1014-00003 · Optiwave Semiconductor Corp · Expert Engagement

reasonable-royalty-georgia-pacific-draft.docx DOCUMENT Litigation

34 / 83 ← → ✕

PRIVILEGED AND CONFIDENTIAL — ATTORNEY WORK PRODUCT

Draft Memorandum: Reasonable Royalty Analysis — Georgia-Pacific Factor Framework

January 2023

Right-click to update Table of Contents

I. Purpose and Scope

This memorandum has been prepared to develop and document the fifteen-factor analytical framework established in Georgia-Pacific Corp. v. United States Plywood Corp., 318 F. Supp. 1116 (S.D.N.Y. 1970), for use by Dr. Lawrence Kessler of Apex Claim Analytics LLC (Chicago, Illinois) in preparing his reasonable royalty damages expert report and testimony in this matter. Dr. Kessler has been retained by Optiwave Semiconductor Corporation ("OWS" or "Optiwave") as its designated damages expert. This memorandum is intended to provide Dr. Kessler with a structured preliminary analysis of each Georgia-Pacific factor, marshaling the factual record developed to date and identifying areas requiring further investigation as discovery proceeds.

The scope of this memorandum encompasses an analysis of each of the fifteen Georgia-Pacific factors in the context of the four patents asserted by OWS in this action — U.S. Patent Nos. 10,234,567 (the "'567 Patent"), 10,891,234 (the "'234 Patent"), 11,045,678 (the "'678 Patent"), and 11,312,890 (the "'890 Patent") — and the two accused products manufactured and sold by Defendant Lumos Photonics, Inc. ("Lumos" or "Lumos Photonics"): the LumosBridge™ 800G Optical Engine and the LumosEdge™ 400ZR+ Pluggable Transceiver. The analysis is structured around the hypothetical negotiation between OWS (as willing licensor) and Lumos (as willing licensee) that the Georgia-Pacific framework requires.

The applicable legal standard for this analysis is rooted in 35 U.S.C. § 284, which provides that "[u]pon finding for the claimant the court shall award the claimant damages adequate to compensate for the infringement, but in no event less than a reasonable royalty for the use made of the invention by the infringer." The reasonable royalty represents the floor for patent damages, and is determined through the construct of a hypothetical negotiation between the patentee and the infringer at the time infringement began. The seminal framework for this analysis was set forth in Georgia-Pacific Corp. v. United States Plywood Corp., 318 F. Supp. 1116 (S.D.N.Y. 1970), which identified fifteen factors bearing on the determination of a reasonable royalty. That framework has been adopted and applied by the Federal Circuit in numerous subsequent decisions and remains the standard analytical tool for reasonable royalty determinations.

For purposes of this memorandum, the hypothetical negotiation date is assumed to be immediately before the earliest date of infringement. For the '567 Patent, the '234 Patent, and the '678 Patent, the hypothetical negotiation date is set at Q3 2021, immediately before the commencement of production of the LumosEdge 400ZR+ Pluggable Transceiver. For the '890 Patent, which was not issued until April 26, 2022, the hypothetical negotiation date is set at Q1 2022, immediately before the first shipment of the LumosBridge 800G Optical Engine, as infringement could not predate issuance of the patent. The implications of these different negotiation dates are discussed further in Section III.B below.

Important Note Regarding Status: This memorandum was prepared in January 2023. The analysis was paused and has not been finalized following the Court's Markman Order issued February 17, 2023 (Dkt. 87), which substantially narrowed the scope of the asserted claims by adopting Defendant-aligned constructions for eight of eleven disputed claim terms. The implications of the Markman Order for this damages analysis are addressed in Section X.

II. Summary of Key Financial Data

The following summary sets forth the principal financial data points relevant to the reasonable royalty analysis. These figures are drawn from publicly available financial disclosures, discovery responses produced to date, and pre-litigation valuations commissioned by OWS. This summary is intended to provide a convenient reference for the factor-by-factor analysis that follows.

Optiwave Semiconductor Corporation (OWS):

• FY 2022 Revenue: $187.4 million

• Headquarters: San Jose, California

• Approximate employees: 620

• Majority ownership: Vantage Point Capital Partners Fund IV, LP (approximately 62% equity interest)

• Patent portfolio: 87 issued U.S. utility patents; 14 pending applications

• Key products: OWS-400ZR and OWS-800LR coherent optical transceiver chipsets

• Fabrication: fabless model; chips fabricated by Pinnacle Semiconductor Foundry Co., Ltd. (Hsinchu, Taiwan) on a 45nm silicon-on-insulator (SOI) process

• Pre-litigation patent portfolio valuation (Clearpath IP Valuation Services, Inc.; lead analyst Jennifer Armitage, CFA, CLP): $215 million (full portfolio); $42.8 million (four asserted patents collectively)

• OWS's preliminary estimated reasonable royalty rate: 7.5%

Lumos Photonics, Inc. (Lumos):

• FY 2022 Revenue: $312.8 million

• Headquarters: Austin, Texas

• Fabrication facility: Hillsboro, Oregon

• Approximate employees: 1,450

• Publicly traded on Nasdaq under ticker symbol LMPH since 2021 IPO

• Market capitalization as of complaint filing (June 3, 2022): approximately $1.87 billion

• Market capitalization as of date of this memorandum (January 2023): in excess of $2.0 billion

• Largest institutional shareholder: Ridgeline Asset Management, LLC (approximately 9.2% stake)

• Operating model: integrated device manufacturer (IDM) of silicon photonics components

Accused Product Financial Data:

Preliminary Damages Calculations:

• Base annual royalty: $148.6 million × 7.5% = $11.145 million (applied to full CY 2022 accused product revenue)

• Pre-suit damages (Q1–Q2 2022): $62.3 million × 7.5% = $4.6725 million

• Total claimed damages per Complaint: in excess of $75 million, inclusive of pre-suit damages, post-suit damages accruing through trial, and potential willfulness enhancement (up to treble damages under 35 U.S.C. § 284)

These financial data points are referenced throughout the factor analysis that follows and will be refined as additional discovery is completed, particularly with respect to Lumos's product-level cost, margin, and profit data, which are expected to be produced in response to OWS's outstanding discovery requests.

III. Hypothetical Negotiation Framework

A. Legal Standard

The reasonable royalty determination under 35 U.S.C. § 284 is anchored in the construct of a hypothetical negotiation between a willing licensor and a willing licensee, negotiating at arm's length at the time infringement began. This is not an inquiry into what the parties actually would have agreed to — it is a legal construct designed to approximate the royalty that would have emerged from a voluntary negotiation conducted under certain assumed conditions. See Georgia-Pacific Corp. v. United States Plywood Corp., 318 F. Supp. 1116, 1120 (S.D.N.Y. 1970); Lucent Techs., Inc. v. Gateway, Inc., 580 F.3d 1301, 1324–25 (Fed. Cir. 2009).

The hypothetical negotiation proceeds under the foundational assumption that both parties are aware that the patents at issue are valid and infringed. See Lucent, 580 F.3d at 1325. This assumption eliminates the parties' actual uncertainty about patent validity and infringement scope, which in a real-world negotiation would tend to depress the royalty rate. The construct thus ensures that the reasonable royalty reflects the full value of the patented technology, rather than a rate discounted for litigation risk.

Both parties are further assumed to be willing — the licensor willing to grant a license, and the licensee willing to take one — and the negotiation is conducted at arm's length, free from any duress, compulsion, or other factors that might distort the bargain. The reasonable royalty is the rate that would have been agreed upon, not the rate that one party might have wished to impose unilaterally.

Critical to this analysis, the Federal Circuit has emphasized that the hypothetical negotiation must be grounded in the facts of the case and supported by sound economic reasoning. Speculative or conclusory damages theories are insufficient. Uniloc USA, Inc. v. Kestrel Partners LLC, 632 F.3d 1292, 1315 (Fed. Cir. 2011). In particular, the Federal Circuit has scrutinized the application of the entire market value rule, holding that "the entire market value rule allows for the recovery of damages based on the value of an entire apparatus containing several features, when the feature patented constitutes the basis for customer demand." Id. at 1318 (internal quotation marks omitted). Where the patented feature is not the sole basis for customer demand, apportionment is required — either through the selection of an appropriately narrower royalty base or through a royalty rate that reflects the relative contribution of the patented feature to the overall product value. These principles are discussed further in Section VI below.

B. Assumed Negotiation Date

The hypothetical negotiation date is a critical parameter in the Georgia-Pacific analysis because it defines the universe of information available to the parties and the economic conditions under which the negotiation occurs. For each patent-in-suit, the negotiation date is set immediately before the earliest date of infringement.

'567 Patent (U.S. Patent No. 10,234,567): This patent was filed on March 14, 2017 and issued on March 19, 2019. Because the patent was already issued well before the commencement of Lumos's accused infringing activity, the hypothetical negotiation date is determined by the date of first infringement, which corresponds to the commencement of production of the LumosEdge 400ZR+ in Q3 2021.

'234 Patent (U.S. Patent No. 10,891,234): This patent was filed on September 8, 2017 and issued on January 14, 2021. Similarly, the '234 Patent issued before the commencement of production of the LumosEdge, and the hypothetical negotiation date is set at Q3 2021.

'678 Patent (U.S. Patent No. 11,045,678): This patent was filed on June 22, 2018 and issued on June 29, 2021, shortly before the LumosEdge entered production in Q3 2021. The hypothetical negotiation date for the '678 Patent is likewise set at Q3 2021.

'890 Patent (U.S. Patent No. 11,312,890): This patent was filed on November 15, 2018 and did not issue until April 26, 2022. Because infringement liability under 35 U.S.C. § 271 generally does not attach until after the patent issues, the hypothetical negotiation date for the '890 Patent is set at Q1 2022, immediately before Lumos's first shipment of the LumosBridge 800G. The '890 Patent presents a unique issue: its hypothetical negotiation postdates the negotiation for the other three patents by approximately six months. At the Q1 2022 negotiation date, the parties would have had access to additional information not available at the Q3 2021 date, including the actual market reception of the LumosEdge product, Lumos's Q3–Q4 2021 revenue performance, and the announced launch timeline for the LumosBridge 800G. This additional information may slightly affect the royalty analysis for the '890 Patent relative to the other three patents.

As a practical matter, however, Dr. Kessler may determine that the differences in negotiation dates do not materially alter the overall reasonable royalty determination, particularly if the analysis focuses on a blended royalty rate applicable to all four patents collectively. The question of whether to present a single consolidated royalty rate or separate rates for each patent should be discussed with Dr. Kessler during the expert report preparation process.

C. Parties' Bargaining Positions at Hypothetical Negotiation

Understanding the parties' respective bargaining positions at the hypothetical negotiation date is essential to assessing what royalty rate would have emerged from an arm's-length negotiation.

Optiwave Semiconductor Corporation (OWS) — The Willing Licensor:

At the hypothetical negotiation date, OWS was a fabless semiconductor company headquartered in San Jose, California, with approximately 620 employees and FY 2022 revenue of $187.4 million. OWS's core business is the design and development of silicon photonics chipsets for high-speed optical transceivers used in data center interconnects and telecommunications applications. Its key products — the OWS-400ZR and OWS-800LR coherent optical transceiver chipsets — are manufactured by Pinnacle Semiconductor Foundry Co., Ltd., a leading foundry located in Hsinchu, Taiwan, using a 45nm silicon-on-insulator (SOI) process.

OWS is majority-owned (approximately 62%) by Vantage Point Capital Partners Fund IV, LP, a technology-focused private equity fund with a vintage of 2018 and a fund size of $2.1 billion. Vantage Point's initial equity investment in OWS totaled $74.5 million across Series C and D financing rounds. The Vantage Point investment thesis for OWS includes the company's intellectual property portfolio as a significant component of enterprise value, a point that is directly relevant to OWS's incentives and bargaining posture in any licensing negotiation.

