density baseline of claim 4's environment). But it discloses only that uniform-density fin — it is silent on a second, adjacent fin with a zero-or-lower fin density at the lower portion near the fin base (claim 1), on intersecting-orientation fins (claim 4), and on the per-fin density gradient (claims 10, 11). Because §102 requires every claimed element in a single reference, the JSME page anticipates no claim. It is instead the admitted prior-art baseline against which the claimed gradient is measured — i.e., the starting point for a §103 obviousness analysis, and the reference the patentee effectively conceded.

Flagged discrepancy (strict rule): The record prints the JSME Data Book as "(4th Edition) (1989)." One library catalog record I retrieved (JAEA OPAC) describes the 改訂第4版 (revised 4th ed.) as published 1986.10 by Maruzen. I do not auto-correct the patent's stated 1989 date; I report both and note the citation date is the 1989 figure the examiner used for §102(b) purposes.

---

3. Direct answer: which claims each reference potentially anticipates under §102

No reference of record, alone or in combination, discloses the single feature that is the point of novelty of every independent claim: an elemental-fin density that is zero or lower at the lower portion close to the fin base as compared with the upper portion — mounted adjacent to a uniform-density fin (claims 1, 13), combined with intersecting fin orientation (claims 4, 14), or claimed as the fin per se (claims 10, 11). That is why none of the cited art anticipates any of the 14 claims.

This is consistent with (and helps explain) the litigation/search finding carried over from the earlier sections: no §102/§103 challenge to this patent appears in the public record, and its 2017 expiration closed the matter. It also explains why the examiner's citation set is so small and so weakly on-point — there simply was no close anticipatory art for the fin-density gradient.

---

4. Caveats and confidence

• High confidence in the identity and count of the citations (2 patent + 1 non-patent work) and in the content of each reference, because both come from the authoritative full-text record supplied and from the references' own text retrieved via Google Patents / EveryPatent / Unified Patents.
• Moderate confidence in the exact bibliographic dates for US 5,191,511 (three sources vary by one day — see above; reported literally).
• Low/unverified: the inventor of US 5,097,385 — I could not confirm it and have deliberately not assigned a name. (Balan/EP 0447819 belongs to the different IBM patent US 5,063,475.)
• Cannot verify a negative with certainty: I could not directly query USPTO PatentCenter's internal citation pane for US 08/800,460; the citation set above is taken from the authoritative Google Patents front matter for US5818694A, which mirrors the printed patent. If you need an examiner-of-record opinion (e.g., the actual Form PTO-892 "References Cited" and the examiner's reasons for allowance), pull the USPTO PatentCenter file wrapper for application 08/800,460 — that is the definitive source and would confirm whether any additional art was cited or considered.
• Not prior art: the 59/118 "Cited By" documents and the "Families Citing this family" entry (JP 5381561 B2, Fuji Electric, 2014) are forward citations and post-date the patent.

Sources: https://patents.google.com/patent/US5818694/en ; https://patents.google.com/patent/[US5097385A](/patent/US5097385A)/en ; https://portal.unifiedpatents.com/patents/patent/US-5191511-A ; everypatent.com records for US5097385 and US5191511.

I'll research the cited prior-art references before constructing the §103 analysis.

Let me examine a closely-related Hitachi reference that appears in the page's "Similar Documents" list, plus check for any examiner reasoning.

Let me verify the filing date and content of the related Hitachi reference US5705854A.

Obviousness Analysis of U.S. Patent No. 5,818,694 under 35 U.S.C. § 103

Reference frame: This builds on the prior sections (bibliography, claim taxonomy of the six independent claims, and the negative litigation finding). I will not repeat those. One flag: the previously generated sections were produced against "2026-09-26/27," whereas this task specifies "April 26, 2026." That discrepancy is immaterial to the §103 analysis (the patent expired 2017-02-14 either way) but I am noting it rather than silently reconciling it.

Source discipline: The operative prior art for this analysis is the material listed in the Prior Art portion of the Google Patents record — the two Patent Citations and the two Non-Patent Citations. I treat the page's "Cited By" and "Similar Documents" lists as non-prior-art context (they post-date the 1996-02-16 priority date) and I label them as such wherever used.

