1014-00003 · Optiwave Semiconductor Corp · Post-Markman Assessment

post-markman-viability-567-patent.docx DOCUMENT Litigation

62 / 83 ← → ✕

PRIVILEGED AND CONFIDENTIAL — ATTORNEY WORK PRODUCT

This memorandum is protected by the attorney-client privilege and the work product doctrine. It was prepared in anticipation of litigation and reflects the mental impressions, conclusions, opinions, and legal theories of counsel. It is not to be disclosed to any third party without the prior written authorization of Calderwood & Harkness LLP.

CALDERWOOD & HARKNESS LLP One Embarcadero Center, 34th Floor San Francisco, California 94111 Telephone: (415) 555-7200 Facsimile: (415) 555-7201

MEMORANDUM

February 22, 2023

TO: Alan Ngo, Responsible Partner, intellectual Property Group

FROM: Omar Siddiqui, Lead Associate, intellectual Property Group

CC: Christy Bauer, Associate; Marcus Delgado, Billing Partner

RE: Post-Markman Infringement Viability Analysis — U.S. Patent No. 10,234,567 ("the '567 Patent") — Asserted Claims 1, 4, 7, 12, and 15

MATTER: Optiwave Semiconductor Corp. v. Lumos Photonics, Inc., Case No. 5:22-cv-03417-EJD (N.D. Cal.)

C&H MATTER NO.: 1014-00003

CLIENT: Optiwave Semiconductor Corp.

Right-click to update Table of Contents

I. Executive Summary

The purpose of this memorandum is to assess the viability of Optiwave Semiconductor Corp.'s ("OWS" or "Plaintiff") infringement theories for U.S. Patent No. 10,234,567 ("the '567 Patent") following the Markman Order issued on February 17, 2023, by the Honorable Edward J. Torvelis of the United States District Court for the Northern District of California. This analysis addresses all five asserted claims of the '567 Patent — Claims 1, 4, 7, 12, and 15 — and evaluates whether any viable path to establishing infringement remains in light of the Court's claim constructions.

The Court's 34-page Markman Order construed eleven disputed claim terms across the four patents-in-suit in this action. Of those eleven terms, the Court adopted Defendant Lumos Photonics, Inc.'s ("Lumos" or "Defendant") proposed construction — or a construction substantially aligned with Defendant's position — on eight of the eleven terms. The Court adopted Plaintiff OWS's proposed construction on two terms and arrived at a modified construction that departed from both parties' proposals on one term. This outcome represents a significant and, candidly, decisive setback for OWS's infringement case across the patent portfolio.

For the '567 Patent specifically, the critical disputed term was "adiabatically coupled," which appears in independent Claim 1 and is inherited by all four asserted dependent claims (Claims 4, 7, 12, and 15). The parties advanced materially different constructions of this term:

• OWS's proposed construction: "coupled such that optical mode transitions occur without significant power loss."

• Lumos's proposed construction: "coupled through a gradual, continuous taper that prevents excitation of higher-order modes throughout the entire transition region."

The Court adopted a construction closely aligned with Defendant's proposal, construing "adiabatically coupled" to mean: "coupled through a gradual, continuous taper designed and configured to prevent excitation of higher-order modes over the entire transition length." The Court's construction introduced minor modifications from Lumos's proposal — replacing "throughout the entire transition region" with "over the entire transition length" and adding the phrase "designed and configured to" — but substantively embraced Defendant's structural interpretation of the term. The Court's construction imposes structural requirements on the coupling mechanism rather than defining it solely by its functional result, which was the cornerstone of OWS's infringement theory.

Bottom-line conclusion: Infringement theories for all five asserted claims of the '567 Patent (Claims 1, 4, 7, 12, and 15) are untenable under the Court's construction. Lumos's LumosBridge™ 800G Optical Engine and LumosEdge™ 400ZR+ Pluggable Transceiver products employ a segmented (non-continuous) taper coupler architecture with discrete coupling sections. This design falls squarely outside the Court's narrowed construction, which requires a "gradual, continuous taper" over the "entire transition length." The segmented taper design used by Lumos is structurally distinct from the continuous adiabatic taper contemplated by the claims as construed.

Moreover, doctrine-of-equivalents arguments are foreclosed by prosecution history estoppel. The "adiabatically coupled" limitation was added to Claim 1 by amendment during prosecution of the '567 Patent application specifically to overcome a prior art rejection based on the Tanaka patent (U.S. Patent No. 9,456,012). The applicants' prosecution remarks explicitly distinguished the invention by reference to a "gradual, continuous taper transition," thereby disclaiming non-continuous coupling approaches. Under Festo Corp. v. Fenmore & Co., 535 U.S. 722 (2002), this narrowing amendment creates a presumption of estoppel that OWS cannot rebut, as segmented taper couplers were well-known and foreseeable at the time of the amendment.

Recommendation: Infringement of the '567 Patent should not be pursued further in this litigation. This assessment is consistent with, and reinforces, the broader recommendation for settlement discussions set forth in the consolidated Post-Markman Impact Assessment Memorandum dated February 21, 2023.

II. Background

A. The '567 Patent Overview

U.S. Patent No. 10,234,567, entitled "Integrated Waveguide Coupler with Tapered Mode Converter for Silicon Photonics Platforms," was filed on March 14, 2017, and issued on March 19, 2019, after prosecution before the United States Patent and Trademark Office. The patent is assigned to Optiwave Semiconductor Corp., our client.

The named inventors on the '567 Patent are: (1) Dr. Kevin Fong, who serves as Chief Technology Officer and Co-Founder of OWS; (2) Dr. Priya Ramanathan, who serves as Chief Executive Officer and Co-Founder of OWS; and (3) Hideo Tanaka, a former OWS engineer who was employed by the company from 2016 to 2019. As discussed in further detail below, Mr. Tanaka's prior patent — U.S. Patent No. 9,456,012 — played a significant role in the prosecution of the '567 Patent and, ultimately, in the claim construction at issue.

The '567 Patent is directed to an integrated optical coupler designed for use on silicon photonics platforms. More specifically, the invention concerns a tapered silicon waveguide that is adiabatically coupled to a silicon nitride (SiN) waveguide layer, thereby enabling efficient fiber-to-chip and chip-to-fiber optical coupling in high-speed optical transceivers. The technology addresses a well-known challenge in silicon photonics: the large mode-size mismatch between standard single-mode optical fibers and the sub-micron silicon waveguides used in integrated photonic circuits. The '567 Patent's approach uses a tapered mode converter to gradually transform the optical mode from the larger fiber mode to the smaller waveguide mode (and vice versa), minimizing insertion loss and back-reflection.

OWS's commercial silicon photonics chips incorporating the patented technology are fabricated by Pinnacle Semiconductor Foundry Co., Ltd. on its 45nm silicon-on-insulator (SOI) process. These chips are deployed in OWS's own optical transceiver products and are also licensed to certain third-party transceiver manufacturers under OWS's technology licensing program.

B. Asserted Claims

OWS asserts Claims 1, 4, 7, 12, and 15 of the '567 Patent in this action. Claim 1 is the principal independent claim, and Claims 4, 7, 12, and 15 each depend — directly or indirectly — from Claim 1.

Claim 1 recites, in relevant part:

"An integrated optical coupler comprising a tapered silicon waveguide having a tip width of less than 200 nanometers adiabatically coupled to a silicon nitride waveguide layer, said silicon nitride waveguide layer disposed on a silicon photonics platform, wherein the tapered silicon waveguide is configured to convert an optical mode from a first mode profile to a second mode profile for coupling between an optical fiber and a photonic integrated circuit."

(Emphasis added to the disputed term.)

Claim 4 depends directly from Claim 1 and adds a limitation requiring a specific taper length-to-width ratio for the tapered silicon waveguide. Claim 7 depends directly from Claim 1 and adds a limitation regarding the material composition of the silicon nitride waveguide layer, specifying a refractive index range. Claim 12 depends directly from Claim 1 and adds a limitation directed to integration with a photodetector on the same silicon photonics platform. Claim 15 depends from Claim 12 (and thus indirectly from Claim 1) and adds a limitation requiring a specific germanium-on-silicon photodetector integration with the waveguide coupler.

Critically, all five asserted claims inherit the "adiabatically coupled" limitation from Claim 1. Because the Court's construction of this term is outcome-determinative — as explained in detail in Sections III and IV below — the infringement analysis for all five claims rises or falls together.

C. Accused Products

OWS accuses two Lumos Photonics products of infringing the asserted claims of the '567 Patent:

1. LumosBridge™ 800G Optical Engine. The LumosBridge™ 800G is a co-packaged optics (CPO) engine designed for integration with high-bandwidth switch ASICs in hyperscale data center environments. The product was first shipped commercially in the first quarter of 2022. Based on publicly available pricing information and discovery responses, the unit price of the LumosBridge™ 800G ranges from approximately $1,200 to $1,800 per unit, depending on configuration and volume tier.

2. LumosEdge™ 400ZR+ Pluggable Transceiver. The LumosEdge™ 400ZR+ is a QSFP-DD form factor pluggable optical transceiver designed for metro and data center interconnect (DCI) applications. The product has been in production and commercially available since the third quarter of 2021. Unit pricing for the LumosEdge™ 400ZR+ ranges from approximately $600 to $900 per unit.

Based on Lumos's annual report on Form 10-K for fiscal year 2022 and discovery responses produced in this litigation, the combined revenue attributable to the two accused products for calendar year 2022 was approximately $148.6 million. Lumos's total fiscal year 2022 revenue was $312.8 million, meaning the accused products represent approximately 47.5% of Lumos's overall revenue.

Lumos Photonics, Inc. (Nasdaq: LMPH) is a Delaware corporation headquartered in Austin, Texas. The company operates a semiconductor fabrication facility in Hillsboro, Oregon, and maintains a sales and engineering office at 4750 Great America Parkway, Suite 320, Santa Clara, California 95054. At the time OWS filed the complaint in this action (June 3, 2022), Lumos's market capitalization was approximately $1.87 billion.

D. Pre-Markman Infringement Theory

Prior to the Markman hearing, the viability of OWS's infringement case against the '567 Patent was assessed in detail in a pre-suit infringement analysis memorandum prepared by this team, dated April 18, 2022 (the "Pre-Suit Memo"). That analysis identified the "adiabatically coupled" limitation as a potential vulnerability in OWS's infringement theory, but assessed the risk as manageable under the then-prevailing litigation strategy.

OWS's pre-Markman infringement theory relied on the broader proposed construction of "adiabatically coupled" — namely, "coupled such that optical mode transitions occur without significant power loss." Under this functional construction, the term would encompass any coupling mechanism that achieved low-loss mode transfer between the silicon waveguide and the silicon nitride waveguide layer, regardless of the specific structural approach used to achieve that result. Critically, this construction would have encompassed Lumos's segmented taper design, because the segmented taper does, in fact, achieve low-loss coupling — it simply does so through a different structural mechanism than a continuous adiabatic taper.

Under OWS's proposed construction, the focus of the infringement inquiry was on the functional result (low power loss during mode transitions) rather than the structural requirement (a gradual, continuous taper). This approach was grounded in the specification's recurrent emphasis on the purpose of adiabatic coupling — i.e., minimizing power loss — and was supported by the declaration of Dr. Harold Whitfield of Whitfield Technical Consulting Group, Inc. (Palo Alto, California), who opined that persons of ordinary skill in the art of integrated photonics understand "adiabatic coupling" primarily by reference to its functional result.

The Pre-Suit Memo flagged the risk that the Court could adopt a narrower, structurally-focused construction of "adiabatically coupled," which would be more consistent with Lumos's expected defensive position. That risk was assessed as moderate at the time, given the breadth of the specification's disclosure and the general principle that claim terms should be given their plain and ordinary meaning to a person of ordinary skill in the art ("POSITA"). However, the Pre-Suit Memo acknowledged that the prosecution history — specifically the amendment adding the "adiabatically coupled" limitation to overcome the Tanaka prior art — presented a meaningful counterargument that could persuade the Court to adopt a narrower construction.

As discussed in the following section, the risk identified in the Pre-Suit Memo has materialized in the most consequential way possible.

III. The Court's Claim Construction — "Adiabatically Coupled"

A. Summary of the Parties' Positions

The parties presented sharply divergent proposed constructions for the term "adiabatically coupled" as it appears in Claim 1 of the '567 Patent. The dispute centered on whether the term should be construed functionally — defining the coupling by the result it achieves — or structurally — defining the coupling by the physical mechanism required to achieve it.