OWS held a substantial patent portfolio consisting of 87 issued U.S. utility patents and 14 pending applications at the time of the hypothetical negotiation. The portfolio was valued by Clearpath IP Valuation Services, Inc. (Boston, Massachusetts; lead analyst Jennifer Armitage, CFA, CLP) at $215 million for the full portfolio and $42.8 million for the four asserted patents collectively. This valuation, while conducted on a pre-litigation basis and subject to the limitations discussed in Section VIII, provides an independent third-party assessment of the economic significance of OWS's patent assets.

As a willing licensor, OWS would have been motivated to extract maximum royalty compensation, recognizing that any license to the patents-in-suit would enable a direct competitor to practice the patented technology and thereby potentially erode OWS's own market share in the highly competitive silicon photonics transceiver market.

Lumos Photonics, Inc. (Lumos) — The Willing Licensee:

At the hypothetical negotiation date, Lumos was an integrated device manufacturer (IDM) of silicon photonics components headquartered in Austin, Texas, with approximately 1,450 employees and FY 2022 revenue of $312.8 million. Unlike OWS, Lumos operates its own fabrication facility in Hillsboro, Oregon, giving it vertical integration capabilities that include control over the manufacturing process and yields of its silicon photonic integrated circuits.

Lumos completed its initial public offering on the Nasdaq exchange (ticker: LMPH) in 2021 and had a market capitalization of approximately $1.87 billion as of the filing of the Complaint on June 3, 2022. Its largest institutional shareholder is Ridgeline Asset Management, LLC, which holds an approximately 9.2% stake. By the date of this memorandum, Lumos's market capitalization has grown to in excess of $2.0 billion, reflecting continued investor confidence in the company's growth trajectory.

Lumos directly competes with OWS in the high-speed optical transceiver market, targeting hyperscale data center customers for its LumosBridge 800G Optical Engine and LumosEdge 400ZR+ Pluggable Transceiver products. The accused products have been commercially successful, generating approximately $148.6 million in combined revenue in CY 2022. As a willing licensee, Lumos would have been motivated to negotiate the lowest possible royalty rate while securing the right to continue manufacturing and selling its accused products.

The competitive dynamic between the parties is a critical factor in the hypothetical negotiation. In a negotiation between direct competitors, the licensor's willingness to grant a license is tempered by the recognition that doing so enables competition in its own market, while the licensee's willingness to pay is informed by the value of avoiding the risk of injunctive relief and the cost of designing around the patented technology.

IV. Georgia-Pacific Factor Analysis

The following sections address each of the fifteen Georgia-Pacific factors as applied to the facts of this case. For each factor, we identify the relevant evidence, assess its directional impact on the reasonable royalty determination (i.e., whether it favors a higher royalty, a lower royalty, or is neutral), and note any areas where additional evidence is needed.

Factor 1: Royalties Received by the Patentee for Licensing the Patents-in-Suit

The first Georgia-Pacific factor considers the royalties received by the patentee for the licensing of the patent in suit, proving or tending to prove an established royalty. An established royalty is the most reliable indicator of a reasonable royalty because it reflects the actual market value placed on the patented technology by arm's-length negotiation.

OWS has not previously licensed the four asserted patents — U.S. Patent Nos. 10,234,567; 10,891,234; 11,045,678; and 11,312,890 — to any third party. No established royalty rate exists for these specific patents. The absence of any licensing history for the patents-in-suit means that there is no direct evidence from which to infer the market rate for the patented technology. This is not unusual in cases involving relatively recently issued patents held by companies that pursue an exclusionary patent strategy, as OWS does.

OWS's broader patent portfolio has been the subject of two cross-license agreements with non-competitor companies. Based on information obtained to date, these cross-licenses involved different technology fields — specifically, modulator driver architectures for telecommunications applications, rather than the data center interconnect technologies at issue in this case. The details of these cross-licenses, including their financial terms and the specific patents covered, are subject to further confirmation with Dr. Kessler upon completion of document review. However, the limited comparability of these cross-licenses — arising from differences in technology field, competitive posture of the counterparties, and license scope — substantially reduces their probative value as indicators of the appropriate royalty rate for the patents-in-suit.

The absence of an established royalty for the patents-in-suit does not preclude the determination of a meaningful reasonable royalty. Rather, it shifts the analytical burden to the remaining Georgia-Pacific factors and to comparable license evidence from the broader market. See Georgia-Pacific, 318 F. Supp. at 1120 (identifying fifteen factors, of which an established royalty is only the first).

Assessment: This factor is neutral to modestly favorable for OWS. The absence of a prior license does not set a ceiling on the royalty; instead, it necessitates reliance on other indicia of value. OWS can argue that its refusal to license reflects the high value of the patents, while Lumos may argue that the absence of any licensing activity suggests the patents have not been independently valued by the market.

Factor 2: Rates Paid by the Licensee for Use of Comparable Patents

The second factor examines the rates paid by the licensee for the use of other patents comparable to the patent in suit. This factor looks to the licensee's own licensing history for insight into what rates the market has established for similar technology.

Lumos's existing patent license agreements are a critical source of comparable royalty data. OWS's First Set of Requests for Production, Nos. 67–78, served on November 21, 2022, sought the production of all license agreements involving silicon photonics technology, optical transceiver technology, and related semiconductor intellectual property. As of the date of this memorandum, Lumos's responses to these requests are pending, and the full scope of Lumos's licensing history has not been established.

Preliminary analysis of publicly available licensing data in the silicon photonics space suggests that royalty rates for component-level patent licenses generally fall in the range of 3% to 10% of net sales, with the rate depending on factors including portfolio breadth, technology criticality, the competitive relationship between licensor and licensee, and the geographic scope of the license. This range is drawn from public filings, analyst reports, and prior litigation outcomes in the semiconductor industry, and is subject to refinement as more specific data becomes available.

Lumos's most recent Form 10-K, filed with the Securities and Exchange Commission, references certain intellectual property licensing costs as a component of its cost of goods sold. However, the specific royalty rates, licensors, and patents covered are not broken out in the public filing and must be obtained through discovery. The 10-K disclosure does confirm that Lumos has entered into IP licensing arrangements, which will provide useful data points once the underlying agreements are produced.

Assessment: This factor is pending completion of discovery. The preliminary market data suggesting a 3%–10% range is broadly consistent with OWS's proposed 7.5% rate, but the analysis cannot be finalized until Lumos's actual license agreements are reviewed and assessed for comparability. Dr. Kessler should be prepared to analyze each of Lumos's licenses for comparability to the hypothetical license at issue in this case.

Factor 3: Nature and Scope of the License

The third factor addresses the nature and scope of the license, including whether it is exclusive or non-exclusive, whether it is restricted or unrestricted in terms of territory or field of use, and who is granted the license.

For purposes of the hypothetical negotiation, the assumed license is a non-exclusive, worldwide license limited to the four patents-in-suit. The license would cover the manufacture, use, sale, offer for sale, and importation of products embodying the claimed inventions — specifically, Lumos's LumosBridge 800G Optical Engine and LumosEdge 400ZR+ Pluggable Transceiver products, along with reasonably foreseeable derivative products incorporating the patented technology.

The non-exclusive nature of the hypothetical license is significant because a non-exclusive license does not prevent OWS from granting additional licenses to other parties or from continuing to practice the patented inventions itself. Non-exclusive licenses generally command lower royalty rates than exclusive licenses, all else being equal, because the licensee does not receive the benefit of exclusivity. However, the practical significance of this distinction is tempered in this case by the fact that OWS has not licensed the patents-in-suit to any other party and has historically maintained an exclusionary posture regarding its core patent portfolio.

The hypothetical license is limited to the four patents-in-suit and to specific products, rather than being a broad portfolio cross-license or a field-of-use license. A narrower license covering a smaller number of patents and limited to specific products would tend to yield a lower total royalty payment than a broad portfolio license, but the per-patent or per-technology royalty rate may actually be higher for a narrower license because the licensee is obtaining rights to specific, targeted technology rather than a portfolio of mixed-value patents.

With respect to royalty structure, a running royalty (calculated as a percentage of net sales on an ongoing basis) is recommended over a lump-sum payment. The running royalty structure is more appropriate in this context because: (i) Lumos's accused products are generating ongoing revenue and are expected to continue doing so throughout the remaining patent life; (ii) a running royalty aligns the royalty payment with the licensee's actual use of the patented technology and avoids the need for speculative estimates of future sales volumes; and (iii) running royalties are the more commonly applied structure in the silicon photonics and semiconductor industries for component-level patent licenses. Dr. Kessler should address the running royalty versus lump-sum question in his expert report and be prepared to testify on the economic basis for the recommended structure.

Assessment: This factor modestly favors a lower royalty rate relative to what would be appropriate for a broader or exclusive license, but the narrow scope of the license is offset by the targeted, high-value nature of the patented technology covered. The recommended running royalty structure is consistent with industry practice and provides for fair compensation proportional to Lumos's actual sales.

Factor 4: Licensor's Established Policy of Maintaining Patent Monopoly

The fourth factor examines the licensor's established policy and marketing program to maintain its patent monopoly by not licensing others to use the invention, or by granting licenses under special conditions designed to preserve that monopoly. This factor recognizes that a licensor who has consistently refused to license its patents would require a higher royalty to compensate for the departure from its exclusionary strategy.

OWS has historically maintained a strong exclusionary posture regarding its core silicon photonics patent portfolio. As noted under Factor 1, OWS has not granted any licenses for the four patents-in-suit. The two cross-license agreements involving OWS's broader portfolio covered different technology fields (modulator driver architectures for telecommunications) and involved non-competitor counterparties, and thus do not represent a departure from OWS's general policy of maintaining exclusivity over its core data center interconnect patents.

Internal communications produced in discovery reinforce OWS's exclusionary licensing policy. Dr. Priya Ramanathan, OWS's Chief Executive Officer, has emphasized in multiple internal memoranda and board presentations that the company's patent portfolio is a "strategic competitive asset" that provides OWS with a critical differentiator in the silicon photonics market. Ted Marchetti, OWS's Vice President of intellectual Property, has echoed this position in correspondence with external counsel and in internal IP strategy documents, consistently advocating against licensing OWS's core patents to competitors.

The investment thesis of Vantage Point Capital Partners Fund IV, LP — OWS's majority shareholder — further supports the importance of the patent portfolio to OWS's enterprise value. Vantage Point's initial equity investment of $74.5 million across Series C and Series D financing rounds was premised in part on the strength and defensibility of OWS's intellectual property position. The Clearpath IP Valuation Services pre-litigation valuation of $215 million for the full patent portfolio — representing a significant portion of OWS's total enterprise value — confirms that the patent portfolio is viewed by OWS and its investors as a foundational asset, not a commodity to be freely licensed.

Assessment: This factor weighs strongly in favor of a higher royalty rate. OWS's consistent refusal to license its core patents, combined with the strategic importance of the patent portfolio to OWS's competitive position and enterprise value, indicates that OWS would demand substantial compensation for any departure from its exclusionary policy. A willing licensor in OWS's position would insist on a premium royalty rate that reflects the competitive harm of enabling a direct competitor's products.

Factor 5: Commercial Relationship Between the Licensor and Licensee

The fifth factor considers the commercial relationship between the licensor and licensee — specifically, whether they are competitors in the same territory and in the same line of business, or whether they are inventor and promoter.

OWS and Lumos are direct competitors in the silicon photonics market for high-speed data center interconnects and telecommunications applications. Both companies design and sell optical transceiver chipsets and engines targeting the hyperscale data center market, and both compete for design wins with the same universe of major customers, including hyperscale cloud service providers and data center switch manufacturers.

OWS's OWS-400ZR coherent optical transceiver chipset competes directly with Lumos's LumosEdge 400ZR+ Pluggable Transceiver in the 400G coherent optical interconnect segment. Similarly, OWS's OWS-800LR platform competes with Lumos's LumosBridge 800G Optical Engine in the emerging 800G co-packaged optics segment. The products occupy overlapping market positions and are evaluated by customers as substitutes or alternatives in competitive procurement processes.

