CALDERWOOD & HARKNESS LLP INTERNAL MEMORANDUM — PRIVILEGED AND CONFIDENTIAL ATTORNEY WORK PRODUCT
Date: Anticipated July 2025 (to be finalized upon receipt of Dr. Volkov's rebuttal expert report, due July 15, 2025)
TO: Alan Ngo, Responsible Partner; Wendy Schultz, Lead Associate
FROM: Omar Siddiqui, Associate
RE: Outline of Dr. Lawrence Keating's Anticipated Rebuttal to Dr. Natasha Volkov's Damages Opinions
MATTER: Optiwave Semiconductor Corp. v. Lumicore Technologies, Inc., Case No. 2:24-cv-00319-JRG (E.D. Tex.)
CLIENT: Optiwave Semiconductor Corp. ("OWS")
OPPOSING PARTY: Lumicore Technologies, Inc. ("Lumicore")
OPPOSING COUNSEL: Hawthorne Vega & Cruz LLP, 600 Congress Avenue, Suite 2400, Austin, TX 78701
STATUS: Draft — Subject to revision upon receipt of Dr. Volkov's rebuttal expert report
CONFIDENTIAL TREATMENT REQUESTED — This document is attorney work product prepared by Calderwood & Harkness LLP in anticipation of litigation. This memorandum is protected by the attorney-client privilege and the work-product doctrine. It is prepared solely for internal use in connection with the firm's representation of Optiwave Semiconductor Corp. Distribution to any third party is strictly prohibited without the express written authorization of Alan Ngo, Responsible Partner.
Right-click to update Table of Contents
I. PURPOSE AND SCOPE OF MEMORANDUM
This internal memorandum outlines the anticipated scope and structure of Dr. Lawrence Keating's rebuttal expert report, to be prepared in response to Dr. Natasha Volkov's rebuttal damages report. Dr. Volkov, a Managing Director at Pinnacle Analytics Partners LLP (Houston, Texas), serves as Lumicore Technologies, Inc.'s retained damages expert in this matter. Her rebuttal report is expected to be served by Lumicore on or before July 15, 2025, in accordance with the Court's Scheduling Order entered on August 12, 2024 (D.E. 22).
Dr. Keating is a Managing Director of Meridian Economic Consulting Group, LLC, with offices at 233 South Wacker Drive, Suite 4400, Chicago, Illinois 60606. He has been retained by OWS as its damages expert in this matter at a rate of $875 per hour, with an estimated total budget for his engagement (opening report, rebuttal report, deposition preparation, deposition testimony, and trial testimony) of $950,000. Dr. Keating served his opening expert report on June 15, 2025, in which he presented three damages theories in support of OWS's claims for patent infringement damages:
1. Lost Profits under the *Panduit* Framework. Dr. Keating opined that OWS is entitled to lost profits of approximately $128.9 million, based on an analysis under Panduit Corp. v. Stahlin Bros. Fibre Works, Inc., 575 F.2d 1152 (6th Cir. 1978), demonstrating that OWS satisfies all four Panduit factors for the full infringement period (Q3 2022 through the projected trial date in Q1 2026).
2. Reasonable Royalty under *Georgia-Pacific*. In the alternative, Dr. Keating opined that a reasonable royalty rate of 8.5%, applied to total accused-product revenue of $714.6 million, yields a reasonable royalty of approximately $37.5 million (adjusted downward from the lost-profits figure to account for the different legal standard). This analysis was conducted under the framework established in Georgia-Pacific Corp. v. United States Plywood Corp., 318 F. Supp. 1116 (S.D.N.Y. 1970).
3. Enhanced Damages for Willful Infringement. Dr. Keating provided the economic foundation for an enhanced-damages calculation under 35 U.S.C. § 284, based on Lumicore's willful and deliberate infringement of OWS's patents during the post-notice period (beginning January 15, 2023, the date of the first licensing letter from OWS's prior counsel, FairPoint intellectual Property Group PLLC).