1. OWS's Proposed Construction

OWS proposed that "adiabatically coupled" be construed to mean: "coupled such that optical mode transitions occur without significant power loss."

OWS's construction was grounded in several arguments. First, OWS contended that the term should be given its plain and ordinary meaning to a person of ordinary skill in the art of integrated photonics. OWS argued that in the field, "adiabatic" is understood as a term describing the efficiency of a mode transition — i.e., a transition that occurs slowly enough relative to the coupling length that the system remains in its fundamental mode throughout the process, avoiding excitation of parasitic higher-order modes and thereby minimizing power loss. Under this interpretation, any coupling mechanism that achieves this functional result — low-loss, single-mode transition — qualifies as "adiabatic," regardless of the specific physical structure used.

Second, OWS argued that the specification of the '567 Patent supports this functional interpretation. The specification repeatedly describes the purpose of adiabatic coupling as minimizing insertion loss and maintaining efficient mode transfer. While the specification discloses specific embodiments using a gradual, continuous taper, OWS contended that the specification should not be read to limit the claims to those specific embodiments.

Third, OWS relied on the declaration and report of Dr. Harold Whitfield of Whitfield Technical Consulting Group, Inc. (Palo Alto, California), an expert in integrated photonics design. Dr. Whitfield opined that persons of ordinary skill in the art understand "adiabatic coupling" by reference to its functional result — efficient, low-loss mode transfer — and would not necessarily limit the term to a single structural implementation such as a continuous taper. Dr. Whitfield cited several textbooks and peer-reviewed publications that describe "adiabatic" coupling in functional terms.

2. Lumos's Proposed Construction

Lumos proposed that "adiabatically coupled" be construed to mean: "coupled through a gradual, continuous taper that prevents excitation of higher-order modes throughout the entire transition region."

Lumos's construction was grounded in both intrinsic and extrinsic evidence. Lumos argued, first, that the prosecution history compelled a structural interpretation. The "adiabatically coupled" limitation was not present in the original application as filed on March 14, 2017. It was added by amendment during prosecution in response to the second office action, which rejected the claims under 35 U.S.C. § 102(a)(1) as anticipated by the Tanaka prior art (U.S. Patent No. 9,456,012, issued September 6, 2016). In the applicants' remarks accompanying the second amendment, OWS's prosecution counsel explicitly distinguished the invention from the Tanaka prior art by stating that the present invention requires "adiabatic coupling through a gradual, continuous taper transition" — language that Lumos argued constituted a clear disclaimer of non-continuous coupling approaches.

Second, Lumos pointed to the specification's consistent description of the taper coupler in exclusively gradual and continuous terms. Every embodiment disclosed in the specification employs a continuous, smoothly-varying taper profile. No embodiment discloses or suggests a segmented, stepped, or discontinuous taper. Lumos argued that the specification thus confirms the structural meaning of "adiabatically coupled" as requiring a gradual, continuous taper.

Third, Lumos presented extrinsic evidence from the technical literature defining "adiabatic" in the specific context of waveguide coupling. The term "adiabatic" in waveguide optics, Lumos argued, has a well-established meaning that encompasses not just the functional result but also the structural requirement of gradual, continuous variation to maintain single-mode propagation throughout the transition.

Fourth, Lumos relied on the testimony and report of Dr. Raymond Ng, who serves as Of Counsel and Technical Consultant to Lumos's counsel, Brannick Howell & Stein LLP. Dr. Ng is a former professor of photonics at the University of Colorado Boulder, with over 25 years of experience in integrated photonics research and waveguide design. Dr. Ng testified at the Markman hearing held on January 30, 2023, and opined that a person of ordinary skill in the art would understand "adiabatically coupled" in the context of the '567 Patent to require a gradual, continuous taper that prevents excitation of higher-order modes. Dr. Ng distinguished between the thermodynamic usage of "adiabatic" (referring generally to a process without energy exchange) and the specific usage in waveguide optics (referring to a gradual, continuous transition that maintains single-mode propagation).

B. The Court's Construction

In its Markman Order dated February 17, 2023, the Court construed "adiabatically coupled" as follows:

"coupled through a gradual, continuous taper designed and configured to prevent excitation of higher-order modes over the entire transition length"

See Markman Order at 12.

The Court's construction closely tracks Lumos's proposed construction, with two minor modifications. First, the Court replaced "throughout the entire transition region" with "over the entire transition length" — a change that is substantively immaterial but reflects the Court's preference for the precision of "transition length" over the slightly broader "transition region." Second, the Court added the phrase "designed and configured to" before "prevent excitation of higher-order modes," which introduces a design-intent qualification but does not alter the fundamental structural requirement of a gradual, continuous taper. Substantively, the Court adopted Defendant's structural interpretation of the term.

The Court's reasoning, as set forth on pages 8 through 15 of the Markman Order, proceeded through the standard claim construction framework established in Phillips v. AWH Corp., 415 F.3d 1303 (Fed. Cir. 2005) (en banc), giving primacy to intrinsic evidence.

Intrinsic Evidence — Specification. The Court found that the specification of the '567 Patent describes the taper coupler in exclusively gradual, continuous terms. The Court noted that every disclosed embodiment employs a continuous taper with a smoothly varying width profile, and that the specification contains no disclosure of a segmented, stepped, or otherwise discontinuous taper. The Court observed that while claims are not ordinarily limited to the disclosed embodiments, the specification's uniform treatment of the taper as continuous — combined with the prosecution history discussed below — supported a structural interpretation of "adiabatically coupled."

Intrinsic Evidence — Prosecution History. The Court found the prosecution history to be particularly instructive, and indeed dispositive, on the meaning of "adiabatically coupled." The Court recounted the prosecution timeline in detail: the original application did not include the "adiabatically coupled" limitation; the limitation was added by amendment in response to the second office action, which maintained a § 102(a)(1) rejection based on the Tanaka prior art. The Court quoted the applicants' remarks accompanying the amendment, in which prosecution counsel stated that "the present invention is distinguished from Tanaka by the requirement of adiabatic coupling through a gradual, continuous taper transition, which prevents excitation of higher-order modes — a feature absent from Tanaka's non-adiabatic coupler design." Markman Order at 13. The Court found that these remarks created a clear disclaimer of non-continuous coupling approaches, and that the prosecution history narrowed the scope of "adiabatically coupled" to require the gradual, continuous taper structure described in both the specification and the applicants' own characterization of the term during prosecution.

Extrinsic Evidence. The Court credited Dr. Ng's testimony regarding the structural meaning of "adiabatic" in the specific context of waveguide coupling, finding it consistent with and confirmatory of the intrinsic record. The Court found Dr. Whitfield's purely functional interpretation to be inconsistent with the intrinsic record — particularly the prosecution history — and therefore declined to adopt it. The Court emphasized that while Dr. Whitfield's testimony may reflect a usage of "adiabatic" in some broader contexts, the intrinsic evidence of this particular patent establishes a narrower, structural meaning.

Rejection of OWS's Plain and Ordinary Meaning Argument. The Court explicitly rejected OWS's argument that "adiabatically coupled" should be given its broadest plain and ordinary meaning to a POSITA. The Court acknowledged that the term may have a broader functional meaning in some contexts, but found that the prosecution history narrowed the claim scope and that the plain and ordinary meaning argument could not override the clear disclaimer established during prosecution. See Markman Order at 14 (citing Omega Eng'g, Inc. v. Raytek Corp., 334 F.3d 1314, 1323 (Fed. Cir. 2003) (noting that "the prosecution history limits the interpretation of claim terms so as to exclude any interpretation that was disclaimed during prosecution")).

C. Impact on Scope of "Adiabatically Coupled"

The Court's construction of "adiabatically coupled" imposes four distinct requirements that any accused coupler design must satisfy:

1. A gradual taper — the taper must vary smoothly and gradually in width, not abruptly or in discrete steps.

2. A continuous taper — the taper must be continuous, without segmentation, discontinuities, or discrete coupling sections. A taper composed of discrete segments separated by transition discontinuities does not satisfy this requirement.

3. Designed and configured to prevent higher-order mode excitation — the taper must be designed and configured such that higher-order optical modes are not excited during the mode transition. This is a design-intent requirement that reinforces the structural character of the limitation.

4. Over the entire transition length — the above properties must be maintained over the entire length of the taper transition, not merely at discrete points or in isolated sections.

The critical consequence of this construction is that it imposes a structural limitation on the coupling mechanism, not merely a functional one. A coupler that achieves low-loss mode transfer through a mechanism other than a gradual, continuous taper — even if it achieves the same functional result (minimal power loss) — does not satisfy the "adiabatically coupled" limitation as construed. Specifically, any coupler design using discrete segments, stepped transitions, or non-continuous taper profiles falls outside the scope of the Court's construction.

This is precisely the situation with Lumos's accused products, as analyzed in the following section.

IV. Claim-by-Claim Infringement Analysis Under the Court's Construction

A. Claim 1

Claim 1 is the independent claim from which all other asserted claims depend. The dispositive limitation for purposes of the infringement analysis is "adiabatically coupled," as construed by the Court.

1. The Dispositive Limitation: "Adiabatically Coupled"

Lumos's Accused Design. Based on the discovery produced to date — including Lumos's technical design documents (LMPH-000147523 through LMPH-000147891), process design kit specifications, and internal engineering presentations — as well as publicly available technical documentation (including Lumos's published papers, conference presentations, and product datasheets for the LumosBridge™ 800G and LumosEdge™ 400ZR+ products), the accused products employ a segmented taper coupler architecture that is structurally distinct from the continuous adiabatic taper required by the Court's construction.

Specifically, Lumos's segmented taper coupler design employs a series of discrete coupling sections (segments) rather than a single gradual, continuous taper. Each segment is a distinct waveguide section with its own width and length parameters. The segments are separated by transition discontinuities — abrupt changes in waveguide geometry between adjacent segments. The design achieves low-loss coupling through a fundamentally different mechanism than a continuous adiabatic taper: rather than maintaining continuous mode evolution along a smooth taper profile, Lumos's approach relies on optimized segment geometry — carefully selecting the width, length, and spacing of each discrete segment to achieve constructive interference and efficient power transfer across the aggregate structure.

This segmented taper approach was described by Dr. Nikolai Petrov, Lumos's Vice President of Engineering, Photonics Division, in Lumos's technical publications, including a 2021 conference paper presented at the Optical Fiber Communication Conference (OFC). Dr. Petrov's description of the LumosBridge™ coupler design is consistent with the segmented taper coupler methodology described in the Chen & Liu reference — "Segmented Taper Coupler Design for Low-Loss Fiber-to-Chip Coupling," Journal of Photonic Integration, Vol. 14, No. 3, pp. 217–229 (2016) — which was produced by Lumos in discovery and is referenced in Lumos's prior art contentions. The Chen & Liu paper demonstrates that segmented taper designs were well-established in the art prior to the filing of the '567 Patent, and that they represent a recognized alternative to continuous adiabatic tapers.

Internal Lumos engineering documents produced in discovery (LMPH-000148201 through LMPH-000148367) further confirm the segmented nature of the coupler design. These documents include design specifications showing the individual segment parameters (widths, lengths, and gap dimensions) for the waveguide coupler used in both accused products. The documents refer to the coupler as a "segmented mode converter" and do not use the term "adiabatic" to describe the coupling mechanism.

Analysis. Lumos's segmented taper coupler does not meet the "gradual, continuous taper" requirement of the Court's construction. The discontinuities between segments mean that the taper is neither gradual nor continuous "over the entire transition length." Instead, the taper consists of a plurality of discrete sections with abrupt width changes between them. Even though each individual segment may itself have a relatively smooth profile, the aggregate structure is not a single continuous taper but rather a series of discrete elements. The Court's construction requires continuity "over the entire transition length," and the presence of segmentation discontinuities at multiple points along the transition defeats this requirement.

Furthermore, the mode conversion mechanism in Lumos's design is fundamentally different from the adiabatic mode evolution contemplated by the claims. Rather than maintaining the optical signal in its fundamental mode throughout a gradual, continuous transition, Lumos's segmented design relies on discrete coupling events at each segment boundary. This is a structurally distinct approach that falls outside the scope of the "adiabatically coupled" limitation as construed.

Literal Infringement: Not met. The segmented taper design used in the LumosBridge™ 800G and LumosEdge™ 400ZR+ products is excluded by the Court's construction of "adiabatically coupled." Lumos's coupler is not "coupled through a gradual, continuous taper designed and configured to prevent excitation of higher-order modes over the entire transition length."