The competitive relationship between OWS and Lumos is direct and intense. Every unit of LumosEdge or LumosBridge sold by Lumos represents a potential lost sale of OWS-400ZR or OWS-800LR by OWS. The silicon photonics transceiver market, while rapidly growing, is characterized by a limited number of design wins with major hyperscale customers, and the loss of a design win to a competitor can result in years of lost revenue for the displaced supplier. The competitive dynamic is further intensified by the high switching costs faced by data center operators once a particular transceiver platform is deployed at scale.

In the patent licensing context, it is well established that a license to a direct competitor typically commands a higher royalty rate than a license to a non-competitor. See, e.g., Georgia-Pacific, 318 F. Supp. at 1120 (identifying the "commercial relationship between the licensor and licensee" as a factor bearing on the reasonable royalty). The rationale is straightforward: a license to a competitor not only provides the licensee with access to the patented technology but also enables the licensee to compete more effectively against the licensor, thereby inflicting competitive harm that must be compensated through the royalty.

Assessment: This factor weighs strongly in favor of a higher royalty rate. The direct competitive relationship between OWS and Lumos means that any license from OWS to Lumos would enable Lumos to practice the patented technology in products that directly compete with OWS's own products. A willing licensor in OWS's position would demand a substantial premium to compensate for the competitive injury resulting from the license.

Factor 6: Effect of Selling the Patented Specialty in Promoting Sales of Other Products

The sixth factor examines the effect of selling the patented specialty in promoting sales of the licensor's other products, the existing value of the invention to the licensor as a generator of sales of its non-patented items, and the extent of such derivative or convoyed sales.

Lumos's accused products — the LumosBridge 800G Optical Engine and the LumosEdge 400ZR+ Pluggable Transceiver — are standalone revenue-generating products, not accessories or components sold primarily to drive sales of other Lumos product lines. Each product is sold as an independent commercial offering with its own pricing, specification sheets, and customer base. There is no evidence that the accused products serve primarily as loss leaders or promotional items designed to generate convoyed sales of unrelated Lumos products.

However, the data center optical interconnect market exhibits significant platform-stickiness and lock-in effects that are relevant to this factor. Adoption of Lumos's transceiver products by hyperscale data center operators for a given deployment generation creates strong incentives for those operators to continue using Lumos products in subsequent deployment cycles. This platform-stickiness arises from several sources: (i) the cost and complexity of requalifying alternative transceiver suppliers; (ii) the interoperability testing required to validate new transceiver platforms against existing switch and network infrastructure; and (iii) the preference of data center operators for vendor continuity to simplify supply chain management and reduce operational risk.

As a result, every design win that Lumos secures with its current-generation products has an outsized impact on Lumos's competitive position for future product generations. This downstream competitive effect is not captured by the current-period revenue of the accused products alone. From OWS's perspective as a willing licensor, licensing the patented technology to Lumos not only enables Lumos's current products but also strengthens Lumos's competitive position for future generations of transceiver products, compounding the competitive harm to OWS over time.

OWS's own products function similarly as standalone chipsets and engines, and OWS benefits from the same platform-stickiness dynamics when it secures design wins with hyperscale customers. The competitive symmetry between the parties on this factor means that both OWS and Lumos would recognize the long-term competitive implications of the license during the hypothetical negotiation.

Assessment: This factor weighs modestly in favor of a higher royalty rate. While the accused products are not promotional items driving convoyed sales of other products, the platform-stickiness and design-win momentum effects in the data center transceiver market create downstream competitive consequences that a willing licensor would seek to address through the royalty rate.

Factor 7: Duration of the Patent and the Term of the License

The seventh factor considers the duration of the patent and the term of the license. Longer remaining patent life provides the licensee with a longer period of protection against competition from the patented technology, which generally supports a higher royalty rate.

The four asserted patents have the following relevant dates and estimated remaining life:

All four patents have substantial remaining life — approximately 14 to 17 years from the hypothetical negotiation date(s). This extended remaining life means that the patented technology will be enforceable throughout the expected commercial lifetime of the accused products and through multiple future product generations. In a rapidly evolving technology market such as silicon photonics, a 14-to-17-year patent term encompasses multiple product generation cycles and extends well beyond the current market horizon.

The hypothetical license term would most likely be coextensive with the remaining patent terms, as is customary for patent licenses in the semiconductor industry. A license running for the full remaining patent term provides the licensee with certainty of rights and avoids the need for periodic renegotiation, while ensuring that the licensor receives compensation for the full duration of the patent monopoly.

Assessment: This factor weighs in favor of a higher royalty rate. The substantial remaining life of all four patents means that the licensee would benefit from the license for an extended period, justifying a meaningful royalty that reflects the long duration of the patented technology's commercial relevance. The long remaining life also increases the total present value of the royalty stream, which is a relevant consideration in calibrating the appropriate royalty rate.

Factor 8: Established Profitability of the Product Made Under the Patent; Its Commercial Success; and Its Current Popularity

The eighth factor examines the established profitability of the product made under the patent, its commercial success, and its current popularity. Commercially successful products practicing the patented technology are evidence that the patented features contribute meaningful value, supporting a higher royalty rate.

Lumos's accused products have demonstrated strong commercial success. The LumosBridge 800G and LumosEdge 400ZR+ generated combined revenue of approximately $148.6 million in CY 2022, representing approximately 47.5% of Lumos's total FY 2022 revenue of $312.8 million. The accused products are thus a core driver of Lumos's business, not peripheral or marginal product lines.

The LumosBridge 800G Optical Engine first shipped in Q1 2022 and is priced at $1,200 to $1,800 per unit depending on configuration. The LumosBridge is positioned as a next-generation co-packaged optics platform designed for integration with hyperscale data center switch ASICs, targeting the emerging 800G and beyond interconnect segment. The LumosBridge has attracted significant attention from hyperscale customers and industry analysts, reflecting the market's strong demand for co-packaged optics solutions capable of supporting the bandwidth requirements of AI/ML workloads.

The LumosEdge 400ZR+ has been in production since Q3 2021 and is priced at $600 to $900 per unit depending on configuration. The LumosEdge is a QSFP-DD form factor pluggable transceiver designed for metro and data center interconnect (DCI) applications. It has achieved broad market adoption and competes directly with OWS's OWS-400ZR product for 400G coherent optical interconnect design wins.

Lumos's overall revenue growth trajectory and its successful Nasdaq IPO in 2021 further indicate strong market demand and investor confidence in the company's products and technology platform. The market capitalization growth from approximately $1.87 billion at the time of the Complaint to over $2.0 billion as of the date of this memorandum reinforces the commercial attractiveness and growth prospects of Lumos's silicon photonics product portfolio, of which the accused products are the centerpiece.

For comparison, OWS's own products employing the patented technology — the OWS-400ZR and OWS-800LR — have also achieved commercial traction, with OWS generating $187.4 million in total FY 2022 revenue. While OWS's product-level revenue breakdown has not been disclosed publicly, the competitive performance of OWS's products in the same market segments further corroborates the commercial significance of the patented technology.

Assessment: This factor weighs in favor of a moderate-to-higher royalty rate. The strong commercial performance of the accused products, both in absolute revenue terms and as a proportion of Lumos's total revenue, indicates that the patented technology is commercially significant and contributes meaningfully to the products' market success. The profitability of the accused products (to be confirmed through discovery of Lumos's cost and margin data) will further inform this assessment.

Factor 9: Utility and Advantages of the Patent Property Over Old Modes or Devices

The ninth factor examines the utility and advantages of the patent property over the old modes or devices, if any, that had been used for working out similar results. This factor assesses the technical superiority of the patented inventions over the prior art and the availability of alternatives.

Each of the four asserted patents addresses a distinct performance-critical aspect of silicon photonics transceiver design:

'567 Patent — Integrated Waveguide Coupler with Tapered Mode Converter: The '567 Patent claims an integrated waveguide coupler incorporating an adiabatic taper design for fiber-to-chip optical coupling. The claimed adiabatic taper mode converter provides significantly lower optical coupling losses compared to prior art edge couplers, which relied on abrupt transitions between fiber and waveguide modes. The adiabatic taper design enables efficient, low-loss coupling between standard single-mode fiber and sub-micron silicon waveguides, which is critical for high-speed optical transceivers where every decibel of coupling loss directly impacts link budget and system performance.

'234 Patent — Ge-on-Si Photodetector with Graded Buffer: The '234 Patent claims a germanium-on-silicon photodetector employing a continuously graded buffer layer between the silicon substrate and the germanium absorber. The continuously graded buffer reduces threading dislocation density at the Ge/Si interface compared to prior art approaches using abrupt or step-graded interfaces, thereby enhancing photodetector responsivity and bandwidth. This is particularly important for 50+ Gbaud photodetection, where defect-related dark current and bandwidth limitations are significant performance constraints.

'678 Patent — Multi-Channel WDM on Silicon: The '678 Patent claims a monolithically integrated wavelength division multiplexing (WDM) system on a silicon photonic chip, eliminating the need for external discrete WDM multiplexing and demultiplexing components. Monolithic integration reduces package size, lowers manufacturing cost, and eliminates the optical losses and alignment challenges associated with discrete WDM components. This is particularly advantageous for multi-channel transceiver modules where aggregate bandwidth is achieved through wavelength parallelism.

'890 Patent — MZM Driver Co-Integration: The '890 Patent claims a Mach-Zehnder modulator (MZM) driver architecture in which the pre-emphasis equalization stage is co-integrated with the optical modulator on a common substrate. Co-integration reduces parasitic capacitance at the interface between the driver and modulator, improving high-frequency modulation performance. This is critical for achieving the high baud rates required for 400G and 800G optical signaling, where parasitic capacitance at the driver-modulator interface is a dominant bandwidth limiter.

Each of these patented inventions offers measurable performance advantages over prior art approaches. However, it must be acknowledged that design-around alternatives exist for each patent. For the '567 Patent, alternative coupling approaches include segmented tapers, grating couplers, and inverse taper designs that achieve mode conversion through non-adiabatic transitions. For the '234 Patent, step-graded germanium buffer layers and selective epitaxial growth techniques are known alternatives, albeit with performance trade-offs. For the '678 Patent, hybrid integration approaches using external arrayed waveguide gratings (AWGs) or thin-film filters can achieve WDM functionality without monolithic integration. For the '890 Patent, flip-chip bonded driver ICs and wire-bonded driver-modulator assemblies represent alternative approaches, though with higher parasitic capacitance.

The availability of these alternatives is relevant because it affects the licensee's outside option in the hypothetical negotiation — if Lumos could readily design around the asserted patents, it would face less pressure to accept a license and would be willing to pay a lower royalty. Conversely, if the design-around alternatives impose significant performance penalties, increased manufacturing cost, or longer development timelines, the patented technology's value to the licensee is commensurately higher.

Assessment: This factor weighs in favor of a moderate royalty. The patented inventions offer genuine and measurable technical advantages over prior art approaches, but the availability of design-around alternatives for each patent prevents this factor from commanding the full weight it would carry if the patented technology were truly indispensable. Dr. Whitfield's technical expert testimony will be critical in quantifying the performance advantages and the cost and feasibility of alternatives.

Factor 10: Nature of the Patented Invention, Character of Commercial Embodiment, and Benefits to Users

The tenth factor considers the nature of the patented invention, the character of the commercial embodiment of it as owned and produced by the licensor, and the benefits to those who have used the invention.

The patented inventions relate to fundamental building blocks of silicon photonic integrated circuits (PICs) used in high-speed data center optical interconnects. These building blocks — fiber-to-chip coupling, photodetection, wavelength division multiplexing, and modulator drive — are core functional elements that are present in essentially all modern silicon photonic transceiver designs. The inventions are not directed to peripheral or ornamental features; they address the physical and electrical interfaces that determine the transceiver's optical performance, bandwidth, and power efficiency.