---

1. Legal framework applied

• § 103 / Graham v. John Deere (1966): scope and content of the prior art; differences between prior art and claims; level of ordinary skill; and secondary considerations.
• KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007): the TSM test is not the sole test; a combination is obvious where the improvement is a "predictable variation," a "use of a known technique to improve a similar device in the same way," or an "obvious to try" solution from a finite number of identified, predictable options.
• MPEP 2143 / 2144: recognized rationales include (A) combining known elements with predictable results, (B) simple substitution, (C) known technique to improve a similar device, (D) applying a known technique to a known device ready for improvement, (F) design incentives, (G) predictable variations; and design choice (2144) and optimization of ranges (2144.04).
• MPEP 2129 / In re Nomiya: the applicant's own characterization of a reference in the specification is a binding admission of what that reference discloses. This matters greatly here, because the '694 specification itself characterizes the JSME Data Book.

Level of ordinary skill (POSITA), as I would construe it from the intrinsic record: a person with a bachelor's degree in mechanical or electrical engineering (or equivalent), plus 2–4 years' experience in thermal management of electronic packages, familiar with fin-array heat-transfer design (fin spacing, thickness, pressure drop, boundary-layer/leading-edge effects) and with multi-chip-module cooling. This is a modest skill level, which broadens what counts as an obvious design choice.

Analogous art: All three of-record references are analogous. The JSME Data Book is the field's foundational text; US 5,097,385 is electronics cooling hardware; US 5,191,511 is LSI packaging. Even on the narrower "reasonably pertinent to the problem" test (In re Bigio), all three qualify — they are directed at, or plainly useful for, dissipating heat from LSI packages.

---

2. The prior art of record — what each reference actually teaches

The key observation: the only of-record reference that discloses fin geometry varying along the height of a heat-transfer member with the upper region populated and the lower region depopulated/substituted is US 5,097,385. US 5,191,511 supplies no element of the fin-density limitations at all — it goes only to the claim-1/claim-4 preamble environment (multilayer ceramic board, LSI packages, a heat-radiation fin on the package).

---

3. Claim-by-claim §103 assessment

3.1 Class A — claims vulnerable on the of-record art

Claims 8 and 9 (impingement from above, then flow to the second fin) — Strong.
Claim 8 (dep. from 4) requires cooling fluid "impinged onto the plurality of elemental fins of said first cooling fin from an upper portion… and after cooling the first cooling fin, the cooling fluid is allowed to flow to the second cooling fin." That is the stated purpose of US 5,097,385: coolant delivered through cylindrical ducts/jets from above, flowing over piston fin sections and then into the channels between adjacent structure. The motivation is explicit in the reference itself ("jet enhanced convective cooling"), and the combination with the JSME duct baseline is a "known technique to improve a similar device" (MPEP 2143(C)). This pair of claims is the cleanest obviousness case on the face of the record. The patent's own FIG. 7 embodiment (nozzles 12a/12b over the central column) is a direct read on this.

Claim 12 (mixing effect due to interaction of the first and second fins) — Strong.
This is a result/functional limitation with no structural addition. Where the recited result is the inherent and expected consequence of the claimed structure, it cannot confer patentability (In re Schreiber; MPEP 2114). Independently, the mixing effect is expressly discussed in Hitachi application JPH02168697A (published 1990-06-28, "冷却装置 for 電子機器"), which explains that at the merge of main and bypass flows "the heat-transfer characteristic of the heat-generating body in that vicinity improves due to the mixing effect (ミキシング効果)" — and which criticizes the prior art precisely for lacking a way to control the velocity distribution of the coolant around the heat-radiation fins ("放熱フィン廻りの冷却流体の速度分布を制御する構成とはなっていなかった"). Source: https://patents.google.com/patent/JPH02168697A/en. (Caveat: JPH02168697A appears in the page's "Similar Documents," not the citation list. I am using it as corroboration of what the field already knew, not as prior art "of record.")

Claims 3 and 6 and 7 (fin shape; wider spacing at bottom; thinner fins at bottom) — Strong, on design-choice/optimization grounds.

• Claim 3's "circular or rectangular" alternative is a bare shape election; US 5,097,385 already shows rectangular, diamond and brick fin shapes.
• Claim 6 (larger spacing at upper portion, smaller at lower) and claim 7 (greater thickness at upper portion, lesser at lower) recite purely dimensional, position-dependent variables. Fin spacing and fin thickness are the two canonical control variables in fin-array design taught by the JSME text (heat-transfer coefficient vs. pressure drop trade-off). The specification itself concedes these as mere alternatives: "the thickness of the elemental fins at the lower portion may be made smaller or the spacing between them may be rendered wider." Under MPEP 2144.04 (optimization of a result-effective variable) and MPEP 2144 (design choice), these claims are prima facie obvious absent a showing of criticality or unexpected results. The patent provides no comparative data establishing criticality of any spacing or thickness value.