Remaining Claim Elements. Even assuming arguendo that all other elements of Claim 1 are met — including the "tapered silicon waveguide having a tip width of less than 200 nanometers," the "silicon nitride waveguide layer," the "silicon photonics platform," and the mode conversion configuration — the failure to meet the "adiabatically coupled" limitation is independently sufficient to defeat literal infringement of Claim 1. Under well-established law, every limitation of a claim must be met, either literally or by a substantial equivalent, for infringement to exist. See Laitram Corp. v. Rexnord, Inc., 939 F.2d 1533, 1535 (Fed. Cir. 1991). The failure to satisfy even a single limitation — here, "adiabatically coupled" — is fatal.

B. Claim 4 (Dependent on Claim 1)

Claim 4 depends directly from Claim 1 and adds a limitation requiring a specific taper length-to-width ratio for the tapered silicon waveguide. Because Claim 4 inherits all limitations of Claim 1, including the "adiabatically coupled" limitation, it necessarily requires that the coupler satisfy the Court's construction of that term.

Because Claim 1 is not infringed under the Court's construction for the reasons set forth above, Claim 4 is likewise not infringed. A dependent claim cannot be infringed unless the independent claim from which it depends is also infringed. See Wahpeton Canvas Co. v. Frontier, Inc., 870 F.2d 1546, 1552 (Fed. Cir. 1989).

Moreover, even if the Claim 1 analysis were not dispositive, the additional limitation of Claim 4 presents independent difficulties. The limitation requires a specific taper length-to-width ratio, which presupposes a single, continuous taper with a definable length and a definable width variation. Lumos's segmented taper geometry does not present a single continuous taper with a single definable length-to-width ratio in the manner contemplated by the claim. Instead, each discrete segment has its own length and width parameters, and the overall coupler structure does not have a single "taper length" or "taper width" that can be meaningfully compared to the claim limitation. This further supports the conclusion that Claim 4 is not infringed.

C. Claim 7 (Dependent on Claim 1)

Claim 7 depends directly from Claim 1 and adds a limitation regarding the material composition of the silicon nitride waveguide layer, specifying that the silicon nitride layer must have a refractive index within a defined range. As with Claim 4, Claim 7 inherits all limitations of Claim 1, including the "adiabatically coupled" limitation.

Because Claim 1 is not infringed under the Court's construction, Claim 7 is not infringed.

As an independent matter, the evidentiary record regarding Lumos's specific silicon nitride refractive index remains incomplete. While certain publicly available Lumos publications reference a SiN layer with optical properties that may or may not fall within the range specified in Claim 7, the detailed process specifications for Lumos's SiN deposition have not been fully produced or analyzed. However, this limitation is moot given the dispositive failure on the Claim 1 "adiabatically coupled" limitation, and further investigation of this issue is not warranted.

D. Claim 12 (Dependent on Claim 1)

Claim 12 depends directly from Claim 1 and adds a limitation directed to integration of the waveguide coupler with a photodetector on the same silicon photonics platform. Claim 12 inherits all limitations of Claim 1, including the "adiabatically coupled" limitation.

Because Claim 1 is not infringed under the Court's construction, Claim 12 is not infringed.

It bears noting that both accused Lumos products — the LumosBridge™ 800G and the LumosEdge™ 400ZR+ — do, in fact, integrate photodetectors on the same silicon photonics platform as the waveguide coupler. Under OWS's original infringement theory, the Claim 12 photodetector limitation was not expected to be a point of contention. However, the fact that Lumos's products satisfy this additional limitation is wholly immaterial given the dispositive failure on the "adiabatically coupled" limitation. Satisfaction of one claim element does not cure the failure to satisfy another.

E. Claim 15 (Dependent on Claim 1 via Claim 12)

Claim 15 depends from Claim 12, which in turn depends from Claim 1. Claim 15 adds a further limitation requiring a specific germanium-on-silicon (Ge-on-Si) photodetector integration with the waveguide coupler. As a claim that depends indirectly from Claim 1 through Claim 12, Claim 15 inherits all limitations of both Claim 1 and Claim 12, including the "adiabatically coupled" limitation.

Because Claim 1 is not infringed under the Court's construction, Claim 15 is not infringed. The analysis of the additional Ge-on-Si photodetector limitation of Claim 15 is unnecessary and has not been undertaken for purposes of this memorandum.

Summary Table: Claim-by-Claim Assessment

V. Doctrine of Equivalents Analysis

Even where literal infringement cannot be established, a patentee may sometimes rely on the doctrine of equivalents ("DOE") to capture accused products that perform substantially the same function, in substantially the same way, to achieve substantially the same result as the claimed invention. See Warner-Jenkinson Co. v. Harborline Enterprises Davis Chem. Co., 520 U.S. 17 (1997). However, as discussed below, the doctrine of equivalents is foreclosed in this case by prosecution history estoppel, and would in any event fail on the merits.

A. Prosecution History Estoppel

1. The Narrowing Amendment

The "adiabatically coupled" limitation was not present in the original claims of the '567 Patent application as filed on March 14, 2017. It was added by amendment during prosecution in response to the United States Patent and Trademark Office's second office action, which maintained a rejection of the claims under 35 U.S.C. § 102(a)(1) as anticipated by the Tanaka prior art — U.S. Patent No. 9,456,012, issued September 6, 2016.

It is noteworthy — and presents an awkward dynamic for OWS — that Hideo Tanaka, a named inventor on the '567 Patent, is also a co-inventor of the Tanaka prior art patent (U.S. 9,456,012). Mr. Tanaka obtained the '012 patent before joining OWS in 2016, and the '012 patent discloses a non-adiabatic coupler design that was cited by the examiner as anticipating the original claims of the '567 Patent application. The '567 Patent applicants were thus compelled to distinguish their invention from Mr. Tanaka's own prior work — a distinction that was accomplished by adding the "adiabatically coupled" limitation to narrow the claims beyond the scope of the Tanaka prior art.

The applicants' remarks accompanying the second amendment are directly relevant to the prosecution history estoppel analysis. In those remarks, OWS's prosecution counsel stated:

This statement constitutes a clear and unambiguous characterization of the scope of the "adiabatically coupled" limitation, explicitly defining it in terms of a "gradual, continuous taper transition." By these remarks, the applicants disclaimed coupling approaches that do not employ a gradual, continuous taper — including, critically, the segmented taper approach used by Lumos.

2. Application of *Festo*

Under the Supreme Court's decision in Festo Corp. v. Fenmore & Co., 535 U.S. 722 (2002), when a patent holder narrows a claim by amendment during prosecution for a reason related to patentability, there is a presumption that the amendment surrendered the territory between the original claim and the amended claim. This presumption of prosecution history estoppel prevents the patent holder from using the doctrine of equivalents to recapture the surrendered subject matter.

Here, the addition of the "adiabatically coupled" limitation was unquestionably a narrowing amendment made for reasons of patentability — specifically, to overcome the Tanaka prior art rejection. The original claims encompassed any coupler design meeting the other claim elements; the amended claims narrowed the scope to require adiabatic coupling. The territory surrendered by this amendment encompasses non-adiabatic coupling approaches, including segmented taper designs that do not employ a gradual, continuous taper.

3. Analysis of the *Festo* Rebuttal Exceptions

Under Festo, the presumption of prosecution history estoppel can be rebutted in three narrow circumstances. We analyze each in turn.

Exception 1: Unforeseeability. The patentee may rebut the presumption by showing that the alleged equivalent was unforeseeable at the time of the amendment. Here, segmented taper couplers were well-known and well-documented in the art at the time of the amendment (2017–2018). The Chen & Liu publication — "Segmented Taper Coupler Design for Low-Loss Fiber-to-Chip Coupling," Journal of Photonic Integration, Vol. 14, No. 3, pp. 217–229 (2016) — was published a full year before the original filing date of the '567 Patent and demonstrates that segmented taper designs were an established alternative to continuous adiabatic tapers in the field. Additional references in the prior art, including conference papers from the 2015 and 2016 OFC conferences, further confirm the foreseeability of segmented taper approaches. The alleged equivalent — a segmented taper coupler — was therefore entirely foreseeable at the time of the narrowing amendment. This rebuttal exception fails.

Exception 2: Tangential Relation. The patentee may rebut the presumption by showing that the rationale underlying the amendment bears only a tangential relation to the equivalent in question. Here, the amendment was made precisely to distinguish the prior art's coupling mechanism — i.e., to require adiabatic coupling through a gradual, continuous taper, as opposed to the non-adiabatic coupling disclosed in the Tanaka prior art. The equivalent at issue — Lumos's segmented taper coupler — is a coupler design that employs a non-continuous coupling mechanism. The rationale for the amendment (distinguishing non-adiabatic, non-continuous coupling) is directly and centrally related to the equivalent at issue (a segmented, non-continuous coupler). This rebuttal exception fails.

Exception 3: Other Reason. The patentee may rebut the presumption by showing some other reason suggesting that the patentee could not reasonably be expected to have described the insubstantial substitute in question. Here, no such reason exists. The patentee — OWS, through its inventors Dr. Fong, Dr. Ramanathan, and Mr. Tanaka — had extensive knowledge of the integrated photonics field, including familiarity with segmented taper designs. The applicants had ample opportunity to draft claims that encompassed segmented tapers, or to add claims directed to alternative coupler architectures, but chose not to do so. Indeed, the prosecution counsel's remarks specifically characterized the invention as requiring a "gradual, continuous taper transition," affirmatively excluding segmented approaches. This rebuttal exception fails.

4. Conclusion on Estoppel

Because all three Festo rebuttal exceptions fail, the presumption of prosecution history estoppel is unrebutted. OWS is estopped from asserting the doctrine of equivalents with respect to the "adiabatically coupled" limitation against Lumos's segmented taper coupler design. This forecloses any DOE-based infringement theory for all five asserted claims.

B. Independent DOE Analysis (Even Absent Estoppel)

Even if prosecution history estoppel did not apply — which, for the reasons stated above, it does — a doctrine of equivalents argument would fail on the merits under the traditional function-way-result test.

Function. Both the continuous taper of the '567 Patent and the segmented taper of Lumos's accused products achieve mode conversion for fiber-to-chip optical coupling. The function is arguably the same or substantially similar: converting the optical mode from a larger fiber mode profile to a smaller waveguide mode profile (and vice versa) while minimizing power loss.

Way. The two approaches achieve mode conversion through substantially different ways. The continuous taper achieves mode conversion through gradual, smooth mode evolution along a continuously varying waveguide geometry. The optical mode is maintained in its fundamental state throughout the transition, with the mode profile evolving continuously as the waveguide width changes. The segmented taper, by contrast, achieves mode conversion through a series of discrete, optimized coupling sections. Each segment has its own fixed geometry, and coupling between segments occurs at discrete boundary interfaces. The mode conversion mechanism relies on constructive interference across the aggregate segmented structure rather than continuous adiabatic evolution. These are fundamentally different physical mechanisms, and the "way" prong of the function-way-result test is not satisfied.

Result. Both approaches achieve low-loss optical coupling between the fiber and the photonic integrated circuit. The result is arguably the same or substantially similar.

Because the "way" prong is substantially different, the function-way-result test fails, and DOE would not apply even in the absence of prosecution history estoppel.

Vitiation. Additionally, applying the doctrine of equivalents to read the "continuous" taper requirement to encompass a "segmented" taper would effectively vitiate the "gradual, continuous taper" limitation as construed by the Court. A claim limitation cannot be effectively read out of the claim through the doctrine of equivalents. See Brilliant Instruments, Inc. v. GuideTech, LLC, 707 F.3d 1342, 1347 (Fed. Cir. 2013) ("If a court determines that a finding of infringement under the doctrine of equivalents would entirely vitiate a particular claim limitation, then the court should rule that there is no infringement under the doctrine of equivalents."). If "continuous" is construed to encompass "segmented," the distinction between continuous and segmented tapers — which was the very basis of the Court's claim construction and the applicants' prosecution disclaimer — would be rendered meaningless. This further defeats any DOE argument.

VI. Impact Assessment: Comparison to Pre-Markman Position

The Court's Markman Order has fundamentally altered the litigation posture with respect to the '567 Patent. A comparison of the pre-Markman and post-Markman positions illustrates the magnitude of this shift.