The commercial embodiment of the inventions is demonstrated by OWS's chipsets manufactured by Pinnacle Semiconductor Foundry Co., Ltd. on a 45nm SOI process. The OWS-400ZR and OWS-800LR products incorporate the patented technologies and have been successfully deployed in production volumes, demonstrating the practical applicability and manufacturability of the inventions at commercial scale. The fact that the patented inventions have been reduced to production-ready silicon validates their commercial utility and distinguishes them from theoretical or laboratory-stage concepts.

End-user benefits derived from the patented technologies are substantial and directly aligned with the key performance metrics that drive purchasing decisions among hyperscale data center operators. These benefits include:

• Higher data rates: The patented technologies collectively enable 400G and 800G per-lane signaling, supporting the aggregate bandwidth requirements of modern hyperscale data center architectures driven by AI/ML workload expansion.

• Lower power consumption: Monolithic integration of WDM components ('678 Patent) and co-integrated driver architectures ('890 Patent) reduce power consumption per bit, which is a critical metric for data centers where power and cooling represent a significant proportion of total operating cost.

• Reduced form factor: Integration of previously discrete components onto a single silicon photonic chip reduces the physical size of the transceiver module, enabling higher port density in switch and router line cards.

• Lower cost per bit: The combination of monolithic integration, reduced packaging complexity, and higher data rates drives down the cost per bit of transmitted data, which is the ultimate economic metric for data center operators.

The data center optical interconnect market is experiencing rapid growth driven by the expansion of hyperscale data centers, the proliferation of AI/ML training and inference workloads, and the transition from 100G to 400G and 800G interconnect speeds. This market growth amplifies the commercial significance of the patented technologies, as they address performance-critical bottlenecks in the transceiver value chain.

Assessment: This factor supports a moderate royalty reflecting the importance of the patented features to the overall product value proposition. The patented technologies are core functional elements of the transceiver products, and the end-user benefits are directly aligned with the most commercially significant performance metrics. The factor does not by itself indicate a specific royalty rate but reinforces the conclusion that the patented features contribute meaningful value to the accused products.

Factor 11: Extent to Which the Infringer Has Made Use of the Invention and Evidence of the Value of That Use

The eleventh factor examines the extent to which the infringer has made use of the invention and any evidence probative of the value of that use. This factor looks at both the volume of the infringer's use and the centrality of the patented technology to the infringer's products and business.

Lumos's accused products — the LumosBridge 800G and LumosEdge 400ZR+ — generated combined revenue of approximately $148.6 million in CY 2022. This represents substantial commercial use of the products alleged to embody the patented inventions. The volume of Lumos's sales demonstrates that the accused products have achieved broad market adoption and are a material component of Lumos's overall business.

The centrality of the patented features to the accused products is a critical sub-issue under this factor. Based on the preliminary infringement analysis, the four patented features — the waveguide coupler with adiabatic taper ('567 Patent), the Ge-on-Si photodetector with graded buffer ('234 Patent), the monolithically integrated WDM system ('678 Patent), and the co-integrated MZM driver ('890 Patent) — are integrated into the core architecture of both the LumosBridge and LumosEdge products. These features are not peripheral add-ons or optional configurations; they are fundamental components of the silicon photonic integrated circuit at the heart of each accused product.

However, the accused products incorporate numerous other innovations and components that are not covered by the asserted patents. Lumos's independent contributions include, among others: (i) Lumos's proprietary digital signal processing (DSP) architecture, which performs forward error correction, equalization, and signal processing functions critical to transceiver performance; (ii) advanced packaging technology, including micro-transfer-printed III-V gain elements; (iii) thermal management solutions necessary for reliable operation in high-density data center environments; and (iv) firmware, software, and management interfaces required for integration into customer network architectures. These non-patented features collectively represent a substantial portion of the value of the accused products.

The apportionment analysis required by this factor — determining what portion of the accused products' value is attributable to the patented features versus other technology — is a complex undertaking that will require both technical and economic expert input. Preliminary estimates, subject to refinement by Dr. Kessler based on detailed technical and economic analysis, suggest that the four patented features collectively account for approximately 15% to 25% of the total value of the accused transceiver platforms. This estimate reflects the importance of the patented building blocks to the transceiver's optical performance while recognizing the substantial value contributed by non-patented technology.

Assessment: This factor weighs in favor of a moderate royalty reflecting significant but not total contribution of the patented technology to the accused products' commercial value. The extensive use of the accused products ($148.6 million in CY 2022 revenue) demonstrates the commercial significance of the technology, but apportionment is required to ensure that the royalty reflects only the value attributable to the patented features.

Factor 12: Portion of the Profit or Selling Price Customarily Allowed for Use of the Invention or Analogous Inventions

The twelfth factor considers the portion of the profit or of the selling price that is customary in the particular business or in comparable businesses to allow for the use of the invention or analogous inventions.

Industry analysis of silicon photonics component licensing reveals the following royalty rate benchmarks:

• Individual patent licenses: Royalty rates typically range from 2% to 8% of net sales for licenses covering individual patents or small groups of closely related patents in the silicon photonics and optical transceiver space.

• Portfolio licenses: For broader portfolio licenses covering core platform technology, royalty rates typically range from 5% to 12% of net sales, reflecting the broader scope and higher aggregate value of multi-patent portfolios.

These ranges are drawn from publicly available data, including SEC filings, analyst reports, and prior litigation outcomes in the semiconductor industry. Dr. Kessler will identify and analyze specific comparable royalty benchmarks in his expert report, drawing on both publicly available data and the comparable license evidence produced through discovery.

It is important to note the constraints imposed by the entire market value rule on the selection of the royalty base. Under Uniloc USA, Inc. v. Kestrel Partners LLC, 632 F.3d 1292 (Fed. Cir. 2011), and subsequent Federal Circuit authority, if the royalty base is the entire product price, the patented feature must be the basis for customer demand for the entire product; otherwise, the royalty must be apportioned to reflect only the value of the patented feature. In this case, the question of the appropriate royalty base is somewhat simplified by the fact that the smallest saleable patent-practicing unit is the integrated photonic chip or engine itself, which is effectively the accused product (the LumosBridge engine or LumosEdge transceiver module). The accused products are not components of larger systems for which the entire market value rule would require the identification of a smaller unit. Accordingly, the accused product revenue of $148.6 million in CY 2022 appears to be an appropriate royalty base, provided that the royalty rate is adjusted to reflect the contribution of the patented features to the overall product value.

Assessment: This factor supports a royalty rate in the range of 5% to 8% when applied to accused product revenue, consistent with industry practice for patent licenses covering core silicon photonics technology. The 7.5% preliminary rate falls within this range and is supported by the market data identified to date. This assessment will be refined by Dr. Kessler based on the comparable license evidence produced through discovery.

Factor 13: Portion of Realizable Profit Attributable to the Invention vs. Non-Patented Elements, Manufacturing Process, Business Risks, and Improvements Added by the Infringer

The thirteenth factor examines the portion of the realizable profit that should be credited to the invention as distinguished from non-patented elements, the manufacturing process, business risks, or significant features or improvements added by the infringer.

Lumos has made substantial independent contributions to the commercial value of its accused products. These contributions include:

Research and Development: Lumos has invested significantly in its own R&D capabilities. Lumos's Chief Technology Officer, Dr. Mei-Lin Chen, is a recognized expert in silicon photonics who holds 14 patents in her own name. Dr. Nikolai Petrov, Lumos's Vice President of Engineering, leads a substantial photonics engineering team that has developed proprietary innovations across the transceiver design stack, from photonic circuit design to packaging and test.

Vertical Integration and Manufacturing: Unlike OWS, which operates as a fabless company relying on Pinnacle Semiconductor Foundry for chip fabrication, Lumos operates its own fabrication facility in Hillsboro, Oregon. This vertical integration gives Lumos significant advantages in process development, yield optimization, and manufacturing cost control that are not derived from the asserted patents. Lumos's investment in its Hillsboro fab represents a substantial capital commitment and a source of competitive differentiation that is independent of the patented technology.

Proprietary Technology: Lumos's LumosBridge platform incorporates several proprietary innovations that are not covered by the asserted patents, including: (i) micro-transfer-printed III-V gain elements, which enable the integration of indium phosphide laser sources onto the silicon photonic platform; (ii) proprietary DSP algorithms for coherent optical signal processing; and (iii) advanced thermal management solutions, including micro-channel liquid cooling for the LumosBridge 800G co-packaged optics module.

Business Risk: Lumos has assumed significant business risk in bringing the LumosBridge 800G to market as a first-generation co-packaged optics product. The co-packaged optics market segment is nascent and involves substantial technical risk (integration with switch ASICs, thermal management challenges) and commercial risk (uncertainty about hyperscale customer adoption timelines). Lumos's willingness to invest in and commercialize co-packaged optics products represents an independent source of value creation that should be credited to Lumos in the royalty analysis.

These independent contributions by Lumos are significant and must be properly accounted for in the apportionment analysis. A royalty rate that failed to credit Lumos for its own innovations, manufacturing capabilities, and business risks would overcompensate OWS at Lumos's expense and would not reflect the outcome of an arm's-length negotiation.

Estimated profit split: Based on the preliminary analysis, approximately 60% to 70% of the accused products' realizable profit is attributable to Lumos's non-patented contributions (including its proprietary technology, manufacturing capabilities, and business risk-taking), with approximately 30% to 40% attributable to the patented technology and favorable market conditions. This profit split is preliminary and will be refined by Dr. Kessler based on detailed economic analysis of the accused products' cost structure, profit margins, and the incremental value of the patented features.

Assessment: This factor favors a lower royalty rate, recognizing Lumos's substantial independent contributions to the commercial value of the accused products. The factor serves as an important counterweight to Factors 4, 5, 8, and 9, which collectively support a higher royalty. The net effect is to moderate the royalty rate to a level that appropriately compensates OWS for the patented technology while giving due credit to Lumos's independent innovations.

Factor 14: Opinion Testimony of Qualified Experts

The fourteenth factor considers the opinion testimony of qualified experts on the relevant issues.

OWS's Technical Expert — Dr. Harold Whitfield:

OWS has retained Dr. Harold Whitfield as its technical expert in this litigation. Dr. Whitfield holds a Ph.D. in Electrical Engineering from the University of California, Berkeley, and is Professor Emeritus at the University of California, Santa Barbara. He has authored over 200 publications in the field of integrated photonics and is a Fellow of the IEEE and the Optical Society of America. Dr. Whitfield operates through his consulting firm, Whitfield Technical Consulting Group, Inc., based in Palo Alto, California.

Dr. Whitfield's declaration, filed in December 2022 in support of OWS's claim construction brief, opines on the technical significance of the patented innovations and their role in the accused products. Dr. Whitfield's opinions address the advantages of the patented approaches over prior art alternatives, the difficulty and cost of designing around the asserted claims, and the centrality of the patented features to the performance characteristics of the accused products. His technical opinions will provide critical input to the apportionment analysis under Factors 11 and 13 and will inform Dr. Kessler's economic analysis of the value attributable to the patented technology.

OWS's Damages Expert — Dr. Lawrence Kessler:

OWS has retained Dr. Lawrence Kessler as its damages expert. Dr. Kessler holds a Ph.D. in Economics from Northwestern University and is a principal at Apex Claim Analytics LLC in Chicago, Illinois. This memorandum is being prepared for Dr. Kessler's use in formulating his expert opinion on reasonable royalty damages. Dr. Kessler will independently review and refine the Georgia-Pacific factor analysis presented in this memorandum, conduct his own economic analysis of comparable licenses and industry royalty benchmarks, and prepare his expert report in accordance with Federal Rule of Civil Procedure 26(a)(2)(B) and the Court's expert discovery schedule.