The total damages demand as pled in OWS's Complaint (D.E. 1) is approximately $210.0 million. This figure represents a blended, conservative estimate incorporating lost profits, reasonable royalty, and a selected enhancement on the willful-infringement component, as described in greater detail in Section VII below.
The purpose of this memorandum is threefold. First, it identifies the likely lines of attack in Dr. Volkov's report, based on the litigation positions Lumicore has taken to date in its discovery responses, depositions, and motion practice. Second, it maps out Dr. Keating's anticipated rebuttal to each of Dr. Volkov's expected opinions, providing the substantive foundation for the rebuttal report. Third, it identifies coordination points with the other members of OWS's expert team and the Calderwood & Harkness litigation team, to ensure that the rebuttal report is consistent with the broader trial strategy and positions in the case.
In preparing this outline, we have drawn upon several sources to predict the content and structure of Dr. Volkov's report:
• Lumicore's Answers and Objections to OWS's First Set of Interrogatories (served September 15, 2024), particularly Interrogatory Nos. 8–12 (damages contentions).
• Lumicore's Responses and Objections to OWS's First Set of Requests for Admission (served October 1, 2024), particularly Request Nos. 14–22 (relating to damages methodology and factual predicates).
• The deposition transcript of Janet Moreno, Lumicore's Chief Financial Officer, taken on October 28, 2024, in which Ms. Moreno testified regarding Lumicore's financial performance, product-level margins, and revenue attributable to the accused products.
• The deposition transcript of Dr. Sandra Michaud, Lumicore's Chief Technology Officer, taken on November 19, 2024, in which Dr. Michaud testified regarding the design of the accused products, the absence of any formal design-around analysis, and her awareness of OWS's published patent applications before the launch of the LumiLink product line.
• The deposition transcript of Raymond Bautista, Lumicore's General Counsel, taken on December 3, 2024, in which Mr. Bautista testified regarding Lumicore's pre-suit knowledge of OWS's patents, the referenced (but unproduced) "competent legal opinions," and the absence of a formal design-around analysis before the launch of the LumiEdge 1600G-LR in Q4 2023.
• Dr. Volkov's prior published expert reports in comparable semiconductor patent cases, including Meridian Semiconductor Corp. v. Marlowe Asset Management IC Solutions, Inc., No. 6:21-cv-00412-RWS (E.D. Tex. 2023), and Photon Dynamics, LLC v. Coreline Technologies, Inc., No. 3:20-cv-01887-WHO (N.D. Cal. 2022), which provide insight into her typical report structure, analytical methodology, and the types of arguments she is likely to advance.
• The defense damages strategies employed by Hawthorne Vega & Cruz LLP in prior representations, including their approach to apportionment, royalty base, and Georgia-Pacific analysis in Prisma Semiconductor Corp. v. ArcLight Systems, Inc., No. 2:22-cv-00891-JRG (E.D. Tex. 2024), and NovaTech Photonics, Inc. v. Stratos Communications Corp., No. 1:21-cv-01156-LY (W.D. Tex. 2023).
This memorandum is organized to correspond to the anticipated structure of Dr. Volkov's report, as predicted from the sources described above. Upon receipt of her actual report, Omar Siddiqui will prepare a first-pass analysis comparing the actual report against this outline to flag any unanticipated arguments or methodological approaches that require additional attention.
II. ANTICIPATED STRUCTURE OF DR. VOLKOV'S REBUTTAL REPORT
A. Overview of Anticipated Volkov Opinions
Based on Lumicore's litigation positions, interrogatory responses, and deposition testimony from CFO Janet Moreno and CTO Dr. Sandra Michaud, we anticipate that Dr. Volkov's core opinions will be as follows:
1. OWS is not entitled to lost profits because it cannot satisfy all four *Panduit* factors. Specifically, we expect Dr. Volkov to argue that OWS fails on Factor 2 (absence of acceptable non-infringing substitutes) and possibly Factor 3 (manufacturing and marketing capability to exploit the demand). This is consistent with the defense positions articulated in Lumicore's interrogatory responses and Ms. Moreno's deposition testimony, in which she identified several purported "competitors" in the transceiver market that she claimed offered comparable products.