Pre-Markman Position. Under OWS's proposed functional construction of "adiabatically coupled," the infringement case against the '567 Patent was assessed in the Pre-Suit Memo and subsequent strategy memoranda as moderately strong. The primary risk identified was the claim construction outcome — specifically, the possibility that the Court would adopt a structurally-focused construction that would exclude Lumos's segmented taper design. However, the team assessed this as a moderate risk, not a certainty, given the arguments available in support of a broader functional construction. The '567 Patent was regarded as a cornerstone of OWS's overall case because the integrated waveguide coupler technology addressed by the patent is present in both accused products — the LumosBridge™ 800G and the LumosEdge™ 400ZR+ — making it the patent with the broadest applicability across the accused product line.

Post-Markman Position. The Court's adoption of the structural, Defendant-aligned construction eliminates all viable infringement theories for the '567 Patent. There is no remaining path to establish literal infringement or infringement under the doctrine of equivalents for any of the five asserted claims. The risk identified in the pre-Markman analysis has materialized fully, and the '567 Patent is no longer a viable assertion vehicle in this litigation.

Valuation Impact. Prior to filing suit, OWS engaged Clearpath IP Valuation Services to assess the litigation value of the four asserted patents collectively. Clearpath's pre-litigation valuation attributed significant value to the '567 Patent as part of the $42.8 million collective valuation of the four-patent portfolio. That valuation was based, in part, on the assumption that the Court would adopt OWS's proposed functional construction (or at least a construction that would encompass Lumos's accused products). In light of the Markman Order, the post-Markman litigation value of the '567 Patent in the context of this action is effectively zero. The '567 Patent retains intrinsic value as intellectual property — it remains a valid patent with potential applicability to other products and competitors — but it has no remaining value as an assertion weapon against Lumos's currently accused products.

Compounding Effect. The loss of the '567 Patent infringement theory does not occur in isolation. As analyzed in the separate parallel Post-Markman viability memoranda for the '234 Patent, the '678 Patent, and the '890 Patent (each dated February 21–22, 2023), the remaining three patents-in-suit face similar construction-driven infringement challenges. The cumulative effect of the Markman Order across all four patents significantly weakens OWS's overall litigation position and supports the recommendation for settlement discussions.

VII. Consideration of Alternative Theories

A. Indirect Infringement

We have considered whether OWS's infringement theories for the '567 Patent could be reformulated as claims of indirect infringement — either induced infringement under 35 U.S.C. § 271(b) or contributory infringement under 35 U.S.C. § 271(c). The answer is no.

Both theories of indirect infringement require, as an essential predicate, the existence of an underlying act of direct infringement by some party. See Limelight Networks, Inc. v. Akamai Techs., Inc., 572 U.S. 915, 921 (2014) ("there can be no indirect infringement without direct infringement"). Induced infringement under § 271(b) requires proof that the defendant actively induced a third party to commit an act of direct infringement, with knowledge that the induced acts constitute patent infringement. Contributory infringement under § 271(c) requires proof that the defendant sold a component of a patented invention, knowing it to be especially made or adapted for use in infringing the patent and not a staple article of commerce suitable for substantial noninfringing use.

Because the accused products — the LumosBridge™ 800G and LumosEdge™ 400ZR+ — do not literally or equivalently practice the "adiabatically coupled" limitation as construed by the Court, there is no act of direct infringement by Lumos, its customers, or any other party. The segmented taper coupler used in these products falls outside the scope of the asserted claims. Absent direct infringement, no theory of indirect infringement is viable. Reformulating the claims as indirect infringement theories would not cure the fundamental deficiency in the infringement analysis.

B. Redesign / Future Products

This analysis is limited to the currently accused products — the LumosBridge™ 800G and LumosEdge™ 400ZR+ — as they exist in the marketplace and as they have been characterized in the discovery record. The analysis does not address potential future Lumos products or redesigns.

It is theoretically possible that Lumos could, in the future, adopt a continuous taper coupler design in a new or redesigned product. If such a product were to employ a "gradual, continuous taper designed and configured to prevent excitation of higher-order modes over the entire transition length" — consistent with the Court's construction — infringement of the '567 Patent could become viable with respect to that future product. However, there is no evidence in the current record of any such redesign by Lumos. To the contrary, Lumos's publicly stated technology roadmap and product development plans (as disclosed in its investor presentations and SEC filings) indicate continued reliance on its proprietary segmented taper architecture.

This consideration is not relevant to the current litigation posture and does not alter the recommendation set forth in this memorandum.

VIII. Conclusions and Recommendations

Based on the foregoing analysis, we reach the following conclusions and recommendations:

Conclusion. Infringement of asserted Claims 1, 4, 7, 12, and 15 of U.S. Patent No. 10,234,567 is untenable under the Court's Markman Order construction of "adiabatically coupled." Lumos's segmented (non-continuous) taper coupler architecture, as implemented in both the LumosBridge™ 800G Optical Engine and the LumosEdge™ 400ZR+ Pluggable Transceiver, falls squarely outside the Court's construction, which requires a "gradual, continuous taper designed and configured to prevent excitation of higher-order modes over the entire transition length." The failure to satisfy the "adiabatically coupled" limitation is dispositive as to all five asserted claims, because each dependent claim (Claims 4, 7, 12, and 15) inherits this limitation from independent Claim 1.

The doctrine of equivalents is foreclosed by prosecution history estoppel arising from the narrowing amendment that added the "adiabatically coupled" limitation to Claim 1 during prosecution to overcome the Tanaka prior art (U.S. Patent No. 9,456,012). The applicants' prosecution remarks explicitly characterized the invention as requiring "adiabatic coupling through a gradual, continuous taper transition," creating an unrebutted presumption of estoppel under Festo. Even absent estoppel, the doctrine of equivalents would fail because the "way" in which Lumos's segmented taper achieves mode conversion is substantially different from the "way" in which the continuous taper of the '567 Patent achieves mode conversion, and applying DOE would vitiate the "gradual, continuous taper" limitation.

No alternative theory — including indirect infringement — can salvage the '567 Patent infringement case given the fundamental failure to establish that the accused products practice the dispositive claim limitation.

Recommendation. The '567 Patent should not be pursued further in this litigation. The resources currently allocated to the '567 Patent infringement theory — including expert retention, damages analysis, and discovery — should be redirected or wound down in accordance with the team's revised litigation budget.

This conclusion is consistent with, and reinforces, the overall post-Markman assessment set forth in the separate consolidated Post-Markman Impact Assessment Memorandum dated February 21, 2023, which recommends that OWS pursue settlement discussions with Lumos. The loss of the '567 Patent, combined with the similar construction-driven setbacks on the '234 Patent, '678 Patent, and '890 Patent (as analyzed in parallel memoranda), substantially diminishes OWS's litigation leverage and makes continued prosecution of this action inadvisable absent a favorable settlement opportunity.

This analysis should be read in conjunction with the parallel post-Markman viability memoranda for the '234 Patent, the '678 Patent, and the '890 Patent, as well as the consolidated Post-Markman Impact Assessment Memorandum. We are available to discuss these findings and recommendations at your convenience and recommend scheduling a case assessment call with the client at the earliest opportunity.

Appendix A: Claim Chart — Claim 1 Under Court's Construction

The following claim chart analyzes each limitation of Claim 1 of U.S. Patent No. 10,234,567 against the accused Lumos products (LumosBridge™ 800G Optical Engine and LumosEdge™ 400ZR+ Pluggable Transceiver) under the Court's claim constructions as set forth in the Markman Order dated February 17, 2023. The dispositive non-infringement finding is highlighted in the "adiabatically coupled" row.

Summary: All limitations of Claim 1 are met by the accused products except the dispositive "adiabatically coupled" limitation. Because this limitation is not met, Claim 1 is not infringed.

Appendix B: Summary of Prosecution History — "Adiabatically Coupled" Amendment

The following is a chronological summary of the prosecution history of U.S. Patent No. 10,234,567, focusing on the events leading to the addition of the "adiabatically coupled" limitation to Claim 1 and the prosecution remarks that are the basis for the prosecution history estoppel analysis in Section V.A of this memorandum.

March 14, 2017 — Original Application Filing. The application for what would become the '567 Patent was filed on March 14, 2017, with seventeen original claims. Original Claim 1 recited an integrated optical coupler comprising a tapered silicon waveguide coupled to a silicon nitride waveguide layer on a silicon photonics platform, with mode conversion functionality. Critically, original Claim 1 did not include the term "adiabatically coupled." The coupling between the tapered silicon waveguide and the silicon nitride waveguide layer was described without any limitation as to the nature or mechanism of the coupling.

October 5, 2017 — First Office Action. The USPTO examiner issued a first office action rejecting Claims 1–17 under 35 U.S.C. § 102(a)(1) as anticipated by Tanaka, U.S. Patent No. 9,456,012 (issued September 6, 2016). The examiner found that the Tanaka patent disclosed all elements of the original claims, including a tapered silicon waveguide coupled to a silicon nitride waveguide layer on a silicon photonics platform. The examiner's analysis noted that the Tanaka patent disclosed a waveguide coupler with a tapered transition region, and that the original claims did not recite any structural or functional feature that distinguished the claimed coupler from the Tanaka disclosure.

January 22, 2018 — Applicant's First Response. The applicants filed a response to the first office action, amending several claims and presenting arguments to distinguish the claims from the Tanaka prior art. However, the applicants did not add the "adiabatically coupled" limitation in this first response. Instead, the applicants attempted to distinguish the claims by arguing that other claim elements — including the specific tip width limitation ("less than 200 nanometers") and the mode conversion configuration — were not disclosed or suggested by the Tanaka prior art. The applicants also added narrowing limitations to certain dependent claims regarding the waveguide geometry and material composition.

May 3, 2018 — Second Office Action. The examiner issued a second office action maintaining the § 102(a)(1) rejection based on the Tanaka prior art. The examiner found that the amendments and arguments presented in the applicants' first response were insufficient to overcome the rejection. Specifically, the examiner found that the Tanaka patent disclosed a tapered silicon waveguide with a tip width within the claimed range and that the applicants' arguments regarding other claim elements did not distinguish the invention from the Tanaka disclosure. The examiner noted that the Tanaka patent's coupler, while employing a different coupling mechanism than the applicants' described embodiments, nonetheless met the broad language of the original claims.

August 15, 2018 — Applicant's Second Response. The applicants filed a second response to the office action, this time adding the "adiabatically coupled" limitation to independent Claim 1. The amended Claim 1 now recited "a tapered silicon waveguide having a tip width of less than 200 nanometers adiabatically coupled to a silicon nitride waveguide layer." Conforming amendments were made to the dependent claims.

In the remarks accompanying this amendment, the applicants' prosecution counsel stated:

"The present invention is distinguished from Tanaka by the requirement of adiabatic coupling through a gradual, continuous taper transition, which prevents excitation of higher-order modes — a feature absent from Tanaka's non-adiabatic coupler design. Tanaka's coupler employs a fundamentally different coupling mechanism that does not maintain single-mode propagation through a gradual, continuous transition. The addition of the 'adiabatically coupled' limitation to Claim 1 clearly distinguishes the present invention from the Tanaka reference, and Applicants respectfully request withdrawal of the rejection."

These remarks are significant for two independent reasons. First, they explicitly define "adiabatic coupling" in structural terms — as requiring "a gradual, continuous taper transition" — rather than in purely functional terms. Second, they explicitly disclaim the Tanaka prior art's coupling mechanism as "fundamentally different" from the claimed adiabatic coupling approach.

November 9, 2018 — Notice of Allowance. The examiner issued a Notice of Allowance, finding that the addition of the "adiabatically coupled" limitation distinguished the claimed invention from the Tanaka prior art and rendered the claims allowable. The examiner's reasons for allowance cited the applicants' arguments regarding the distinction between the adiabatic coupling of the claimed invention and the non-adiabatic coupling of the Tanaka reference.

March 19, 2019 — Patent Issuance. U.S. Patent No. 10,234,567 was issued on March 19, 2019, with fifteen claims. Claim 1, as issued, includes the "adiabatically coupled" limitation that was added during prosecution.