Lumos's Designated Damages Expert — Dr. Patricia Moreno:

Lumos has designated Dr. Patricia Moreno as its damages expert. Dr. Moreno is a Certified Public Accountant (CPA) with an Accredited in Business Valuation (ABV) credential and is a partner at Compass Forensic Accounting LLP in Dallas, Texas. Dr. Moreno is anticipated to offer rebuttal opinions on reasonable royalty damages, likely favoring a lower royalty rate or no royalty. Her expert opinions are not yet available, as expert reports have not yet been exchanged pursuant to the Cour

PRIVILEGED AND CONFIDENTIAL — ATTORNEY WORK PRODUCT
Draft Memorandum: Reasonable Royalty Analysis — Georgia-Pacific Factor Framework
January 2023
Right-click to update Table of Contents
I. Purpose and Scope
This memorandum has been prepared to develop and document the fifteen-factor analytical framework established in Georgia-Pacific Corp. v. United States Plywood Corp., 318 F. Supp. 1116 (S.D.N.Y. 1970), for use by Dr. Lawrence Kessler of Apex Claim Analytics LLC (Chicago, Illinois) in preparing his reasonable royalty damages expert report and testimony in this matter. Dr. Kessler has been retained by Optiwave Semiconductor Corporation ("OWS" or "Optiwave") as its designated damages expert. This memorandum is intended to provide Dr. Kessler with a structured preliminary analysis of each Georgia-Pacific factor, marshaling the factual record developed to date and identifying areas requiring further investigation as discovery proceeds.
The scope of this memorandum encompasses an analysis of each of the fifteen Georgia-Pacific factors in the context of the four patents asserted by OWS in this action — U.S. Patent Nos. 10,234,567 (the "'567 Patent"), 10,891,234 (the "'234 Patent"), 11,045,678 (the "'678 Patent"), and 11,312,890 (the "'890 Patent") — and the two accused products manufactured and sold by Defendant Lumos Photonics, Inc. ("Lumos" or "Lumos Photonics"): the LumosBridge™ 800G Optical Engine and the LumosEdge™ 400ZR+ Pluggable Transceiver. The analysis is structured around the hypothetical negotiation between OWS (as willing licensor) and Lumos (as willing licensee) that the Georgia-Pacific framework requires.
The applicable legal standard for this analysis is rooted in 35 U.S.C. § 284, which provides that "[u]pon finding for the claimant the court shall award the claimant damages adequate to compensate for the infringement, but in no event less than a reasonable royalty for the use made of the invention by the infringer." The reasonable royalty represents the floor for patent damages, and is determined through the construct of a hypothetical negotiation between the patentee and the infringer at the time infringement began. The seminal framework for this analysis was set forth in Georgia-Pacific Corp. v. United States Plywood Corp., 318 F. Supp. 1116 (S.D.N.Y. 1970), which identified fifteen factors bearing on the determination of a reasonable royalty. That framework has been adopted and applied by the Federal Circuit in numerous subsequent decisions and remains the standard analytical tool for reasonable royalty determinations.
For purposes of this memorandum, the hypothetical negotiation date is assumed to be immediately before the earliest date of infringement. For the '567 Patent, the '234 Patent, and the '678 Patent, the hypothetical negotiation date is set at Q3 2021, immediately before the commencement of production of the LumosEdge 400ZR+ Pluggable Transceiver. For the '890 Patent, which was not issued until April 26, 2022, the hypothetical negotiation date is set at Q1 2022, immediately before the first shipment of the LumosBridge 800G Optical Engine, as infringement could not predate issuance of the patent. The implications of these different negotiation dates are discussed further in Section III.B below.
Important Note Regarding Status: This memorandum was prepared in January 2023. The analysis was paused and has not been finalized following the Court's Markman Order issued February 17, 2023 (Dkt. 87), which substantially narrowed the scope of the asserted claims by adopting Defendant-aligned constructions for eight of eleven disputed claim terms. The implications of the Markman Order for this damages analysis are addressed in Section X.
II. Summary of Key Financial Data
The following summary sets forth the principal financial data points relevant to the reasonable royalty analysis. These figures are drawn from publicly available financial disclosures, discovery responses produced to date, and pre-litigation valuations commissioned by OWS. This summary is intended to provide a convenient reference for the factor-by-factor analysis that follows.
Optiwave Semiconductor Corporation (OWS):
• FY 2022 Revenue: $187.4 million
• Headquarters: San Jose, California
• Approximate employees: 620
• Majority ownership: Vantage Point Capital Partners Fund IV, LP (approximately 62% equity interest)
• Patent portfolio: 87 issued U.S. utility patents; 14 pending applications
• Key products: OWS-400ZR and OWS-800LR coherent optical transceiver chipsets
• Fabrication: fabless model; chips fabricated by Pinnacle Semiconductor Foundry Co., Ltd. (Hsinchu, Taiwan) on a 45nm silicon-on-insulator (SOI) process
• Pre-litigation patent portfolio valuation (Clearpath IP Valuation Services, Inc.; lead analyst Jennifer Armitage, CFA, CLP): $215 million (full portfolio); $42.8 million (four asserted patents collectively)
• OWS's preliminary estimated reasonable royalty rate: 7.5%
Lumos Photonics, Inc. (Lumos):
• FY 2022 Revenue: $312.8 million
• Headquarters: Austin, Texas
• Fabrication facility: Hillsboro, Oregon
• Approximate employees: 1,450
• Publicly traded on Nasdaq under ticker symbol LMPH since 2021 IPO
• Market capitalization as of complaint filing (June 3, 2022): approximately $1.87 billion
• Market capitalization as of date of this memorandum (January 2023): in excess of $2.0 billion
• Largest institutional shareholder: Ridgeline Asset Management, LLC (approximately 9.2% stake)
• Operating model: integrated device manufacturer (IDM) of silicon photonics components
Accused Product Financial Data:
Preliminary Damages Calculations:
• Base annual royalty: $148.6 million × 7.5% = $11.145 million (applied to full CY 2022 accused product revenue)
• Pre-suit damages (Q1–Q2 2022): $62.3 million × 7.5% = $4.6725 million
• Total claimed damages per Complaint: in excess of $75 million, inclusive of pre-suit damages, post-suit damages accruing through trial, and potential willfulness enhancement (up to treble damages under 35 U.S.C. § 284)
These financial data points are referenced throughout the factor analysis that follows and will be refined as additional discovery is completed, particularly with respect to Lumos's product-level cost, margin, and profit data, which are expected to be produced in response to OWS's outstanding discovery requests.
III. Hypothetical Negotiation Framework
A. Legal Standard
The reasonable royalty determination under 35 U.S.C. § 284 is anchored in the construct of a hypothetical negotiation between a willing licensor and a willing licensee, negotiating at arm's length at the time infringement began. This is not an inquiry into what the parties actually would have agreed to — it is a legal construct designed to approximate the royalty that would have emerged from a voluntary negotiation conducted under certain assumed conditions. See Georgia-Pacific Corp. v. United States Plywood Corp., 318 F. Supp. 1116, 1120 (S.D.N.Y. 1970); Lucent Techs., Inc. v. Gateway, Inc., 580 F.3d 1301, 1324–25 (Fed. Cir. 2009).
The hypothetical negotiation proceeds under the foundational assumption that both parties are aware that the patents at issue are valid and infringed. See Lucent, 580 F.3d at 1325. This assumption eliminates the parties' actual uncertainty about patent validity and infringement scope, which in a real-world negotiation would tend to depress the royalty rate. The construct thus ensures that the reasonable royalty reflects the full value of the patented technology, rather than a rate discounted for litigation risk.
Both parties are further assumed to be willing — the licensor willing to grant a license, and the licensee willing to take one — and the negotiation is conducted at arm's length, free from any duress, compulsion, or other factors that might distort the bargain. The reasonable royalty is the rate that would have been agreed upon, not the rate that one party might have wished to impose unilaterally.
Critical to this analysis, the Federal Circuit has emphasized that the hypothetical negotiation must be grounded in the facts of the case and supported by sound economic reasoning. Speculative or conclusory damages theories are insufficient. Uniloc USA, Inc. v. Kestrel Partners LLC, 632 F.3d 1292, 1315 (Fed. Cir. 2011). In particular, the Federal Circuit has scrutinized the application of the entire market value rule, holding that "the entire market value rule allows for the recovery of damages based on the value of an entire apparatus containing several features, when the feature patented constitutes the basis for customer demand." Id. at 1318 (internal quotation marks omitted). Where the patented feature is not the sole basis for customer demand, apportionment is required — either through the selection of an appropriately narrower royalty base or through a royalty rate that reflects the relative contribution of the patented feature to the overall product value. These principles are discussed further in Section VI below.
B. Assumed Negotiation Date
The hypothetical negotiation date is a critical parameter in the Georgia-Pacific analysis because it defines the universe of information available to the parties and the economic conditions under which the negotiation occurs. For each patent-in-suit, the negotiation date is set immediately before the earliest date of infringement.
'567 Patent (U.S. Patent No. 10,234,567): This patent was filed on March 14, 2017 and issued on March 19, 2019. Because the patent was already issued well before the commencement of Lumos's accused infringing activity, the hypothetical negotiation date is determined by the date of first infringement, which corresponds to the commencement of production of the LumosEdge 400ZR+ in Q3 2021.
'234 Patent (U.S. Patent No. 10,891,234): This patent was filed on September 8, 2017 and issued on January 14, 2021. Similarly, the '234 Patent issued before the commencement of production of the LumosEdge, and the hypothetical negotiation date is set at Q3 2021.
'678 Patent (U.S. Patent No. 11,045,678): This patent was filed on June 22, 2018 and issued on June 29, 2021, shortly before the LumosEdge entered production in Q3 2021. The hypothetical negotiation date for the '678 Patent is likewise set at Q3 2021.
'890 Patent (U.S. Patent No. 11,312,890): This patent was filed on November 15, 2018 and did not issue until April 26, 2022. Because infringement liability under 35 U.S.C. § 271 generally does not attach until after the patent issues, the hypothetical negotiation date for the '890 Patent is set at Q1 2022, immediately before Lumos's first shipment of the LumosBridge 800G. The '890 Patent presents a unique issue: its hypothetical negotiation postdates the negotiation for the other three patents by approximately six months. At the Q1 2022 negotiation date, the parties would have had access to additional information not available at the Q3 2021 date, including the actual market reception of the LumosEdge product, Lumos's Q3–Q4 2021 revenue performance, and the announced launch timeline for the LumosBridge 800G. This additional information may slightly affect the royalty analysis for the '890 Patent relative to the other three patents.
As a practical matter, however, Dr. Kessler may determine that the differences in negotiation dates do not materially alter the overall reasonable royalty determination, particularly if the analysis focuses on a blended royalty rate applicable to all four patents collectively. The question of whether to present a single consolidated royalty rate or separate rates for each patent should be discussed with Dr. Kessler during the expert report preparation process.
C. Parties' Bargaining Positions at Hypothetical Negotiation
Understanding the parties' respective bargaining positions at the hypothetical negotiation date is essential to assessing what royalty rate would have emerged from an arm's-length negotiation.
Optiwave Semiconductor Corporation (OWS) — The Willing Licensor:
At the hypothetical negotiation date, OWS was a fabless semiconductor company headquartered in San Jose, California, with approximately 620 employees and FY 2022 revenue of $187.4 million. OWS's core business is the design and development of silicon photonics chipsets for high-speed optical transceivers used in data center interconnects and telecommunications applications. Its key products — the OWS-400ZR and OWS-800LR coherent optical transceiver chipsets — are manufactured by Pinnacle Semiconductor Foundry Co., Ltd., a leading foundry located in Hsinchu, Taiwan, using a 45nm silicon-on-insulator (SOI) process.
OWS is majority-owned (approximately 62%) by Vantage Point Capital Partners Fund IV, LP, a technology-focused private equity fund with a vintage of 2018 and a fund size of $2.1 billion. Vantage Point's initial equity investment in OWS totaled $74.5 million across Series C and D financing rounds. The Vantage Point investment thesis for OWS includes the company's intellectual property portfolio as a significant component of enterprise value, a point that is directly relevant to OWS's incentives and bargaining posture in any licensing negotiation.
OWS held a substantial patent portfolio consisting of 87 issued U.S. utility patents and 14 pending applications at the time of the hypothetical negotiation. The portfolio was valued by Clearpath IP Valuation Services, Inc. (Boston, Massachusetts; lead analyst Jennifer Armitage, CFA, CLP) at $215 million for the full portfolio and $42.8 million for the four asserted patents collectively. This valuation, while conducted on a pre-litigation basis and subject to the limitations discussed in Section VIII, provides an independent third-party assessment of the economic significance of OWS's patent assets.
As a willing licensor, OWS would have been motivated to extract maximum royalty compensation, recognizing that any license to the patents-in-suit would enable a direct competitor to practice the patented technology and thereby potentially erode OWS's own market share in the highly competitive silicon photonics transceiver market.
Lumos Photonics, Inc. (Lumos) — The Willing Licensee:
At the hypothetical negotiation date, Lumos was an integrated device manufacturer (IDM) of silicon photonics components headquartered in Austin, Texas, with approximately 1,450 employees and FY 2022 revenue of $312.8 million. Unlike OWS, Lumos operates its own fabrication facility in Hillsboro, Oregon, giving it vertical integration capabilities that include control over the manufacturing process and yields of its silicon photonic integrated circuits.