3.2 Class B — claims with a viable but contestable §103 case

Claim 1 (uniform first fin + gradient second fin, adjacent).

• Taught: the "first cooling fin … uniform density" is the admitted JSME/p. 206 baseline. The "second cooling fin … first fin density at an upper portion … second, lower fin density at a lower portion close to a fin base" is, in substance, US 5,097,385's finned upper piston sections over the lower header/feet region (fins concentrated above, flow access maximized below).
• Missing: neither reference discloses two adjacent fin structures of these two different characters mounted in series in a common duct. This is the crux, and it requires a synthesis that neither reference states in so many words.
• Motivation to combine (the strongest available rationale): (i) the problem is identified in the prior art itself — JPH02168697A describes exactly the downstream temperature-rise problem for series-arranged heat generators (a 1989 Hitachi filing); (ii) the physics is elementary and stated in the JSME text (higher flow velocity in the free space above the fins, higher pressure drop with denser fin packing, leading-edge enhancement); (iii) KSR endorses "a use of a known technique to improve a similar device in the same way" and "predictable variation." A POSITA with the duct-velocity-profile problem in hand would predictably draw the lower fins back to redirect cold, fast air to the fin bases.
• Assessment: obvious with a reasonable expectation of success, but the rejection depends on motivation reasoning rather than a single clean disclosure. This is precisely the kind of case where allowance below turned on motivation/teaching-away, not on novelty of any single element.

Claims 5, 13, 14 (gradient on the second fin; zero fins at the bottom).
Claim 13 hardens claim 1 to no elemental fins at the lower portion. US 5,097,385's pistons present an upper finned region with a fin-free lower header in contact with the chip. That is close in substance to claim 13's zero-density lower region, and the rationale for removing (rather than merely thinning) the lower fins — maximizing unobstructed flow to the base/highest-resistance region — is exactly the rationale US 5,097,385 gives for its jet/channel geometry. Claim 14 is claim 13's requirement grafted onto claim 4's mixed-orientation geometry. Claim 5 is claim 4 plus the gradient of claim 1. Same analysis: well-supported obviousness theory, dependent on motivation reasoning.

Claims 4, 10, 11 (the geometry/mixed-orientation and fin-level claims).

• Claim 4 requires no density gradient at all — only that a first fin carry perpendicular (vertical-plate) fins on the base and an adjacent second fin carry parallel (stacked-plate) fins on a perpendicular support fin. Vertical-plate fin sinks and stacked-plate fin sinks were each separately ubiquitous; placing them side by side in a duct is a simple substitution / predictable variation under MPEP 2143(A)–(B). IBM's piston (upper finned shaft above a rectangular header/feet with channels between the feet) and the JSME plate-fin baseline together supply the two orientations. Claim 4 is, in my view, the weakest claim in the patent.
• Claims 10 and 11 are the same concept drawn at the fin level (denser top, sparser bottom). Claim 10's closest art remains US 5,097,385's upper-portion microfins, but the IBM member is a liquid-cooled piston, not an air-cooled duct fin — a difference in kind that a defendant would have to bridge with motivation. Claim 11's "parallel, rectangular elemental fins … attached to a support fin mounted perpendicularly to the fin base" is a commonplace stacked-plate heat sink; adding the top-denser gradient is the same predictable variation.

3.3 What US 5,191,511 contributes

Honestly: little for §103 purposes. It supplies the claim 1/claim 4 preamble environment — a chip carried on a multi-layer ceramic board with an area-array of terminals and a heat-radiation fin on the package front surface — which corresponds to the '694 FIG. 1/2 multi-chip module on multilayer ceramic board 2 with micropins 4. It does not teach or suggest any height-varying fin density, fin spacing, or fin thickness. Any §103 rejection built on '511 would have to use it only as background/supporting art for the "heat generating units"/"electronic device" environment; a rejection resting on '511 for the fin features would fail.