Significance for Prosecution History Estoppel. The prosecution history set forth above establishes a textbook case of a narrowing amendment made for reasons of patentability under Festo Corp. v. Fenmore & Co., 535 U.S. 722 (2002). The "adiabatically coupled" limitation was added specifically to overcome the examiner's § 102(a)(1) rejection based on the Tanaka prior art. The applicants' remarks explicitly characterized the limitation as requiring "a gradual, continuous taper transition" — language that disclaims non-continuous coupling approaches. The scope of surrender encompasses non-continuous coupling mechanisms, including segmented taper designs such as those used in Lumos's accused products. The three Festo rebuttal exceptions a

PRIVILEGED AND CONFIDENTIAL — ATTORNEY WORK PRODUCT
This memorandum is protected by the attorney-client privilege and the work product doctrine. It was prepared in anticipation of litigation and reflects the mental impressions, conclusions, opinions, and legal theories of counsel. It is not to be disclosed to any third party without the prior written authorization of Calderwood & Harkness LLP.
CALDERWOOD & HARKNESS LLP One Embarcadero Center, 34th Floor San Francisco, California 94111 Telephone: (415) 555-7200 Facsimile: (415) 555-7201
MEMORANDUM
February 22, 2023
TO: Alan Ngo, Responsible Partner, intellectual Property Group
FROM: Omar Siddiqui, Lead Associate, intellectual Property Group
CC: Christy Bauer, Associate; Marcus Delgado, Billing Partner
RE: Post-Markman Infringement Viability Analysis — U.S. Patent No. 10,234,567 ("the '567 Patent") — Asserted Claims 1, 4, 7, 12, and 15
MATTER: Optiwave Semiconductor Corp. v. Lumos Photonics, Inc., Case No. 5:22-cv-03417-EJD (N.D. Cal.)
C&H MATTER NO.: 1014-00003
CLIENT: Optiwave Semiconductor Corp.
Right-click to update Table of Contents
I. Executive Summary
The purpose of this memorandum is to assess the viability of Optiwave Semiconductor Corp.'s ("OWS" or "Plaintiff") infringement theories for U.S. Patent No. 10,234,567 ("the '567 Patent") following the Markman Order issued on February 17, 2023, by the Honorable Edward J. Torvelis of the United States District Court for the Northern District of California. This analysis addresses all five asserted claims of the '567 Patent — Claims 1, 4, 7, 12, and 15 — and evaluates whether any viable path to establishing infringement remains in light of the Court's claim constructions.
The Court's 34-page Markman Order construed eleven disputed claim terms across the four patents-in-suit in this action. Of those eleven terms, the Court adopted Defendant Lumos Photonics, Inc.'s ("Lumos" or "Defendant") proposed construction — or a construction substantially aligned with Defendant's position — on eight of the eleven terms. The Court adopted Plaintiff OWS's proposed construction on two terms and arrived at a modified construction that departed from both parties' proposals on one term. This outcome represents a significant and, candidly, decisive setback for OWS's infringement case across the patent portfolio.
For the '567 Patent specifically, the critical disputed term was "adiabatically coupled," which appears in independent Claim 1 and is inherited by all four asserted dependent claims (Claims 4, 7, 12, and 15). The parties advanced materially different constructions of this term:
• OWS's proposed construction: "coupled such that optical mode transitions occur without significant power loss."
• Lumos's proposed construction: "coupled through a gradual, continuous taper that prevents excitation of higher-order modes throughout the entire transition region."
The Court adopted a construction closely aligned with Defendant's proposal, construing "adiabatically coupled" to mean: "coupled through a gradual, continuous taper designed and configured to prevent excitation of higher-order modes over the entire transition length." The Court's construction introduced minor modifications from Lumos's proposal — replacing "throughout the entire transition region" with "over the entire transition length" and adding the phrase "designed and configured to" — but substantively embraced Defendant's structural interpretation of the term. The Court's construction imposes structural requirements on the coupling mechanism rather than defining it solely by its functional result, which was the cornerstone of OWS's infringement theory.
Bottom-line conclusion: Infringement theories for all five asserted claims of the '567 Patent (Claims 1, 4, 7, 12, and 15) are untenable under the Court's construction. Lumos's LumosBridge™ 800G Optical Engine and LumosEdge™ 400ZR+ Pluggable Transceiver products employ a segmented (non-continuous) taper coupler architecture with discrete coupling sections. This design falls squarely outside the Court's narrowed construction, which requires a "gradual, continuous taper" over the "entire transition length." The segmented taper design used by Lumos is structurally distinct from the continuous adiabatic taper contemplated by the claims as construed.
Moreover, doctrine-of-equivalents arguments are foreclosed by prosecution history estoppel. The "adiabatically coupled" limitation was added to Claim 1 by amendment during prosecution of the '567 Patent application specifically to overcome a prior art rejection based on the Tanaka patent (U.S. Patent No. 9,456,012). The applicants' prosecution remarks explicitly distinguished the invention by reference to a "gradual, continuous taper transition," thereby disclaiming non-continuous coupling approaches. Under Festo Corp. v. Fenmore & Co., 535 U.S. 722 (2002), this narrowing amendment creates a presumption of estoppel that OWS cannot rebut, as segmented taper couplers were well-known and foreseeable at the time of the amendment.
Recommendation: Infringement of the '567 Patent should not be pursued further in this litigation. This assessment is consistent with, and reinforces, the broader recommendation for settlement discussions set forth in the consolidated Post-Markman Impact Assessment Memorandum dated February 21, 2023.
II. Background
A. The '567 Patent Overview
U.S. Patent No. 10,234,567, entitled "Integrated Waveguide Coupler with Tapered Mode Converter for Silicon Photonics Platforms," was filed on March 14, 2017, and issued on March 19, 2019, after prosecution before the United States Patent and Trademark Office. The patent is assigned to Optiwave Semiconductor Corp., our client.
The named inventors on the '567 Patent are: (1) Dr. Kevin Fong, who serves as Chief Technology Officer and Co-Founder of OWS; (2) Dr. Priya Ramanathan, who serves as Chief Executive Officer and Co-Founder of OWS; and (3) Hideo Tanaka, a former OWS engineer who was employed by the company from 2016 to 2019. As discussed in further detail below, Mr. Tanaka's prior patent — U.S. Patent No. 9,456,012 — played a significant role in the prosecution of the '567 Patent and, ultimately, in the claim construction at issue.
The '567 Patent is directed to an integrated optical coupler designed for use on silicon photonics platforms. More specifically, the invention concerns a tapered silicon waveguide that is adiabatically coupled to a silicon nitride (SiN) waveguide layer, thereby enabling efficient fiber-to-chip and chip-to-fiber optical coupling in high-speed optical transceivers. The technology addresses a well-known challenge in silicon photonics: the large mode-size mismatch between standard single-mode optical fibers and the sub-micron silicon waveguides used in integrated photonic circuits. The '567 Patent's approach uses a tapered mode converter to gradually transform the optical mode from the larger fiber mode to the smaller waveguide mode (and vice versa), minimizing insertion loss and back-reflection.
OWS's commercial silicon photonics chips incorporating the patented technology are fabricated by Pinnacle Semiconductor Foundry Co., Ltd. on its 45nm silicon-on-insulator (SOI) process. These chips are deployed in OWS's own optical transceiver products and are also licensed to certain third-party transceiver manufacturers under OWS's technology licensing program.
B. Asserted Claims
OWS asserts Claims 1, 4, 7, 12, and 15 of the '567 Patent in this action. Claim 1 is the principal independent claim, and Claims 4, 7, 12, and 15 each depend — directly or indirectly — from Claim 1.
Claim 1 recites, in relevant part:
"An integrated optical coupler comprising a tapered silicon waveguide having a tip width of less than 200 nanometers adiabatically coupled to a silicon nitride waveguide layer, said silicon nitride waveguide layer disposed on a silicon photonics platform, wherein the tapered silicon waveguide is configured to convert an optical mode from a first mode profile to a second mode profile for coupling between an optical fiber and a photonic integrated circuit."
(Emphasis added to the disputed term.)
Claim 4 depends directly from Claim 1 and adds a limitation requiring a specific taper length-to-width ratio for the tapered silicon waveguide. Claim 7 depends directly from Claim 1 and adds a limitation regarding the material composition of the silicon nitride waveguide layer, specifying a refractive index range. Claim 12 depends directly from Claim 1 and adds a limitation directed to integration with a photodetector on the same silicon photonics platform. Claim 15 depends from Claim 12 (and thus indirectly from Claim 1) and adds a limitation requiring a specific germanium-on-silicon photodetector integration with the waveguide coupler.
Critically, all five asserted claims inherit the "adiabatically coupled" limitation from Claim 1. Because the Court's construction of this term is outcome-determinative — as explained in detail in Sections III and IV below — the infringement analysis for all five claims rises or falls together.
C. Accused Products
OWS accuses two Lumos Photonics products of infringing the asserted claims of the '567 Patent:
1. LumosBridge™ 800G Optical Engine. The LumosBridge™ 800G is a co-packaged optics (CPO) engine designed for integration with high-bandwidth switch ASICs in hyperscale data center environments. The product was first shipped commercially in the first quarter of 2022. Based on publicly available pricing information and discovery responses, the unit price of the LumosBridge™ 800G ranges from approximately $1,200 to $1,800 per unit, depending on configuration and volume tier.
2. LumosEdge™ 400ZR+ Pluggable Transceiver. The LumosEdge™ 400ZR+ is a QSFP-DD form factor pluggable optical transceiver designed for metro and data center interconnect (DCI) applications. The product has been in production and commercially available since the third quarter of 2021. Unit pricing for the LumosEdge™ 400ZR+ ranges from approximately $600 to $900 per unit.
Based on Lumos's annual report on Form 10-K for fiscal year 2022 and discovery responses produced in this litigation, the combined revenue attributable to the two accused products for calendar year 2022 was approximately $148.6 million. Lumos's total fiscal year 2022 revenue was $312.8 million, meaning the accused products represent approximately 47.5% of Lumos's overall revenue.
Lumos Photonics, Inc. (Nasdaq: LMPH) is a Delaware corporation headquartered in Austin, Texas. The company operates a semiconductor fabrication facility in Hillsboro, Oregon, and maintains a sales and engineering office at 4750 Great America Parkway, Suite 320, Santa Clara, California 95054. At the time OWS filed the complaint in this action (June 3, 2022), Lumos's market capitalization was approximately $1.87 billion.
D. Pre-Markman Infringement Theory
Prior to the Markman hearing, the viability of OWS's infringement case against the '567 Patent was assessed in detail in a pre-suit infringement analysis memorandum prepared by this team, dated April 18, 2022 (the "Pre-Suit Memo"). That analysis identified the "adiabatically coupled" limitation as a potential vulnerability in OWS's infringement theory, but assessed the risk as manageable under the then-prevailing litigation strategy.
OWS's pre-Markman infringement theory relied on the broader proposed construction of "adiabatically coupled" — namely, "coupled such that optical mode transitions occur without significant power loss." Under this functional construction, the term would encompass any coupling mechanism that achieved low-loss mode transfer between the silicon waveguide and the silicon nitride waveguide layer, regardless of the specific structural approach used to achieve that result. Critically, this construction would have encompassed Lumos's segmented taper design, because the segmented taper does, in fact, achieve low-loss coupling — it simply does so through a different structural mechanism than a continuous adiabatic taper.
Under OWS's proposed construction, the focus of the infringement inquiry was on the functional result (low power loss during mode transitions) rather than the structural requirement (a gradual, continuous taper). This approach was grounded in the specification's recurrent emphasis on the purpose of adiabatic coupling — i.e., minimizing power loss — and was supported by the declaration of Dr. Harold Whitfield of Whitfield Technical Consulting Group, Inc. (Palo Alto, California), who opined that persons of ordinary skill in the art of integrated photonics understand "adiabatic coupling" primarily by reference to its functional result.
The Pre-Suit Memo flagged the risk that the Court could adopt a narrower, structurally-focused construction of "adiabatically coupled," which would be more consistent with Lumos's expected defensive position. That risk was assessed as moderate at the time, given the breadth of the specification's disclosure and the general principle that claim terms should be given their plain and ordinary meaning to a person of ordinary skill in the art ("POSITA"). However, the Pre-Suit Memo acknowledged that the prosecution history — specifically the amendment adding the "adiabatically coupled" limitation to overcome the Tanaka prior art — presented a meaningful counterargument that could persuade the Court to adopt a narrower construction.
As discussed in the following section, the risk identified in the Pre-Suit Memo has materialized in the most consequential way possible.
III. The Court's Claim Construction — "Adiabatically Coupled"
A. Summary of the Parties' Positions
The parties presented sharply divergent proposed constructions for the term "adiabatically coupled" as it appears in Claim 1 of the '567 Patent. The dispute centered on whether the term should be construed functionally — defining the coupling by the result it achieves — or structurally — defining the coupling by the physical mechanism required to achieve it.
1. OWS's Proposed Construction
OWS proposed that "adiabatically coupled" be construed to mean: "coupled such that optical mode transitions occur without significant power loss."