Lumos completed its initial public offering on the Nasdaq exchange (ticker: LMPH) in 2021 and had a market capitalization of approximately $1.87 billion as of the filing of the Complaint on June 3, 2022. Its largest institutional shareholder is Ridgeline Asset Management, LLC, which holds an approximately 9.2% stake. By the date of this memorandum, Lumos's market capitalization has grown to in excess of $2.0 billion, reflecting continued investor confidence in the company's growth trajectory.
Lumos directly competes with OWS in the high-speed optical transceiver market, targeting hyperscale data center customers for its LumosBridge 800G Optical Engine and LumosEdge 400ZR+ Pluggable Transceiver products. The accused products have been commercially successful, generating approximately $148.6 million in combined revenue in CY 2022. As a willing licensee, Lumos would have been motivated to negotiate the lowest possible royalty rate while securing the right to continue manufacturing and selling its accused products.
The competitive dynamic between the parties is a critical factor in the hypothetical negotiation. In a negotiation between direct competitors, the licensor's willingness to grant a license is tempered by the recognition that doing so enables competition in its own market, while the licensee's willingness to pay is informed by the value of avoiding the risk of injunctive relief and the cost of designing around the patented technology.
IV. Georgia-Pacific Factor Analysis
The following sections address each of the fifteen Georgia-Pacific factors as applied to the facts of this case. For each factor, we identify the relevant evidence, assess its directional impact on the reasonable royalty determination (i.e., whether it favors a higher royalty, a lower royalty, or is neutral), and note any areas where additional evidence is needed.
Factor 1: Royalties Received by the Patentee for Licensing the Patents-in-Suit
The first Georgia-Pacific factor considers the royalties received by the patentee for the licensing of the patent in suit, proving or tending to prove an established royalty. An established royalty is the most reliable indicator of a reasonable royalty because it reflects the actual market value placed on the patented technology by arm's-length negotiation.
OWS has not previously licensed the four asserted patents — U.S. Patent Nos. 10,234,567; 10,891,234; 11,045,678; and 11,312,890 — to any third party. No established royalty rate exists for these specific patents. The absence of any licensing history for the patents-in-suit means that there is no direct evidence from which to infer the market rate for the patented technology. This is not unusual in cases involving relatively recently issued patents held by companies that pursue an exclusionary patent strategy, as OWS does.
OWS's broader patent portfolio has been the subject of two cross-license agreements with non-competitor companies. Based on information obtained to date, these cross-licenses involved different technology fields — specifically, modulator driver architectures for telecommunications applications, rather than the data center interconnect technologies at issue in this case. The details of these cross-licenses, including their financial terms and the specific patents covered, are subject to further confirmation with Dr. Kessler upon completion of document review. However, the limited comparability of these cross-licenses — arising from differences in technology field, competitive posture of the counterparties, and license scope — substantially reduces their probative value as indicators of the appropriate royalty rate for the patents-in-suit.
The absence of an established royalty for the patents-in-suit does not preclude the determination of a meaningful reasonable royalty. Rather, it shifts the analytical burden to the remaining Georgia-Pacific factors and to comparable license evidence from the broader market. See Georgia-Pacific, 318 F. Supp. at 1120 (identifying fifteen factors, of which an established royalty is only the first).
Assessment: This factor is neutral to modestly favorable for OWS. The absence of a prior license does not set a ceiling on the royalty; instead, it necessitates reliance on other indicia of value. OWS can argue that its refusal to license reflects the high value of the patents, while Lumos may argue that the absence of any licensing activity suggests the patents have not been independently valued by the market.
Factor 2: Rates Paid by the Licensee for Use of Comparable Patents
The second factor examines the rates paid by the licensee for the use of other patents comparable to the patent in suit. This factor looks to the licensee's own licensing history for insight into what rates the market has established for similar technology.
Lumos's existing patent license agreements are a critical source of comparable royalty data. OWS's First Set of Requests for Production, Nos. 67–78, served on November 21, 2022, sought the production of all license agreements involving silicon photonics technology, optical transceiver technology, and related semiconductor intellectual property. As of the date of this memorandum, Lumos's responses to these requests are pending, and the full scope of Lumos's licensing history has not been established.
Preliminary analysis of publicly available licensing data in the silicon photonics space suggests that royalty rates for component-level patent licenses generally fall in the range of 3% to 10% of net sales, with the rate depending on factors including portfolio breadth, technology criticality, the competitive relationship between licensor and licensee, and the geographic scope of the license. This range is drawn from public filings, analyst reports, and prior litigation outcomes in the semiconductor industry, and is subject to refinement as more specific data becomes available.
Lumos's most recent Form 10-K, filed with the Securities and Exchange Commission, references certain intellectual property licensing costs as a component of its cost of goods sold. However, the specific royalty rates, licensors, and patents covered are not broken out in the public filing and must be obtained through discovery. The 10-K disclosure does confirm that Lumos has entered into IP licensing arrangements, which will provide useful data points once the underlying agreements are produced.
Assessment: This factor is pending completion of discovery. The preliminary market data suggesting a 3%–10% range is broadly consistent with OWS's proposed 7.5% rate, but the analysis cannot be finalized until Lumos's actual license agreements are reviewed and assessed for comparability. Dr. Kessler should be prepared to analyze each of Lumos's licenses for comparability to the hypothetical license at issue in this case.
Factor 3: Nature and Scope of the License
The third factor addresses the nature and scope of the license, including whether it is exclusive or non-exclusive, whether it is restricted or unrestricted in terms of territory or field of use, and who is granted the license.
For purposes of the hypothetical negotiation, the assumed license is a non-exclusive, worldwide license limited to the four patents-in-suit. The license would cover the manufacture, use, sale, offer for sale, and importation of products embodying the claimed inventions — specifically, Lumos's LumosBridge 800G Optical Engine and LumosEdge 400ZR+ Pluggable Transceiver products, along with reasonably foreseeable derivative products incorporating the patented technology.
The non-exclusive nature of the hypothetical license is significant because a non-exclusive license does not prevent OWS from granting additional licenses to other parties or from continuing to practice the patented inventions itself. Non-exclusive licenses generally command lower royalty rates than exclusive licenses, all else being equal, because the licensee does not receive the benefit of exclusivity. However, the practical significance of this distinction is tempered in this case by the fact that OWS has not licensed the patents-in-suit to any other party and has historically maintained an exclusionary posture regarding its core patent portfolio.
The hypothetical license is limited to the four patents-in-suit and to specific products, rather than being a broad portfolio cross-license or a field-of-use license. A narrower license covering a smaller number of patents and limited to specific products would tend to yield a lower total royalty payment than a broad portfolio license, but the per-patent or per-technology royalty rate may actually be higher for a narrower license because the licensee is obtaining rights to specific, targeted technology rather than a portfolio of mixed-value patents.
With respect to royalty structure, a running royalty (calculated as a percentage of net sales on an ongoing basis) is recommended over a lump-sum payment. The running royalty structure is more appropriate in this context because: (i) Lumos's accused products are generating ongoing revenue and are expected to continue doing so throughout the remaining patent life; (ii) a running royalty aligns the royalty payment with the licensee's actual use of the patented technology and avoids the need for speculative estimates of future sales volumes; and (iii) running royalties are the more commonly applied structure in the silicon photonics and semiconductor industries for component-level patent licenses. Dr. Kessler should address the running royalty versus lump-sum question in his expert report and be prepared to testify on the economic basis for the recommended structure.
Assessment: This factor modestly favors a lower royalty rate relative to what would be appropriate for a broader or exclusive license, but the narrow scope of the license is offset by the targeted, high-value nature of the patented technology covered. The recommended running royalty structure is consistent with industry practice and provides for fair compensation proportional to Lumos's actual sales.
Factor 4: Licensor's Established Policy of Maintaining Patent Monopoly
The fourth factor examines the licensor's established policy and marketing program to maintain its patent monopoly by not licensing others to use the invention, or by granting licenses under special conditions designed to preserve that monopoly. This factor recognizes that a licensor who has consistently refused to license its patents would require a higher royalty to compensate for the departure from its exclusionary strategy.
OWS has historically maintained a strong exclusionary posture regarding its core silicon photonics patent portfolio. As noted under Factor 1, OWS has not granted any licenses for the four patents-in-suit. The two cross-license agreements involving OWS's broader portfolio covered different technology fields (modulator driver architectures for telecommunications) and involved non-competitor counterparties, and thus do not represent a departure from OWS's general policy of maintaining exclusivity over its core data center interconnect patents.
Internal communications produced in discovery reinforce OWS's exclusionary licensing policy. Dr. Priya Ramanathan, OWS's Chief Executive Officer, has emphasized in multiple internal memoranda and board presentations that the company's patent portfolio is a "strategic competitive asset" that provides OWS with a critical differentiator in the silicon photonics market. Ted Marchetti, OWS's Vice President of intellectual Property, has echoed this position in correspondence with external counsel and in internal IP strategy documents, consistently advocating against licensing OWS's core patents to competitors.
The investment thesis of Vantage Point Capital Partners Fund IV, LP — OWS's majority shareholder — further supports the importance of the patent portfolio to OWS's enterprise value. Vantage Point's initial equity investment of $74.5 million across Series C and Series D financing rounds was premised in part on the strength and defensibility of OWS's intellectual property position. The Clearpath IP Valuation Services pre-litigation valuation of $215 million for the full patent portfolio — representing a significant portion of OWS's total enterprise value — confirms that the patent portfolio is viewed by OWS and its investors as a foundational asset, not a commodity to be freely licensed.
Assessment: This factor weighs strongly in favor of a higher royalty rate. OWS's consistent refusal to license its core patents, combined with the strategic importance of the patent portfolio to OWS's competitive position and enterprise value, indicates that OWS would demand substantial compensation for any departure from its exclusionary policy. A willing licensor in OWS's position would insist on a premium royalty rate that reflects the competitive harm of enabling a direct competitor's products.
Factor 5: Commercial Relationship Between the Licensor and Licensee
The fifth factor considers the commercial relationship between the licensor and licensee — specifically, whether they are competitors in the same territory and in the same line of business, or whether they are inventor and promoter.
OWS and Lumos are direct competitors in the silicon photonics market for high-speed data center interconnects and telecommunications applications. Both companies design and sell optical transceiver chipsets and engines targeting the hyperscale data center market, and both compete for design wins with the same universe of major customers, including hyperscale cloud service providers and data center switch manufacturers.
OWS's OWS-400ZR coherent optical transceiver chipset competes directly with Lumos's LumosEdge 400ZR+ Pluggable Transceiver in the 400G coherent optical interconnect segment. Similarly, OWS's OWS-800LR platform competes with Lumos's LumosBridge 800G Optical Engine in the emerging 800G co-packaged optics segment. The products occupy overlapping market positions and are evaluated by customers as substitutes or alternatives in competitive procurement processes.
The competitive relationship between OWS and Lumos is direct and intense. Every unit of LumosEdge or LumosBridge sold by Lumos represents a potential lost sale of OWS-400ZR or OWS-800LR by OWS. The silicon photonics transceiver market, while rapidly growing, is characterized by a limited number of design wins with major hyperscale customers, and the loss of a design win to a competitor can result in years of lost revenue for the displaced supplier. The competitive dynamic is further intensified by the high switching costs faced by data center operators once a particular transceiver platform is deployed at scale.
In the patent licensing context, it is well established that a license to a direct competitor typically commands a higher royalty rate than a license to a non-competitor. See, e.g., Georgia-Pacific, 318 F. Supp. at 1120 (identifying the "commercial relationship between the licensor and licensee" as a factor bearing on the reasonable royalty). The rationale is straightforward: a license to a competitor not only provides the licensee with access to the patented technology but also enables the licensee to compete more effectively against the licensor, thereby inflicting competitive harm that must be compensated through the royalty.