---

4. The two-reference combinations, stated in rejector-ready form

---

5. Graham secondary considerations — what the record does and does not show

• Unexpected results — not established. The '694 specification's FIG. 11/12 experiment (34 mm × 93 mm duct, 0.8 m³/min, 10 fins, 0.9 W memory LSIs, 13.5/6.2/5.6 W logic LSIs) shows 2–4 m/s flow across the whole fin region and a temperature rise extending to the fin tops. But there is no comparative control — no equal-volume uniform-density fin array against which the claimed geometry is benchmarked. A design working as predicted is not an "unexpected result" (MPEP 716.02). This is a significant evidentiary gap for the patentee.
• Long-felt need — partially, and without nexus. The specification argues decades of difficulty in cooling high-density LSIs on MCMs. But the problem statement (duct velocity stratification, downstream starvation, pressure-drop ceiling, the mixing effect) is recited almost verbatim in JPH02168697A (1989) and is inherent in the JSME textbook treatment. A need known in the art, addressed by the art's own recognized levers, does not rebut obviousness.
• Commercial success / licensing: nothing in the record. (Note the chain of title shows only the original 1997 assignment to Hitachi, Ltd.; no reissue and no assignment to an asserting entity appears — consistent with the earlier litigation finding.)
• Teaching away — none shown. US 5,097,385's liquid-impingement context might arguably be urged as "different art," but KSR requires that a reference criticize, discredit or discourage the solution; it does not. At most this supports a non-analogous-art argument, which fails here on the Bigio test.
• Design choice / result-effective variable: fin density, spacing and thickness along the fin height are the paradigm "result-effective variables" for fin-array thermal/hydraulic performance.

---

6. Bottom line and honest limits of this opinion

Most likely obvious on the of-record prior art: claims 3, 6, 7, 8, 9 and 12 — plus, on strong motivation reasoning, 13 and 5 (and 14 as its claim-4 analogue).

Viable but contestable: claims 1 and 10 (the core gradient concept) — obvious, but because the two of-record references must be combined and neither states the two-adjacent-fin synthesis, the rejection turns on the sufficiency of the motivation articulation. Claim 4 is the most vulnerable to being held anticipated or obvious on ordinary plate-fin/stacked-fin art alone.

Claims I would not call clearly obvious on this record alone: 11, and claim 4/14 to the extent the mixed-orientation + adjacency requirement is read narrowly — the of-record references lack a clean disclosure of the parallel-fin-on-erect-support-fin geometry adjacent to a perpendicular-fin fin.

Two structural weaknesses in this §103 record, stated plainly:

1. The citation list is thin for the core concept. Only US 5,097,385 speaks to height-varying fin density, and it does so in a liquid-dielectric impingement TCM, not an air-cooled duct fin array. A rejection of claim 1 built only on the of-record art is defensible but not overwhelming — which is a plausible explanation for why the examiner, having both references before him, nonetheless allowed the claims.

2. Highly relevant, non-cited art exists within the same corporate family and same field, which a fuller obviousness challenge would use against the "two adjacent fins / control of the coolant velocity distribution around heat-radiation fins" concept:

• US 5,077,601 (Hitachi; filed 1989-09-07; names Takahiro Daikoku — a named inventor here) — a cooling duct in which fins are attached to heat-generating components arranged in series along the flow, with both a main flow path into the assemblies and an auxiliary flow path directed substantially vertically to impinge upon the heat-generating component. Source: https://patentimages.storage.googleapis.com/51/fd/03/d7d848bde18238/US5077601.pdf
• JPH02168697A (Hitachi; JP 14087689A, 1989) — velocity-distribution control and mixing-effect discussion, quoted above.
• US 5,705,854 A (Hitachi; priority 1995-04-14; granted 1998-01-06, "Cooling apparatus for electronic device") — appears in the page's "Similar Documents." Because its US filing date is not confirmed and it claims only a foreign (JP) priority, I flag it as potentially §102(e)/§103 art but do not rely on it; pre-AIA §102(e) does not reach back to a foreign priority date (the Hilmer doctrine). It should be verified before use.

These are in the page's "Cited By"/"Similar Documents" sections, i.e., not prior art on this record, and I am identifying them as search leads for a fuller §103 case, not as grounds.

If you need a defensible invalidity position rather than an academic one, the gaps to fill are: (a) a pre-1996 reference showing an air-cooled, series-arranged fin array with fin density (or spacing/thickness) deliberately reduced near the fin base to redistribute flow to the base; and (b) a reference showing the mixed perpendicular/parallel fin orientation in series. The of-record citations do not fully supply (a) or (b).