OWS's construction was grounded in several arguments. First, OWS contended that the term should be given its plain and ordinary meaning to a person of ordinary skill in the art of integrated photonics. OWS argued that in the field, "adiabatic" is understood as a term describing the efficiency of a mode transition — i.e., a transition that occurs slowly enough relative to the coupling length that the system remains in its fundamental mode throughout the process, avoiding excitation of parasitic higher-order modes and thereby minimizing power loss. Under this interpretation, any coupling mechanism that achieves this functional result — low-loss, single-mode transition — qualifies as "adiabatic," regardless of the specific physical structure used.
Second, OWS argued that the specification of the '567 Patent supports this functional interpretation. The specification repeatedly describes the purpose of adiabatic coupling as minimizing insertion loss and maintaining efficient mode transfer. While the specification discloses specific embodiments using a gradual, continuous taper, OWS contended that the specification should not be read to limit the claims to those specific embodiments.
Third, OWS relied on the declaration and report of Dr. Harold Whitfield of Whitfield Technical Consulting Group, Inc. (Palo Alto, California), an expert in integrated photonics design. Dr. Whitfield opined that persons of ordinary skill in the art understand "adiabatic coupling" by reference to its functional result — efficient, low-loss mode transfer — and would not necessarily limit the term to a single structural implementation such as a continuous taper. Dr. Whitfield cited several textbooks and peer-reviewed publications that describe "adiabatic" coupling in functional terms.
2. Lumos's Proposed Construction
Lumos proposed that "adiabatically coupled" be construed to mean: "coupled through a gradual, continuous taper that prevents excitation of higher-order modes throughout the entire transition region."
Lumos's construction was grounded in both intrinsic and extrinsic evidence. Lumos argued, first, that the prosecution history compelled a structural interpretation. The "adiabatically coupled" limitation was not present in the original application as filed on March 14, 2017. It was added by amendment during prosecution in response to the second office action, which rejected the claims under 35 U.S.C. § 102(a)(1) as anticipated by the Tanaka prior art (U.S. Patent No. 9,456,012, issued September 6, 2016). In the applicants' remarks accompanying the second amendment, OWS's prosecution counsel explicitly distinguished the invention from the Tanaka prior art by stating that the present invention requires "adiabatic coupling through a gradual, continuous taper transition" — language that Lumos argued constituted a clear disclaimer of non-continuous coupling approaches.
Second, Lumos pointed to the specification's consistent description of the taper coupler in exclusively gradual and continuous terms. Every embodiment disclosed in the specification employs a continuous, smoothly-varying taper profile. No embodiment discloses or suggests a segmented, stepped, or discontinuous taper. Lumos argued that the specification thus confirms the structural meaning of "adiabatically coupled" as requiring a gradual, continuous taper.
Third, Lumos presented extrinsic evidence from the technical literature defining "adiabatic" in the specific context of waveguide coupling. The term "adiabatic" in waveguide optics, Lumos argued, has a well-established meaning that encompasses not just the functional result but also the structural requirement of gradual, continuous variation to maintain single-mode propagation throughout the transition.
Fourth, Lumos relied on the testimony and report of Dr. Raymond Ng, who serves as Of Counsel and Technical Consultant to Lumos's counsel, Brannick Howell & Stein LLP. Dr. Ng is a former professor of photonics at the University of Colorado Boulder, with over 25 years of experience in integrated photonics research and waveguide design. Dr. Ng testified at the Markman hearing held on January 30, 2023, and opined that a person of ordinary skill in the art would understand "adiabatically coupled" in the context of the '567 Patent to require a gradual, continuous taper that prevents excitation of higher-order modes. Dr. Ng distinguished between the thermodynamic usage of "adiabatic" (referring generally to a process without energy exchange) and the specific usage in waveguide optics (referring to a gradual, continuous transition that maintains single-mode propagation).
B. The Court's Construction
In its Markman Order dated February 17, 2023, the Court construed "adiabatically coupled" as follows:
"coupled through a gradual, continuous taper designed and configured to prevent excitation of higher-order modes over the entire transition length"
See Markman Order at 12.
The Court's construction closely tracks Lumos's proposed construction, with two minor modifications. First, the Court replaced "throughout the entire transition region" with "over the entire transition length" — a change that is substantively immaterial but reflects the Court's preference for the precision of "transition length" over the slightly broader "transition region." Second, the Court added the phrase "designed and configured to" before "prevent excitation of higher-order modes," which introduces a design-intent qualification but does not alter the fundamental structural requirement of a gradual, continuous taper. Substantively, the Court adopted Defendant's structural interpretation of the term.
The Court's reasoning, as set forth on pages 8 through 15 of the Markman Order, proceeded through the standard claim construction framework established in Phillips v. AWH Corp., 415 F.3d 1303 (Fed. Cir. 2005) (en banc), giving primacy to intrinsic evidence.
Intrinsic Evidence — Specification. The Court found that the specification of the '567 Patent describes the taper coupler in exclusively gradual, continuous terms. The Court noted that every disclosed embodiment employs a continuous taper with a smoothly varying width profile, and that the specification contains no disclosure of a segmented, stepped, or otherwise discontinuous taper. The Court observed that while claims are not ordinarily limited to the disclosed embodiments, the specification's uniform treatment of the taper as continuous — combined with the prosecution history discussed below — supported a structural interpretation of "adiabatically coupled."
Intrinsic Evidence — Prosecution History. The Court found the prosecution history to be particularly instructive, and indeed dispositive, on the meaning of "adiabatically coupled." The Court recounted the prosecution timeline in detail: the original application did not include the "adiabatically coupled" limitation; the limitation was added by amendment in response to the second office action, which maintained a § 102(a)(1) rejection based on the Tanaka prior art. The Court quoted the applicants' remarks accompanying the amendment, in which prosecution counsel stated that "the present invention is distinguished from Tanaka by the requirement of adiabatic coupling through a gradual, continuous taper transition, which prevents excitation of higher-order modes — a feature absent from Tanaka's non-adiabatic coupler design." Markman Order at 13. The Court found that these remarks created a clear disclaimer of non-continuous coupling approaches, and that the prosecution history narrowed the scope of "adiabatically coupled" to require the gradual, continuous taper structure described in both the specification and the applicants' own characterization of the term during prosecution.
Extrinsic Evidence. The Court credited Dr. Ng's testimony regarding the structural meaning of "adiabatic" in the specific context of waveguide coupling, finding it consistent with and confirmatory of the intrinsic record. The Court found Dr. Whitfield's purely functional interpretation to be inconsistent with the intrinsic record — particularly the prosecution history — and therefore declined to adopt it. The Court emphasized that while Dr. Whitfield's testimony may reflect a usage of "adiabatic" in some broader contexts, the intrinsic evidence of this particular patent establishes a narrower, structural meaning.
Rejection of OWS's Plain and Ordinary Meaning Argument. The Court explicitly rejected OWS's argument that "adiabatically coupled" should be given its broadest plain and ordinary meaning to a POSITA. The Court acknowledged that the term may have a broader functional meaning in some contexts, but found that the prosecution history narrowed the claim scope and that the plain and ordinary meaning argument could not override the clear disclaimer established during prosecution. See Markman Order at 14 (citing Omega Eng'g, Inc. v. Raytek Corp., 334 F.3d 1314, 1323 (Fed. Cir. 2003) (noting that "the prosecution history limits the interpretation of claim terms so as to exclude any interpretation that was disclaimed during prosecution")).
C. Impact on Scope of "Adiabatically Coupled"
The Court's construction of "adiabatically coupled" imposes four distinct requirements that any accused coupler design must satisfy:
1. A gradual taper — the taper must vary smoothly and gradually in width, not abruptly or in discrete steps.
2. A continuous taper — the taper must be continuous, without segmentation, discontinuities, or discrete coupling sections. A taper composed of discrete segments separated by transition discontinuities does not satisfy this requirement.
3. Designed and configured to prevent higher-order mode excitation — the taper must be designed and configured such that higher-order optical modes are not excited during the mode transition. This is a design-intent requirement that reinforces the structural character of the limitation.
4. Over the entire transition length — the above properties must be maintained over the entire length of the taper transition, not merely at discrete points or in isolated sections.
The critical consequence of this construction is that it imposes a structural limitation on the coupling mechanism, not merely a functional one. A coupler that achieves low-loss mode transfer through a mechanism other than a gradual, continuous taper — even if it achieves the same functional result (minimal power loss) — does not satisfy the "adiabatically coupled" limitation as construed. Specifically, any coupler design using discrete segments, stepped transitions, or non-continuous taper profiles falls outside the scope of the Court's construction.
This is precisely the situation with Lumos's accused products, as analyzed in the following section.
IV. Claim-by-Claim Infringement Analysis Under the Court's Construction
A. Claim 1
Claim 1 is the independent claim from which all other asserted claims depend. The dispositive limitation for purposes of the infringement analysis is "adiabatically coupled," as construed by the Court.
1. The Dispositive Limitation: "Adiabatically Coupled"
Lumos's Accused Design. Based on the discovery produced to date — including Lumos's technical design documents (LMPH-000147523 through LMPH-000147891), process design kit specifications, and internal engineering presentations — as well as publicly available technical documentation (including Lumos's published papers, conference presentations, and product datasheets for the LumosBridge™ 800G and LumosEdge™ 400ZR+ products), the accused products employ a segmented taper coupler architecture that is structurally distinct from the continuous adiabatic taper required by the Court's construction.
Specifically, Lumos's segmented taper coupler design employs a series of discrete coupling sections (segments) rather than a single gradual, continuous taper. Each segment is a distinct waveguide section with its own width and length parameters. The segments are separated by transition discontinuities — abrupt changes in waveguide geometry between adjacent segments. The design achieves low-loss coupling through a fundamentally different mechanism than a continuous adiabatic taper: rather than maintaining continuous mode evolution along a smooth taper profile, Lumos's approach relies on optimized segment geometry — carefully selecting the width, length, and spacing of each discrete segment to achieve constructive interference and efficient power transfer across the aggregate structure.
This segmented taper approach was described by Dr. Nikolai Petrov, Lumos's Vice President of Engineering, Photonics Division, in Lumos's technical publications, including a 2021 conference paper presented at the Optical Fiber Communication Conference (OFC). Dr. Petrov's description of the LumosBridge™ coupler design is consistent with the segmented taper coupler methodology described in the Chen & Liu reference — "Segmented Taper Coupler Design for Low-Loss Fiber-to-Chip Coupling," Journal of Photonic Integration, Vol. 14, No. 3, pp. 217–229 (2016) — which was produced by Lumos in discovery and is referenced in Lumos's prior art contentions. The Chen & Liu paper demonstrates that segmented taper designs were well-established in the art prior to the filing of the '567 Patent, and that they represent a recognized alternative to continuous adiabatic tapers.
Internal Lumos engineering documents produced in discovery (LMPH-000148201 through LMPH-000148367) further confirm the segmented nature of the coupler design. These documents include design specifications showing the individual segment parameters (widths, lengths, and gap dimensions) for the waveguide coupler used in both accused products. The documents refer to the coupler as a "segmented mode converter" and do not use the term "adiabatic" to describe the coupling mechanism.
Analysis. Lumos's segmented taper coupler does not meet the "gradual, continuous taper" requirement of the Court's construction. The discontinuities between segments mean that the taper is neither gradual nor continuous "over the entire transition length." Instead, the taper consists of a plurality of discrete sections with abrupt width changes between them. Even though each individual segment may itself have a relatively smooth profile, the aggregate structure is not a single continuous taper but rather a series of discrete elements. The Court's construction requires continuity "over the entire transition length," and the presence of segmentation discontinuities at multiple points along the transition defeats this requirement.
Furthermore, the mode conversion mechanism in Lumos's design is fundamentally different from the adiabatic mode evolution contemplated by the claims. Rather than maintaining the optical signal in its fundamental mode throughout a gradual, continuous transition, Lumos's segmented design relies on discrete coupling events at each segment boundary. This is a structurally distinct approach that falls outside the scope of the "adiabatically coupled" limitation as construed.
Literal Infringement: Not met. The segmented taper design used in the LumosBridge™ 800G and LumosEdge™ 400ZR+ products is excluded by the Court's construction of "adiabatically coupled." Lumos's coupler is not "coupled through a gradual, continuous taper designed and configured to prevent excitation of higher-order modes over the entire transition length."