Assessment: This factor weighs strongly in favor of a higher royalty rate. The direct competitive relationship between OWS and Lumos means that any license from OWS to Lumos would enable Lumos to practice the patented technology in products that directly compete with OWS's own products. A willing licensor in OWS's position would demand a substantial premium to compensate for the competitive injury resulting from the license.
Factor 6: Effect of Selling the Patented Specialty in Promoting Sales of Other Products
The sixth factor examines the effect of selling the patented specialty in promoting sales of the licensor's other products, the existing value of the invention to the licensor as a generator of sales of its non-patented items, and the extent of such derivative or convoyed sales.
Lumos's accused products — the LumosBridge 800G Optical Engine and the LumosEdge 400ZR+ Pluggable Transceiver — are standalone revenue-generating products, not accessories or components sold primarily to drive sales of other Lumos product lines. Each product is sold as an independent commercial offering with its own pricing, specification sheets, and customer base. There is no evidence that the accused products serve primarily as loss leaders or promotional items designed to generate convoyed sales of unrelated Lumos products.
However, the data center optical interconnect market exhibits significant platform-stickiness and lock-in effects that are relevant to this factor. Adoption of Lumos's transceiver products by hyperscale data center operators for a given deployment generation creates strong incentives for those operators to continue using Lumos products in subsequent deployment cycles. This platform-stickiness arises from several sources: (i) the cost and complexity of requalifying alternative transceiver suppliers; (ii) the interoperability testing required to validate new transceiver platforms against existing switch and network infrastructure; and (iii) the preference of data center operators for vendor continuity to simplify supply chain management and reduce operational risk.
As a result, every design win that Lumos secures with its current-generation products has an outsized impact on Lumos's competitive position for future product generations. This downstream competitive effect is not captured by the current-period revenue of the accused products alone. From OWS's perspective as a willing licensor, licensing the patented technology to Lumos not only enables Lumos's current products but also strengthens Lumos's competitive position for future generations of transceiver products, compounding the competitive harm to OWS over time.
OWS's own products function similarly as standalone chipsets and engines, and OWS benefits from the same platform-stickiness dynamics when it secures design wins with hyperscale customers. The competitive symmetry between the parties on this factor means that both OWS and Lumos would recognize the long-term competitive implications of the license during the hypothetical negotiation.
Assessment: This factor weighs modestly in favor of a higher royalty rate. While the accused products are not promotional items driving convoyed sales of other products, the platform-stickiness and design-win momentum effects in the data center transceiver market create downstream competitive consequences that a willing licensor would seek to address through the royalty rate.
Factor 7: Duration of the Patent and the Term of the License
The seventh factor considers the duration of the patent and the term of the license. Longer remaining patent life provides the licensee with a longer period of protection against competition from the patented technology, which generally supports a higher royalty rate.
The four asserted patents have the following relevant dates and estimated remaining life:
All four patents have substantial remaining life — approximately 14 to 17 years from the hypothetical negotiation date(s). This extended remaining life means that the patented technology will be enforceable throughout the expected commercial lifetime of the accused products and through multiple future product generations. In a rapidly evolving technology market such as silicon photonics, a 14-to-17-year patent term encompasses multiple product generation cycles and extends well beyond the current market horizon.
The hypothetical license term would most likely be coextensive with the remaining patent terms, as is customary for patent licenses in the semiconductor industry. A license running for the full remaining patent term provides the licensee with certainty of rights and avoids the need for periodic renegotiation, while ensuring that the licensor receives compensation for the full duration of the patent monopoly.
Assessment: This factor weighs in favor of a higher royalty rate. The substantial remaining life of all four patents means that the licensee would benefit from the license for an extended period, justifying a meaningful royalty that reflects the long duration of the patented technology's commercial relevance. The long remaining life also increases the total present value of the royalty stream, which is a relevant consideration in calibrating the appropriate royalty rate.
Factor 8: Established Profitability of the Product Made Under the Patent; Its Commercial Success; and Its Current Popularity
The eighth factor examines the established profitability of the product made under the patent, its commercial success, and its current popularity. Commercially successful products practicing the patented technology are evidence that the patented features contribute meaningful value, supporting a higher royalty rate.
Lumos's accused products have demonstrated strong commercial success. The LumosBridge 800G and LumosEdge 400ZR+ generated combined revenue of approximately $148.6 million in CY 2022, representing approximately 47.5% of Lumos's total FY 2022 revenue of $312.8 million. The accused products are thus a core driver of Lumos's business, not peripheral or marginal product lines.
The LumosBridge 800G Optical Engine first shipped in Q1 2022 and is priced at $1,200 to $1,800 per unit depending on configuration. The LumosBridge is positioned as a next-generation co-packaged optics platform designed for integration with hyperscale data center switch ASICs, targeting the emerging 800G and beyond interconnect segment. The LumosBridge has attracted significant attention from hyperscale customers and industry analysts, reflecting the market's strong demand for co-packaged optics solutions capable of supporting the bandwidth requirements of AI/ML workloads.
The LumosEdge 400ZR+ has been in production since Q3 2021 and is priced at $600 to $900 per unit depending on configuration. The LumosEdge is a QSFP-DD form factor pluggable transceiver designed for metro and data center interconnect (DCI) applications. It has achieved broad market adoption and competes directly with OWS's OWS-400ZR product for 400G coherent optical interconnect design wins.
Lumos's overall revenue growth trajectory and its successful Nasdaq IPO in 2021 further indicate strong market demand and investor confidence in the company's products and technology platform. The market capitalization growth from approximately $1.87 billion at the time of the Complaint to over $2.0 billion as of the date of this memorandum reinforces the commercial attractiveness and growth prospects of Lumos's silicon photonics product portfolio, of which the accused products are the centerpiece.
For comparison, OWS's own products employing the patented technology — the OWS-400ZR and OWS-800LR — have also achieved commercial traction, with OWS generating $187.4 million in total FY 2022 revenue. While OWS's product-level revenue breakdown has not been disclosed publicly, the competitive performance of OWS's products in the same market segments further corroborates the commercial significance of the patented technology.
Assessment: This factor weighs in favor of a moderate-to-higher royalty rate. The strong commercial performance of the accused products, both in absolute revenue terms and as a proportion of Lumos's total revenue, indicates that the patented technology is commercially significant and contributes meaningfully to the products' market success. The profitability of the accused products (to be confirmed through discovery of Lumos's cost and margin data) will further inform this assessment.
Factor 9: Utility and Advantages of the Patent Property Over Old Modes or Devices
The ninth factor examines the utility and advantages of the patent property over the old modes or devices, if any, that had been used for working out similar results. This factor assesses the technical superiority of the patented inventions over the prior art and the availability of alternatives.
Each of the four asserted patents addresses a distinct performance-critical aspect of silicon photonics transceiver design:
'567 Patent — Integrated Waveguide Coupler with Tapered Mode Converter: The '567 Patent claims an integrated waveguide coupler incorporating an adiabatic taper design for fiber-to-chip optical coupling. The claimed adiabatic taper mode converter provides significantly lower optical coupling losses compared to prior art edge couplers, which relied on abrupt transitions between fiber and waveguide modes. The adiabatic taper design enables efficient, low-loss coupling between standard single-mode fiber and sub-micron silicon waveguides, which is critical for high-speed optical transceivers where every decibel of coupling loss directly impacts link budget and system performance.
'234 Patent — Ge-on-Si Photodetector with Graded Buffer: The '234 Patent claims a germanium-on-silicon photodetector employing a continuously graded buffer layer between the silicon substrate and the germanium absorber. The continuously graded buffer reduces threading dislocation density at the Ge/Si interface compared to prior art approaches using abrupt or step-graded interfaces, thereby enhancing photodetector responsivity and bandwidth. This is particularly important for 50+ Gbaud photodetection, where defect-related dark current and bandwidth limitations are significant performance constraints.
'678 Patent — Multi-Channel WDM on Silicon: The '678 Patent claims a monolithically integrated wavelength division multiplexing (WDM) system on a silicon photonic chip, eliminating the need for external discrete WDM multiplexing and demultiplexing components. Monolithic integration reduces package size, lowers manufacturing cost, and eliminates the optical losses and alignment challenges associated with discrete WDM components. This is particularly advantageous for multi-channel transceiver modules where aggregate bandwidth is achieved through wavelength parallelism.
'890 Patent — MZM Driver Co-Integration: The '890 Patent claims a Mach-Zehnder modulator (MZM) driver architecture in which the pre-emphasis equalization stage is co-integrated with the optical modulator on a common substrate. Co-integration reduces parasitic capacitance at the interface between the driver and modulator, improving high-frequency modulation performance. This is critical for achieving the high baud rates required for 400G and 800G optical signaling, where parasitic capacitance at the driver-modulator interface is a dominant bandwidth limiter.
Each of these patented inventions offers measurable performance advantages over prior art approaches. However, it must be acknowledged that design-around alternatives exist for each patent. For the '567 Patent, alternative coupling approaches include segmented tapers, grating couplers, and inverse taper designs that achieve mode conversion through non-adiabatic transitions. For the '234 Patent, step-graded germanium buffer layers and selective epitaxial growth techniques are known alternatives, albeit with performance trade-offs. For the '678 Patent, hybrid integration approaches using external arrayed waveguide gratings (AWGs) or thin-film filters can achieve WDM functionality without monolithic integration. For the '890 Patent, flip-chip bonded driver ICs and wire-bonded driver-modulator assemblies represent alternative approaches, though with higher parasitic capacitance.
The availability of these alternatives is relevant because it affects the licensee's outside option in the hypothetical negotiation — if Lumos could readily design around the asserted patents, it would face less pressure to accept a license and would be willing to pay a lower royalty. Conversely, if the design-around alternatives impose significant performance penalties, increased manufacturing cost, or longer development timelines, the patented technology's value to the licensee is commensurately higher.
Assessment: This factor weighs in favor of a moderate royalty. The patented inventions offer genuine and measurable technical advantages over prior art approaches, but the availability of design-around alternatives for each patent prevents this factor from commanding the full weight it would carry if the patented technology were truly indispensable. Dr. Whitfield's technical expert testimony will be critical in quantifying the performance advantages and the cost and feasibility of alternatives.
Factor 10: Nature of the Patented Invention, Character of Commercial Embodiment, and Benefits to Users
The tenth factor considers the nature of the patented invention, the character of the commercial embodiment of it as owned and produced by the licensor, and the benefits to those who have used the invention.
The patented inventions relate to fundamental building blocks of silicon photonic integrated circuits (PICs) used in high-speed data center optical interconnects. These building blocks — fiber-to-chip coupling, photodetection, wavelength division multiplexing, and modulator drive — are core functional elements that are present in essentially all modern silicon photonic transceiver designs. The inventions are not directed to peripheral or ornamental features; they address the physical and electrical interfaces that determine the transceiver's optical performance, bandwidth, and power efficiency.
The commercial embodiment of the inventions is demonstrated by OWS's chipsets manufactured by Pinnacle Semiconductor Foundry Co., Ltd. on a 45nm SOI process. The OWS-400ZR and OWS-800LR products incorporate the patented technologies and have been successfully deployed in production volumes, demonstrating the practical applicability and manufacturability of the inventions at commercial scale. The fact that the patented inventions have been reduced to production-ready silicon validates their commercial utility and distinguishes them from theoretical or laboratory-stage concepts.