2. The reasonable royalty rate should be significantly lower than 8.5%, likely in the range of 1.5%–3.0%. We anticipate that Dr. Volkov will rely on what she characterizes as "comparable" license agreements from adjacent but distinct technology fields (e.g., general semiconductor, RF/wireless, digital IC) to support a much lower royalty rate. She may also argue that Dr. Keating's comparable licenses are distinguishable on various grounds.
3. The royalty base should be apportioned to a component level (below the module level), not applied to the entire accused transceiver module price. This is likely to be one of Dr. Volkov's strongest arguments. She will contend that the entire market value rule ("EMVR") does not apply and that the smallest salable patent-practicing unit ("SSPPU") is something smaller than the transceiver module — likely the photonic integrated circuit ("PIC") die or even a specific functional block within the PIC. By reducing the royalty base, she will seek to dramatically reduce the royalty damages figure.
4. Enhanced damages for willful infringement are unwarranted and, in any event, should not be applied to the full post-notice period. Dr. Volkov will likely argue that Lumicore acted in good faith based on "competent legal opinions" (the existence of which has been asserted but not documented in discovery — see pending Motion to Compel, D.E. 38) and that any enhancement would be inappropriate as a matter of law and economics.
5. Total damages. Based on the foregoing opinions, we anticipate that Dr. Volkov's total damages figure will be in the range of $18 million to $35 million. This range is derived by applying her anticipated lower royalty rate (1.5%–3.0%) to a disaggregated royalty base (the PIC-level component value rather than the full module price) over the infringement period, and rejecting any enhancement for willful infringement.
The gap between Dr. Keating's opinions (lost profits of ~$128.9M; reasonable royalty of ~$37.5M; total demand of ~$210.0M including enhancement) and Dr. Volkov's anticipated opinions ($18M–$35M) is significant and will be a central issue at trial. The rebuttal report must demonstrate that Dr. Volkov's opinions are based on unreliable methodology, non-comparable data, and incomplete analysis, and that Dr. Keating's opinions are well-supported and economically sound.
B. Anticipated Organization of Volkov's Report
Based on Dr. Volkov's prior published reports and Hawthorne Vega & Cruz LLP's typical practice, we anticipate that her report will be organized as follows:
(i) Qualifications and Background. A description of Dr. Volkov's educational background (Ph.D. in Economics, University of Texas at Austin; M.B.A., Rice University), professional experience, and prior expert engagements. This section will likely emphasize her experience in semiconductor and technology patent damages cases.
(ii) Summary of Opinions. A high-level overview of her damages conclusions, including her proposed total damages figure, the royalty rate, and the royalty base. This section will also likely include a summary rejection of Dr. Keating's lost-profits analysis.
(iii) Challenge to Lost Profits. A detailed challenge to Dr. Keating's Panduit analysis, focusing primarily on Factors 1 and 2, with secondary arguments on Factors 3 and 4.
(iv) Alternative Reasonable Royalty Analysis. Dr. Volkov's own Georgia-Pacific analysis, presenting a lower royalty rate (1.5%–3.0%) supported by what she will characterize as comparable license agreements.
(v) Apportionment / Royalty Base. Dr. Volkov's argument that the royalty base should be apportioned to a sub-component level, rejecting the EMVR and proposing a disaggregated royalty base.
(vi) Response to Willful Infringement Enhancement. Dr. Volkov's argument that enhanced damages are unwarranted, likely including a discussion of the Read factors and Lumicore's alleged good-faith reliance on legal opinions.
(vii) Conclusion. A summary of Dr. Volkov's opinions and her total damages figure.
We expect Dr. Volkov's report to be accompanied by detailed appendices, including her supporting financial analyses, license comparability charts, and any alternative damages models. Dr. Keating's rebuttal should be structured to address each of these sections in turn, following the organization outlined below.