Remaining Claim Elements. Even assuming arguendo that all other elements of Claim 1 are met — including the "tapered silicon waveguide having a tip width of less than 200 nanometers," the "silicon nitride waveguide layer," the "silicon photonics platform," and the mode conversion configuration — the failure to meet the "adiabatically coupled" limitation is independently sufficient to defeat literal infringement of Claim 1. Under well-established law, every limitation of a claim must be met, either literally or by a substantial equivalent, for infringement to exist. See Laitram Corp. v. Rexnord, Inc., 939 F.2d 1533, 1535 (Fed. Cir. 1991). The failure to satisfy even a single limitation — here, "adiabatically coupled" — is fatal.
B. Claim 4 (Dependent on Claim 1)
Claim 4 depends directly from Claim 1 and adds a limitation requiring a specific taper length-to-width ratio for the tapered silicon waveguide. Because Claim 4 inherits all limitations of Claim 1, including the "adiabatically coupled" limitation, it necessarily requires that the coupler satisfy the Court's construction of that term.
Because Claim 1 is not infringed under the Court's construction for the reasons set forth above, Claim 4 is likewise not infringed. A dependent claim cannot be infringed unless the independent claim from which it depends is also infringed. See Wahpeton Canvas Co. v. Frontier, Inc., 870 F.2d 1546, 1552 (Fed. Cir. 1989).
Moreover, even if the Claim 1 analysis were not dispositive, the additional limitation of Claim 4 presents independent difficulties. The limitation requires a specific taper length-to-width ratio, which presupposes a single, continuous taper with a definable length and a definable width variation. Lumos's segmented taper geometry does not present a single continuous taper with a single definable length-to-width ratio in the manner contemplated by the claim. Instead, each discrete segment has its own length and width parameters, and the overall coupler structure does not have a single "taper length" or "taper width" that can be meaningfully compared to the claim limitation. This further supports the conclusion that Claim 4 is not infringed.
C. Claim 7 (Dependent on Claim 1)
Claim 7 depends directly from Claim 1 and adds a limitation regarding the material composition of the silicon nitride waveguide layer, specifying that the silicon nitride layer must have a refractive index within a defined range. As with Claim 4, Claim 7 inherits all limitations of Claim 1, including the "adiabatically coupled" limitation.
Because Claim 1 is not infringed under the Court's construction, Claim 7 is not infringed.
As an independent matter, the evidentiary record regarding Lumos's specific silicon nitride refractive index remains incomplete. While certain publicly available Lumos publications reference a SiN layer with optical properties that may or may not fall within the range specified in Claim 7, the detailed process specifications for Lumos's SiN deposition have not been fully produced or analyzed. However, this limitation is moot given the dispositive failure on the Claim 1 "adiabatically coupled" limitation, and further investigation of this issue is not warranted.
D. Claim 12 (Dependent on Claim 1)
Claim 12 depends directly from Claim 1 and adds a limitation directed to integration of the waveguide coupler with a photodetector on the same silicon photonics platform. Claim 12 inherits all limitations of Claim 1, including the "adiabatically coupled" limitation.
Because Claim 1 is not infringed under the Court's construction, Claim 12 is not infringed.
It bears noting that both accused Lumos products — the LumosBridge™ 800G and the LumosEdge™ 400ZR+ — do, in fact, integrate photodetectors on the same silicon photonics platform as the waveguide coupler. Under OWS's original infringement theory, the Claim 12 photodetector limitation was not expected to be a point of contention. However, the fact that Lumos's products satisfy this additional limitation is wholly immaterial given the dispositive failure on the "adiabatically coupled" limitation. Satisfaction of one claim element does not cure the failure to satisfy another.
E. Claim 15 (Dependent on Claim 1 via Claim 12)
Claim 15 depends from Claim 12, which in turn depends from Claim 1. Claim 15 adds a further limitation requiring a specific germanium-on-silicon (Ge-on-Si) photodetector integration with the waveguide coupler. As a claim that depends indirectly from Claim 1 through Claim 12, Claim 15 inherits all limitations of both Claim 1 and Claim 12, including the "adiabatically coupled" limitation.
Because Claim 1 is not infringed under the Court's construction, Claim 15 is not infringed. The analysis of the additional Ge-on-Si photodetector limitation of Claim 15 is unnecessary and has not been undertaken for purposes of this memorandum.
Summary Table: Claim-by-Claim Assessment
V. Doctrine of Equivalents Analysis
Even where literal infringement cannot be established, a patentee may sometimes rely on the doctrine of equivalents ("DOE") to capture accused products that perform substantially the same function, in substantially the same way, to achieve substantially the same result as the claimed invention. See Warner-Jenkinson Co. v. Harborline Enterprises Davis Chem. Co., 520 U.S. 17 (1997). However, as discussed below, the doctrine of equivalents is foreclosed in this case by prosecution history estoppel, and would in any event fail on the merits.
A. Prosecution History Estoppel
1. The Narrowing Amendment
The "adiabatically coupled" limitation was not present in the original claims of the '567 Patent application as filed on March 14, 2017. It was added by amendment during prosecution in response to the United States Patent and Trademark Office's second office action, which maintained a rejection of the claims under 35 U.S.C. § 102(a)(1) as anticipated by the Tanaka prior art — U.S. Patent No. 9,456,012, issued September 6, 2016.
It is noteworthy — and presents an awkward dynamic for OWS — that Hideo Tanaka, a named inventor on the '567 Patent, is also a co-inventor of the Tanaka prior art patent (U.S. 9,456,012). Mr. Tanaka obtained the '012 patent before joining OWS in 2016, and the '012 patent discloses a non-adiabatic coupler design that was cited by the examiner as anticipating the original claims of the '567 Patent application. The '567 Patent applicants were thus compelled to distinguish their invention from Mr. Tanaka's own prior work — a distinction that was accomplished by adding the "adiabatically coupled" limitation to narrow the claims beyond the scope of the Tanaka prior art.
The applicants' remarks accompanying the second amendment are directly relevant to the prosecution history estoppel analysis. In those remarks, OWS's prosecution counsel stated:
"The present invention is distinguished from Tanaka by the requirement of adiabatic coupling through a gradual, continuous taper transition, which prevents excitation of higher-order modes — a feature absent from Tanaka's non-adiabatic coupler design."
This statement constitutes a clear and unambiguous characterization of the scope of the "adiabatically coupled" limitation, explicitly defining it in terms of a "gradual, continuous taper transition." By these remarks, the applicants disclaimed coupling approaches that do not employ a gradual, continuous taper — including, critically, the segmented taper approach used by Lumos.
2. Application of *Festo*
Under the Supreme Court's decision in Festo Corp. v. Fenmore & Co., 535 U.S. 722 (2002), when a patent holder narrows a claim by amendment during prosecution for a reason related to patentability, there is a presumption that the amendment surrendered the territory between the original claim and the amended claim. This presumption of prosecution history estoppel prevents the patent holder from using the doctrine of equivalents to recapture the surrendered subject matter.
Here, the addition of the "adiabatically coupled" limitation was unquestionably a narrowing amendment made for reasons of patentability — specifically, to overcome the Tanaka prior art rejection. The original claims encompassed any coupler design meeting the other claim elements; the amended claims narrowed the scope to require adiabatic coupling. The territory surrendered by this amendment encompasses non-adiabatic coupling approaches, including segmented taper designs that do not employ a gradual, continuous taper.
3. Analysis of the *Festo* Rebuttal Exceptions
Under Festo, the presumption of prosecution history estoppel can be rebutted in three narrow circumstances. We analyze each in turn.
Exception 1: Unforeseeability. The patentee may rebut the presumption by showing that the alleged equivalent was unforeseeable at the time of the amendment. Here, segmented taper couplers were well-known and well-documented in the art at the time of the amendment (2017–2018). The Chen & Liu publication — "Segmented Taper Coupler Design for Low-Loss Fiber-to-Chip Coupling," Journal of Photonic Integration, Vol. 14, No. 3, pp. 217–229 (2016) — was published a full year before the original filing date of the '567 Patent and demonstrates that segmented taper designs were an established alternative to continuous adiabatic tapers in the field. Additional references in the prior art, including conference papers from the 2015 and 2016 OFC conferences, further confirm the foreseeability of segmented taper approaches. The alleged equivalent — a segmented taper coupler — was therefore entirely foreseeable at the time of the narrowing amendment. This rebuttal exception fails.
Exception 2: Tangential Relation. The patentee may rebut the presumption by showing that the rationale underlying the amendment bears only a tangential relation to the equivalent in question. Here, the amendment was made precisely to distinguish the prior art's coupling mechanism — i.e., to require adiabatic coupling through a gradual, continuous taper, as opposed to the non-adiabatic coupling disclosed in the Tanaka prior art. The equivalent at issue — Lumos's segmented taper coupler — is a coupler design that employs a non-continuous coupling mechanism. The rationale for the amendment (distinguishing non-adiabatic, non-continuous coupling) is directly and centrally related to the equivalent at issue (a segmented, non-continuous coupler). This rebuttal exception fails.
Exception 3: Other Reason. The patentee may rebut the presumption by showing some other reason suggesting that the patentee could not reasonably be expected to have described the insubstantial substitute in question. Here, no such reason exists. The patentee — OWS, through its inventors Dr. Fong, Dr. Ramanathan, and Mr. Tanaka — had extensive knowledge of the integrated photonics field, including familiarity with segmented taper designs. The applicants had ample opportunity to draft claims that encompassed segmented tapers, or to add claims directed to alternative coupler architectures, but chose not to do so. Indeed, the prosecution counsel's remarks specifically characterized the invention as requiring a "gradual, continuous taper transition," affirmatively excluding segmented approaches. This rebuttal exception fails.
4. Conclusion on Estoppel
Because all three Festo rebuttal exceptions fail, the presumption of prosecution history estoppel is unrebutted. OWS is estopped from asserting the doctrine of equivalents with respect to the "adiabatically coupled" limitation against Lumos's segmented taper coupler design. This forecloses any DOE-based infringement theory for all five asserted claims.
B. Independent DOE Analysis (Even Absent Estoppel)
Even if prosecution history estoppel did not apply — which, for the reasons stated above, it does — a doctrine of equivalents argument would fail on the merits under the traditional function-way-result test.
Function. Both the continuous taper of the '567 Patent and the segmented taper of Lumos's accused products achieve mode conversion for fiber-to-chip optical coupling. The function is arguably the same or substantially similar: converting the optical mode from a larger fiber mode profile to a smaller waveguide mode profile (and vice versa) while minimizing power loss.
Way. The two approaches achieve mode conversion through substantially different ways. The continuous taper achieves mode conversion through gradual, smooth mode evolution along a continuously varying waveguide geometry. The optical mode is maintained in its fundamental state throughout the transition, with the mode profile evolving continuously as the waveguide width changes. The segmented taper, by contrast, achieves mode conversion through a series of discrete, optimized coupling sections. Each segment has its own fixed geometry, and coupling between segments occurs at discrete boundary interfaces. The mode conversion mechanism relies on constructive interference across the aggregate segmented structure rather than continuous adiabatic evolution. These are fundamentally different physical mechanisms, and the "way" prong of the function-way-result test is not satisfied.
Result. Both approaches achieve low-loss optical coupling between the fiber and the photonic integrated circuit. The result is arguably the same or substantially similar.
Because the "way" prong is substantially different, the function-way-result test fails, and DOE would not apply even in the absence of prosecution history estoppel.
Vitiation. Additionally, applying the doctrine of equivalents to read the "continuous" taper requirement to encompass a "segmented" taper would effectively vitiate the "gradual, continuous taper" limitation as construed by the Court. A claim limitation cannot be effectively read out of the claim through the doctrine of equivalents. See Brilliant Instruments, Inc. v. GuideTech, LLC, 707 F.3d 1342, 1347 (Fed. Cir. 2013) ("If a court determines that a finding of infringement under the doctrine of equivalents would entirely vitiate a particular claim limitation, then the court should rule that there is no infringement under the doctrine of equivalents."). If "continuous" is construed to encompass "segmented," the distinction between continuous and segmented tapers — which was the very basis of the Court's claim construction and the applicants' prosecution disclaimer — would be rendered meaningless. This further defeats any DOE argument.
VI. Impact Assessment: Comparison to Pre-Markman Position
The Court's Markman Order has fundamentally altered the litigation posture with respect to the '567 Patent. A comparison of the pre-Markman and post-Markman positions illustrates the magnitude of this shift.
Pre-Markman Position. Under OWS's proposed functional construction of "adiabatically coupled," the infringement case against the '567 Patent was assessed in the Pre-Suit Memo and subsequent strategy memoranda as moderately strong. The primary risk identified was the claim construction outcome — specifically, the possibility that the Court would adopt a structurally-focused construction that would exclude Lumos's segmented taper design. However, the team assessed this as a moderate risk, not a certainty, given the arguments available in support of a broader functional construction. The '567 Patent was regarded as a cornerstone of OWS's overall case because the integrated waveguide coupler technology addressed by the patent is present in both accused products — the LumosBridge™ 800G and the LumosEdge™ 400ZR+ — making it the patent with the broadest applicability across the accused product line.