End-user benefits derived from the patented technologies are substantial and directly aligned with the key performance metrics that drive purchasing decisions among hyperscale data center operators. These benefits include:
• Higher data rates: The patented technologies collectively enable 400G and 800G per-lane signaling, supporting the aggregate bandwidth requirements of modern hyperscale data center architectures driven by AI/ML workload expansion.
• Lower power consumption: Monolithic integration of WDM components ('678 Patent) and co-integrated driver architectures ('890 Patent) reduce power consumption per bit, which is a critical metric for data centers where power and cooling represent a significant proportion of total operating cost.
• Reduced form factor: Integration of previously discrete components onto a single silicon photonic chip reduces the physical size of the transceiver module, enabling higher port density in switch and router line cards.
• Lower cost per bit: The combination of monolithic integration, reduced packaging complexity, and higher data rates drives down the cost per bit of transmitted data, which is the ultimate economic metric for data center operators.
The data center optical interconnect market is experiencing rapid growth driven by the expansion of hyperscale data centers, the proliferation of AI/ML training and inference workloads, and the transition from 100G to 400G and 800G interconnect speeds. This market growth amplifies the commercial significance of the patented technologies, as they address performance-critical bottlenecks in the transceiver value chain.
Assessment: This factor supports a moderate royalty reflecting the importance of the patented features to the overall product value proposition. The patented technologies are core functional elements of the transceiver products, and the end-user benefits are directly aligned with the most commercially significant performance metrics. The factor does not by itself indicate a specific royalty rate but reinforces the conclusion that the patented features contribute meaningful value to the accused products.
Factor 11: Extent to Which the Infringer Has Made Use of the Invention and Evidence of the Value of That Use
The eleventh factor examines the extent to which the infringer has made use of the invention and any evidence probative of the value of that use. This factor looks at both the volume of the infringer's use and the centrality of the patented technology to the infringer's products and business.
Lumos's accused products — the LumosBridge 800G and LumosEdge 400ZR+ — generated combined revenue of approximately $148.6 million in CY 2022. This represents substantial commercial use of the products alleged to embody the patented inventions. The volume of Lumos's sales demonstrates that the accused products have achieved broad market adoption and are a material component of Lumos's overall business.
The centrality of the patented features to the accused products is a critical sub-issue under this factor. Based on the preliminary infringement analysis, the four patented features — the waveguide coupler with adiabatic taper ('567 Patent), the Ge-on-Si photodetector with graded buffer ('234 Patent), the monolithically integrated WDM system ('678 Patent), and the co-integrated MZM driver ('890 Patent) — are integrated into the core architecture of both the LumosBridge and LumosEdge products. These features are not peripheral add-ons or optional configurations; they are fundamental components of the silicon photonic integrated circuit at the heart of each accused product.
However, the accused products incorporate numerous other innovations and components that are not covered by the asserted patents. Lumos's independent contributions include, among others: (i) Lumos's proprietary digital signal processing (DSP) architecture, which performs forward error correction, equalization, and signal processing functions critical to transceiver performance; (ii) advanced packaging technology, including micro-transfer-printed III-V gain elements; (iii) thermal management solutions necessary for reliable operation in high-density data center environments; and (iv) firmware, software, and management interfaces required for integration into customer network architectures. These non-patented features collectively represent a substantial portion of the value of the accused products.
The apportionment analysis required by this factor — determining what portion of the accused products' value is attributable to the patented features versus other technology — is a complex undertaking that will require both technical and economic expert input. Preliminary estimates, subject to refinement by Dr. Kessler based on detailed technical and economic analysis, suggest that the four patented features collectively account for approximately 15% to 25% of the total value of the accused transceiver platforms. This estimate reflects the importance of the patented building blocks to the transceiver's optical performance while recognizing the substantial value contributed by non-patented technology.
Assessment: This factor weighs in favor of a moderate royalty reflecting significant but not total contribution of the patented technology to the accused products' commercial value. The extensive use of the accused products ($148.6 million in CY 2022 revenue) demonstrates the commercial significance of the technology, but apportionment is required to ensure that the royalty reflects only the value attributable to the patented features.
Factor 12: Portion of the Profit or Selling Price Customarily Allowed for Use of the Invention or Analogous Inventions
The twelfth factor considers the portion of the profit or of the selling price that is customary in the particular business or in comparable businesses to allow for the use of the invention or analogous inventions.
Industry analysis of silicon photonics component licensing reveals the following royalty rate benchmarks:
• Individual patent licenses: Royalty rates typically range from 2% to 8% of net sales for licenses covering individual patents or small groups of closely related patents in the silicon photonics and optical transceiver space.
• Portfolio licenses: For broader portfolio licenses covering core platform technology, royalty rates typically range from 5% to 12% of net sales, reflecting the broader scope and higher aggregate value of multi-patent portfolios.
These ranges are drawn from publicly available data, including SEC filings, analyst reports, and prior litigation outcomes in the semiconductor industry. Dr. Kessler will identify and analyze specific comparable royalty benchmarks in his expert report, drawing on both publicly available data and the comparable license evidence produced through discovery.
It is important to note the constraints imposed by the entire market value rule on the selection of the royalty base. Under Uniloc USA, Inc. v. Kestrel Partners LLC, 632 F.3d 1292 (Fed. Cir. 2011), and subsequent Federal Circuit authority, if the royalty base is the entire product price, the patented feature must be the basis for customer demand for the entire product; otherwise, the royalty must be apportioned to reflect only the value of the patented feature. In this case, the question of the appropriate royalty base is somewhat simplified by the fact that the smallest saleable patent-practicing unit is the integrated photonic chip or engine itself, which is effectively the accused product (the LumosBridge engine or LumosEdge transceiver module). The accused products are not components of larger systems for which the entire market value rule would require the identification of a smaller unit. Accordingly, the accused product revenue of $148.6 million in CY 2022 appears to be an appropriate royalty base, provided that the royalty rate is adjusted to reflect the contribution of the patented features to the overall product value.
Assessment: This factor supports a royalty rate in the range of 5% to 8% when applied to accused product revenue, consistent with industry practice for patent licenses covering core silicon photonics technology. The 7.5% preliminary rate falls within this range and is supported by the market data identified to date. This assessment will be refined by Dr. Kessler based on the comparable license evidence produced through discovery.
Factor 13: Portion of Realizable Profit Attributable to the Invention vs. Non-Patented Elements, Manufacturing Process, Business Risks, and Improvements Added by the Infringer
The thirteenth factor examines the portion of the realizable profit that should be credited to the invention as distinguished from non-patented elements, the manufacturing process, business risks, or significant features or improvements added by the infringer.
Lumos has made substantial independent contributions to the commercial value of its accused products. These contributions include:
Research and Development: Lumos has invested significantly in its own R&D capabilities. Lumos's Chief Technology Officer, Dr. Mei-Lin Chen, is a recognized expert in silicon photonics who holds 14 patents in her own name. Dr. Nikolai Petrov, Lumos's Vice President of Engineering, leads a substantial photonics engineering team that has developed proprietary innovations across the transceiver design stack, from photonic circuit design to packaging and test.
Vertical Integration and Manufacturing: Unlike OWS, which operates as a fabless company relying on Pinnacle Semiconductor Foundry for chip fabrication, Lumos operates its own fabrication facility in Hillsboro, Oregon. This vertical integration gives Lumos significant advantages in process development, yield optimization, and manufacturing cost control that are not derived from the asserted patents. Lumos's investment in its Hillsboro fab represents a substantial capital commitment and a source of competitive differentiation that is independent of the patented technology.
Proprietary Technology: Lumos's LumosBridge platform incorporates several proprietary innovations that are not covered by the asserted patents, including: (i) micro-transfer-printed III-V gain elements, which enable the integration of indium phosphide laser sources onto the silicon photonic platform; (ii) proprietary DSP algorithms for coherent optical signal processing; and (iii) advanced thermal management solutions, including micro-channel liquid cooling for the LumosBridge 800G co-packaged optics module.
Business Risk: Lumos has assumed significant business risk in bringing the LumosBridge 800G to market as a first-generation co-packaged optics product. The co-packaged optics market segment is nascent and involves substantial technical risk (integration with switch ASICs, thermal management challenges) and commercial risk (uncertainty about hyperscale customer adoption timelines). Lumos's willingness to invest in and commercialize co-packaged optics products represents an independent source of value creation that should be credited to Lumos in the royalty analysis.
These independent contributions by Lumos are significant and must be properly accounted for in the apportionment analysis. A royalty rate that failed to credit Lumos for its own innovations, manufacturing capabilities, and business risks would overcompensate OWS at Lumos's expense and would not reflect the outcome of an arm's-length negotiation.
Estimated profit split: Based on the preliminary analysis, approximately 60% to 70% of the accused products' realizable profit is attributable to Lumos's non-patented contributions (including its proprietary technology, manufacturing capabilities, and business risk-taking), with approximately 30% to 40% attributable to the patented technology and favorable market conditions. This profit split is preliminary and will be refined by Dr. Kessler based on detailed economic analysis of the accused products' cost structure, profit margins, and the incremental value of the patented features.
Assessment: This factor favors a lower royalty rate, recognizing Lumos's substantial independent contributions to the commercial value of the accused products. The factor serves as an important counterweight to Factors 4, 5, 8, and 9, which collectively support a higher royalty. The net effect is to moderate the royalty rate to a level that appropriately compensates OWS for the patented technology while giving due credit to Lumos's independent innovations.
Factor 14: Opinion Testimony of Qualified Experts
The fourteenth factor considers the opinion testimony of qualified experts on the relevant issues.
OWS's Technical Expert — Dr. Harold Whitfield:
OWS has retained Dr. Harold Whitfield as its technical expert in this litigation. Dr. Whitfield holds a Ph.D. in Electrical Engineering from the University of California, Berkeley, and is Professor Emeritus at the University of California, Santa Barbara. He has authored over 200 publications in the field of integrated photonics and is a Fellow of the IEEE and the Optical Society of America. Dr. Whitfield operates through his consulting firm, Whitfield Technical Consulting Group, Inc., based in Palo Alto, California.
Dr. Whitfield's declaration, filed in December 2022 in support of OWS's claim construction brief, opines on the technical significance of the patented innovations and their role in the accused products. Dr. Whitfield's opinions address the advantages of the patented approaches over prior art alternatives, the difficulty and cost of designing around the asserted claims, and the centrality of the patented features to the performance characteristics of the accused products. His technical opinions will provide critical input to the apportionment analysis under Factors 11 and 13 and will inform Dr. Kessler's economic analysis of the value attributable to the patented technology.
OWS's Damages Expert — Dr. Lawrence Kessler:
OWS has retained Dr. Lawrence Kessler as its damages expert. Dr. Kessler holds a Ph.D. in Economics from Northwestern University and is a principal at Apex Claim Analytics LLC in Chicago, Illinois. This memorandum is being prepared for Dr. Kessler's use in formulating his expert opinion on reasonable royalty damages. Dr. Kessler will independently review and refine the Georgia-Pacific factor analysis presented in this memorandum, conduct his own economic analysis of comparable licenses and industry royalty benchmarks, and prepare his expert report in accordance with Federal Rule of Civil Procedure 26(a)(2)(B) and the Court's expert discovery schedule.
Lumos's Designated Damages Expert — Dr. Patricia Moreno:
Lumos has designated Dr. Patricia Moreno as its damages expert. Dr. Moreno is a Certified Public Accountant (CPA) with an Accredited in Business Valuation (ABV) credential and is a partner at Compass Forensic Accounting LLP in Dallas, Texas. Dr. Moreno is anticipated to offer rebuttal opinions on reasonable royalty damages, likely favoring a lower royalty rate or no royalty. Her expert opinions are not yet available, as expert reports have not yet been exchanged pursuant to the Cour

PRIVILEGED AND CONFIDENTIAL — ATTORNEY WORK PRODUCT

Record

Parties

Dates

Amounts