III. REBUTTAL SECTION 1: DEFENDING THE LOST PROFITS ANALYSIS (PANDUIT FACTORS)
This section addresses Dr. Keating's anticipated rebuttal to Dr. Volkov's challenge to the lost-profits analysis. Dr. Keating's opening report demonstrated that OWS satisfies all four Panduit factors and is therefore entitled to lost profits of approximately $128.9 million. We anticipate that Dr. Volkov will challenge each factor; the rebuttal points for each are set forth below.
A. Anticipated Volkov Challenge: Demand for the Patented Product (Panduit Factor 1)
Dr. Volkov is likely to argue that market demand is for the transceiver module as a whole — a complex, multi-component product — and not specifically for the patented features that OWS alleges are infringed. In particular, she may contend that demand for transceiver modules in the hyperscale data-center market is driven by a combination of factors, including price, form factor, compatibility with existing infrastructure, vendor qualification status, and overall system performance, rather than by the specific patented innovations (waveguide multiplexer per the '847 Patent, thermal management micro-channels per the '562 Patent, coherent transceiver architecture per the '193 Patent, and polarization-diverse grating coupler per the '015 Patent).
This argument mirrors the position Lumicore has taken in its interrogatory responses and is consistent with the approach Dr. Volkov adopted in Meridian Semiconductor Corp. v. Marlowe Asset Management IC Solutions, Inc., where she argued that demand for DRAM modules was driven by industry standards and interoperability rather than specific patented features.
Keating Rebuttal Points:
Dr. Keating should rebut this argument on multiple grounds, drawing on both OWS's own commercial data and the independent analyses of OWS's other retained experts.
First, OWS's own revenue data demonstrates robust market demand for products practicing the asserted patents. OWS's fiscal year 2023 ("FY 2023") revenue was $487.3 million, of which approximately $310 million is attributable to products that practice the asserted patents — specifically, the OWS-400ZR, OWS-800XR, and OWS-1600DR module families. These products represent approximately 63.6% of OWS's total revenue, underscoring the commercial significance of the patented technology. The strong and sustained revenue from these product families over the infringement period demonstrates that the market has consistently demanded products embodying the patented features.
Second, demand for the accused products themselves independently establishes demand under Panduit Factor 1. The accused products — LumiLink 400G-ZR+ ($94.8 million FY 2023 revenue), LumiLink 800G-DR8 ($128.4 million FY 2023 revenue), and LumiEdge 1600G-LR ($41.7 million FY 2023 revenue) — generated total FY 2023 revenue of $264.9 million. That Lumicore was able to generate nearly $265 million in a single fiscal year from products that practice the patented innovations is itself compelling evidence that there is strong demand for products incorporating these features. The accused products represent approximately 84.7% of Lumicore's total FY 2023 revenue of $312.6 million, further demonstrating that the patented features are central to Lumicore's commercial success.
Third, Dr. Keating's rebuttal should cross-reference the industry report prepared by Gregory Falk, an independent industry consultant and formerly Vice President of Engineering at Solara Optics, Inc. Mr. Falk's report establishes that the patented features — specifically, the low-insertion-loss waveguide multiplexing capability of the '847 Patent, the embedded micro-channel thermal management of the '562 Patent, the coherent transceiver integration architecture of the '193 Patent, and the polarization-diverse grating coupling of the '015 Patent — are key differentiators in the hyperscale data-center transceiver market and drive customer purchasing decisions. Mr. Falk's report documents, based on his extensive industry experience and review of customer requirements documents produced in discovery, that hyperscale data-center operators specifically require the optical performance characteristics enabled by these patented innovations (including low insertion loss, high bandwidth density, superior thermal stability under continuous operation, and polarization-diverse signal handling). These are not peripheral features but rather core performance attributes that define the competitive position of transceiver modules in this market.