Post-Markman Position. The Court's adoption of the structural, Defendant-aligned construction eliminates all viable infringement theories for the '567 Patent. There is no remaining path to establish literal infringement or infringement under the doctrine of equivalents for any of the five asserted claims. The risk identified in the pre-Markman analysis has materialized fully, and the '567 Patent is no longer a viable assertion vehicle in this litigation.
Valuation Impact. Prior to filing suit, OWS engaged Clearpath IP Valuation Services to assess the litigation value of the four asserted patents collectively. Clearpath's pre-litigation valuation attributed significant value to the '567 Patent as part of the $42.8 million collective valuation of the four-patent portfolio. That valuation was based, in part, on the assumption that the Court would adopt OWS's proposed functional construction (or at least a construction that would encompass Lumos's accused products). In light of the Markman Order, the post-Markman litigation value of the '567 Patent in the context of this action is effectively zero. The '567 Patent retains intrinsic value as intellectual property — it remains a valid patent with potential applicability to other products and competitors — but it has no remaining value as an assertion weapon against Lumos's currently accused products.
Compounding Effect. The loss of the '567 Patent infringement theory does not occur in isolation. As analyzed in the separate parallel Post-Markman viability memoranda for the '234 Patent, the '678 Patent, and the '890 Patent (each dated February 21–22, 2023), the remaining three patents-in-suit face similar construction-driven infringement challenges. The cumulative effect of the Markman Order across all four patents significantly weakens OWS's overall litigation position and supports the recommendation for settlement discussions.
VII. Consideration of Alternative Theories
A. Indirect Infringement
We have considered whether OWS's infringement theories for the '567 Patent could be reformulated as claims of indirect infringement — either induced infringement under 35 U.S.C. § 271(b) or contributory infringement under 35 U.S.C. § 271(c). The answer is no.
Both theories of indirect infringement require, as an essential predicate, the existence of an underlying act of direct infringement by some party. See Limelight Networks, Inc. v. Akamai Techs., Inc., 572 U.S. 915, 921 (2014) ("there can be no indirect infringement without direct infringement"). Induced infringement under § 271(b) requires proof that the defendant actively induced a third party to commit an act of direct infringement, with knowledge that the induced acts constitute patent infringement. Contributory infringement under § 271(c) requires proof that the defendant sold a component of a patented invention, knowing it to be especially made or adapted for use in infringing the patent and not a staple article of commerce suitable for substantial noninfringing use.
Because the accused products — the LumosBridge™ 800G and LumosEdge™ 400ZR+ — do not literally or equivalently practice the "adiabatically coupled" limitation as construed by the Court, there is no act of direct infringement by Lumos, its customers, or any other party. The segmented taper coupler used in these products falls outside the scope of the asserted claims. Absent direct infringement, no theory of indirect infringement is viable. Reformulating the claims as indirect infringement theories would not cure the fundamental deficiency in the infringement analysis.
B. Redesign / Future Products
This analysis is limited to the currently accused products — the LumosBridge™ 800G and LumosEdge™ 400ZR+ — as they exist in the marketplace and as they have been characterized in the discovery record. The analysis does not address potential future Lumos products or redesigns.
It is theoretically possible that Lumos could, in the future, adopt a continuous taper coupler design in a new or redesigned product. If such a product were to employ a "gradual, continuous taper designed and configured to prevent excitation of higher-order modes over the entire transition length" — consistent with the Court's construction — infringement of the '567 Patent could become viable with respect to that future product. However, there is no evidence in the current record of any such redesign by Lumos. To the contrary, Lumos's publicly stated technology roadmap and product development plans (as disclosed in its investor presentations and SEC filings) indicate continued reliance on its proprietary segmented taper architecture.
This consideration is not relevant to the current litigation posture and does not alter the recommendation set forth in this memorandum.
VIII. Conclusions and Recommendations
Based on the foregoing analysis, we reach the following conclusions and recommendations:
Conclusion. Infringement of asserted Claims 1, 4, 7, 12, and 15 of U.S. Patent No. 10,234,567 is untenable under the Court's Markman Order construction of "adiabatically coupled." Lumos's segmented (non-continuous) taper coupler architecture, as implemented in both the LumosBridge™ 800G Optical Engine and the LumosEdge™ 400ZR+ Pluggable Transceiver, falls squarely outside the Court's construction, which requires a "gradual, continuous taper designed and configured to prevent excitation of higher-order modes over the entire transition length." The failure to satisfy the "adiabatically coupled" limitation is dispositive as to all five asserted claims, because each dependent claim (Claims 4, 7, 12, and 15) inherits this limitation from independent Claim 1.
The doctrine of equivalents is foreclosed by prosecution history estoppel arising from the narrowing amendment that added the "adiabatically coupled" limitation to Claim 1 during prosecution to overcome the Tanaka prior art (U.S. Patent No. 9,456,012). The applicants' prosecution remarks explicitly characterized the invention as requiring "adiabatic coupling through a gradual, continuous taper transition," creating an unrebutted presumption of estoppel under Festo. Even absent estoppel, the doctrine of equivalents would fail because the "way" in which Lumos's segmented taper achieves mode conversion is substantially different from the "way" in which the continuous taper of the '567 Patent achieves mode conversion, and applying DOE would vitiate the "gradual, continuous taper" limitation.
No alternative theory — including indirect infringement — can salvage the '567 Patent infringement case given the fundamental failure to establish that the accused products practice the dispositive claim limitation.
Recommendation. The '567 Patent should not be pursued further in this litigation. The resources currently allocated to the '567 Patent infringement theory — including expert retention, damages analysis, and discovery — should be redirected or wound down in accordance with the team's revised litigation budget.
This conclusion is consistent with, and reinforces, the overall post-Markman assessment set forth in the separate consolidated Post-Markman Impact Assessment Memorandum dated February 21, 2023, which recommends that OWS pursue settlement discussions with Lumos. The loss of the '567 Patent, combined with the similar construction-driven setbacks on the '234 Patent, '678 Patent, and '890 Patent (as analyzed in parallel memoranda), substantially diminishes OWS's litigation leverage and makes continued prosecution of this action inadvisable absent a favorable settlement opportunity.
This analysis should be read in conjunction with the parallel post-Markman viability memoranda for the '234 Patent, the '678 Patent, and the '890 Patent, as well as the consolidated Post-Markman Impact Assessment Memorandum. We are available to discuss these findings and recommendations at your convenience and recommend scheduling a case assessment call with the client at the earliest opportunity.
Appendix A: Claim Chart — Claim 1 Under Court's Construction
The following claim chart analyzes each limitation of Claim 1 of U.S. Patent No. 10,234,567 against the accused Lumos products (LumosBridge™ 800G Optical Engine and LumosEdge™ 400ZR+ Pluggable Transceiver) under the Court's claim constructions as set forth in the Markman Order dated February 17, 2023. The dispositive non-infringement finding is highlighted in the "adiabatically coupled" row.
Summary: All limitations of Claim 1 are met by the accused products except the dispositive "adiabatically coupled" limitation. Because this limitation is not met, Claim 1 is not infringed.
Appendix B: Summary of Prosecution History — "Adiabatically Coupled" Amendment
The following is a chronological summary of the prosecution history of U.S. Patent No. 10,234,567, focusing on the events leading to the addition of the "adiabatically coupled" limitation to Claim 1 and the prosecution remarks that are the basis for the prosecution history estoppel analysis in Section V.A of this memorandum.
March 14, 2017 — Original Application Filing. The application for what would become the '567 Patent was filed on March 14, 2017, with seventeen original claims. Original Claim 1 recited an integrated optical coupler comprising a tapered silicon waveguide coupled to a silicon nitride waveguide layer on a silicon photonics platform, with mode conversion functionality. Critically, original Claim 1 did not include the term "adiabatically coupled." The coupling between the tapered silicon waveguide and the silicon nitride waveguide layer was described without any limitation as to the nature or mechanism of the coupling.
October 5, 2017 — First Office Action. The USPTO examiner issued a first office action rejecting Claims 1–17 under 35 U.S.C. § 102(a)(1) as anticipated by Tanaka, U.S. Patent No. 9,456,012 (issued September 6, 2016). The examiner found that the Tanaka patent disclosed all elements of the original claims, including a tapered silicon waveguide coupled to a silicon nitride waveguide layer on a silicon photonics platform. The examiner's analysis noted that the Tanaka patent disclosed a waveguide coupler with a tapered transition region, and that the original claims did not recite any structural or functional feature that distinguished the claimed coupler from the Tanaka disclosure.
January 22, 2018 — Applicant's First Response. The applicants filed a response to the first office action, amending several claims and presenting arguments to distinguish the claims from the Tanaka prior art. However, the applicants did not add the "adiabatically coupled" limitation in this first response. Instead, the applicants attempted to distinguish the claims by arguing that other claim elements — including the specific tip width limitation ("less than 200 nanometers") and the mode conversion configuration — were not disclosed or suggested by the Tanaka prior art. The applicants also added narrowing limitations to certain dependent claims regarding the waveguide geometry and material composition.
May 3, 2018 — Second Office Action. The examiner issued a second office action maintaining the § 102(a)(1) rejection based on the Tanaka prior art. The examiner found that the amendments and arguments presented in the applicants' first response were insufficient to overcome the rejection. Specifically, the examiner found that the Tanaka patent disclosed a tapered silicon waveguide with a tip width within the claimed range and that the applicants' arguments regarding other claim elements did not distinguish the invention from the Tanaka disclosure. The examiner noted that the Tanaka patent's coupler, while employing a different coupling mechanism than the applicants' described embodiments, nonetheless met the broad language of the original claims.
August 15, 2018 — Applicant's Second Response. The applicants filed a second response to the office action, this time adding the "adiabatically coupled" limitation to independent Claim 1. The amended Claim 1 now recited "a tapered silicon waveguide having a tip width of less than 200 nanometers adiabatically coupled to a silicon nitride waveguide layer." Conforming amendments were made to the dependent claims.
In the remarks accompanying this amendment, the applicants' prosecution counsel stated:
"The present invention is distinguished from Tanaka by the requirement of adiabatic coupling through a gradual, continuous taper transition, which prevents excitation of higher-order modes — a feature absent from Tanaka's non-adiabatic coupler design. Tanaka's coupler employs a fundamentally different coupling mechanism that does not maintain single-mode propagation through a gradual, continuous transition. The addition of the 'adiabatically coupled' limitation to Claim 1 clearly distinguishes the present invention from the Tanaka reference, and Applicants respectfully request withdrawal of the rejection."
These remarks are significant for two independent reasons. First, they explicitly define "adiabatic coupling" in structural terms — as requiring "a gradual, continuous taper transition" — rather than in purely functional terms. Second, they explicitly disclaim the Tanaka prior art's coupling mechanism as "fundamentally different" from the claimed adiabatic coupling approach.
November 9, 2018 — Notice of Allowance. The examiner issued a Notice of Allowance, finding that the addition of the "adiabatically coupled" limitation distinguished the claimed invention from the Tanaka prior art and rendered the claims allowable. The examiner's reasons for allowance cited the applicants' arguments regarding the distinction between the adiabatic coupling of the claimed invention and the non-adiabatic coupling of the Tanaka reference.
March 19, 2019 — Patent Issuance. U.S. Patent No. 10,234,567 was issued on March 19, 2019, with fifteen claims. Claim 1, as issued, includes the "adiabatically coupled" limitation that was added during prosecution.
Significance for Prosecution History Estoppel. The prosecution history set forth above establishes a textbook case of a narrowing amendment made for reasons of patentability under Festo Corp. v. Fenmore & Co., 535 U.S. 722 (2002). The "adiabatically coupled" limitation was added specifically to overcome the examiner's § 102(a)(1) rejection based on the Tanaka prior art. The applicants' remarks explicitly characterized the limitation as requiring "a gradual, continuous taper transition" — language that disclaims non-continuous coupling approaches. The scope of surrender encompasses non-continuous coupling mechanisms, including segmented taper designs such as those used in Lumos's accused products. The three Festo rebuttal exceptions a

PRIVILEGED AND CONFIDENTIAL — ATTORNEY WORK PRODUCT

MEMORANDUM

C&H MATTER NO.: 1014-00003

Record

Parties

Dates

Amounts