Fourth, Dr. Caroline Ashford's infringement report (Dr. Ashford is a Professor of Electrical Engineering and Photonics at the Northfield Institute of Technology) confirms that the patented features are integral to the functionality of the accused products. Dr. Ashford's claim-by-claim infringement analysis demonstrates that each of the four asserted patents covers a distinct but essential aspect of the accused products' architecture and operation. The patented features are not optional add-ons or peripheral enhancements; they are embedded in the core design of the accused transceiver modules and directly enable the performance characteristics that customers demand. Dr. Keating should incorporate Dr. Ashford's technical conclusions into his economic analysis to demonstrate that demand for the accused products is demand for the patented features.
Fifth, to the extent Dr. Volkov argues that customer demand is driven by non-patented features such as form factor, connector type, or compliance with industry standards (e.g., QSFP-DD, OSFP), Dr. Keating should respond that these features are industry-standard baseline requirements — they are necessary but not sufficient conditions for market participation. All transceiver modules in the relevant market comply with these standards. The patented features are the differentiating factors that drive customer selection among competing products that all meet the baseline standards. In other words, form factor and standards compliance are table stakes; the patented innovations are the competitive edge.
B. Anticipated Volkov Challenge: Acceptable Non-Infringing Substitutes (Panduit Factor 2)
Panduit Factor 2 requires OWS to demonstrate the absence of acceptable non-infringing substitutes available to consumers during the infringement period (Q3 2022 through the projected trial date in Q1 2026). This is typically the most contested Panduit factor, and we anticipate that Dr. Volkov will devote considerable attention to identifying purported non-infringing alternatives that she contends were available in the market during the relevant period.
Based on Lumicore's interrogatory responses and the positions taken in its discovery correspondence, Dr. Volkov may identify products from other transceiver manufacturers (such as Tidewater Group/Marlowe & Co., Drayton Group, Quillon Industries, or Dunmore Industries's optical division) as purported non-infringing alternatives. She may also argue that customers could have designed around the patented features by using alternative technical approaches (e.g., arrayed waveguide gratings instead of the patented waveguide multiplexer, or conventional surface-mount heat sinks instead of the patented embedded micro-channel thermal management).
Keating Rebuttal Points:
Dr. Keating's opening report analyzed four potential non-infringing alternatives that had been identified in Lumicore's interrogatory responses and found each of them lacking in one or more of the key patented features. Specifically, Dr. Keating determined that none of the identified alternatives provided the combination of (a) the low-insertion-loss waveguide multiplexing capability of the '847 Patent and (b) the embedded micro-channel thermal management of the '562 Patent. The alternatives that offered waveguide multiplexing used alternative architectures (e.g., arrayed waveguide gratings) that resulted in significantly higher insertion losses (2.5–3.8 dB higher per channel, per published specifications), rendering them unacceptable to hyperscale customers with strict link-budget requirements. The alternatives that offered thermal management solutions used conventional approaches (e.g., external heat sinks, thermoelectric coolers) that failed to meet the power-density and form-factor constraints required for next-generation transceiver modules.
Dr. Keating's rebuttal should reinforce these findings and address any new alternatives that Dr. Volkov identifies in her report. In addition, several key pieces of evidence support Dr. Keating's position:
First, Dr. Sandra Michaud's November 19, 2024 deposition testimony is particularly significant. When questioned about specific non-infringing design alternatives that could replicate the performance characteristics of the accused products without infringing OWS's patents, Dr. Michaud was unable to identify any. She testified that the LumiLink product line's optical performance — particularly its low insertion loss and thermal stability — was directly attributable to the design features that correspond to the asserted patent claims. While Dr. Michaud acknowledged that alternative approaches to waveguide multiplexing existed in the academic literature, she conceded that she was not aware of any commercially available product that achieved comparable performance metrics without employing techniques covered by OWS's patents.
Second, Raymond Bautista's December 3, 2024 deposition testimony is equally damaging to Lumicore's position on this factor. Mr. Bautista, Lumicore's General Counsel, admitted under oath that Lumicore did not commission a formal design-around analysis before launching the LumiEdge 1600G-LR product in Q4 2023. This admission is significant for two reasons: (a) it demonstrates that Lumicore itself did not believe a viable non-infringing alternative existed, and (b) it undercuts any argument that non-infringing substitutes were readily available in the market. If acceptable non-infringing alternatives had been available, a reasonable company in Lumicore's position — aware of OWS's patents and having received two licensing letters (January 15, 2023 and May 15, 2023) — would have been expected to pursue a design-around rather than continue with an infringing design.
Third, any alternatives that Dr. Volkov identifies in her report must satisfy three requirements to qualify as "acceptable non-infringing substitutes" under Panduit: they must have been (a) actually available in the market during the infringement period, (b) acceptable to customers as substitutes for the accused products, and (c) non-infringing of OWS's asserted patents. Merely theoretical alternatives, post-hoc constructions, or products that were in development but not commercially available during the relevant period do not qualify. Dr. Keating should scrutinize each purported alternative against these requirements and demonstrate that none satisfies all three.
Fourth, Dr. Keating's rebuttal should incorporate customer testimony and design-in evidence from OWS's Vice President of Sales, Karen Whitfield, and OWS's regional sales directors. This evidence documents that hyperscale data-center operators — the primary customers in this market — have stringent performance requirements for transceiver modules, including specific specifications for insertion loss, optical signal-to-noise ratio, thermal performance under sustained load, and polarization-handling capability. These specifications are codified in customer qualification requirements and vendor approval processes that typically take 12–18 months to complete. Customer testimony and internal OWS documents produced in discovery establish that the specific performance attributes provided by the patented features are required by customers and that no available alternative products met these specifications during the infringement period.
Fifth, Dr. Keating should address the distinction between theoretical and practical non-infringing alternatives. Even if Dr. Volkov identifies an alternative technical approach that does not literally infringe OWS's patents, that alternative must be shown to be practically available — i.e., embodied in a commercially available product, at a competitive price point, with acceptable performance, and capable of being qualified and designed-in within the relevant timeframe. The silicon-photonics transceiver market is characterized by long design-in cycles, extensive customer qualification processes, and high switching costs. A theoretical alternative that might be engineered in a laboratory does not constitute an "acceptable non-infringing substitute" that would have diverted sales away from OWS in the but-for world.
C. Anticipated Volkov Challenge: Manufacturing and Marketing Capability (Panduit Factor 3)
Panduit Factor 3 requires OWS to demonstrate that it had the manufacturing and marketing capability to exploit the demand for the patented products. Dr. Volkov may argue that OWS lacked sufficient manufacturing capacity to serve all of the demand that would have been diverted from Lumicore in the but-for world — i.e., that even if customers would not have purchased Lumicore's accused products, OWS could not have fulfilled the resulting orders due to capacity constraints.
This argument is weaker than the challenges to Factors 1 and 2, and we do not anticipate that Dr. Volkov will devote extensive attention to it. However, she may raise it as a secondary argument to limit the scope of lost profits.
Keating Rebuttal Points:
Dr. Keating should rebut this challenge on several grounds:
First, OWS operates as a fabless semiconductor company, meaning that it outsources the fabrication of its photonic integrated circuits and the assembly of its transceiver modules to third-party foundries and contract manufacturers. This fabless model provides OWS with inherently scalable manufacturing capacity. OWS's primary foundry partner, Drayton Holdings (with facilities in Malta, New York, and Dresden, Germany), has committed capacity allocations that can be expanded on relatively short notice (typically 60–90 days lead time for capacity increases within existing process nodes). OWS's module assembly partner, Halcyon Partners LLC, similarly provides scalable assembly capacity. As a result, OWS's manufacturing capability is not constrained by fixed in-house capacity limitations.
Second, OWS's existing sales and marketing infrastructure covers all major hyperscale data-center operators. OWS's market share displacement analysis, prepared by Dr. Keating and supported by customer data produced in discovery, demonstrates that 12 of the top 20 hyperscale data-center operators (including, on information and belief, Ridgeway Industries, Crestline Systems, Northgate Corporation, Adriatic Enterprises, and several major Chinese cloud providers) purchase transceiver modules from both OWS and Lumicore. OWS maintains active vendor relationships, approved vendor status, and qualified product listings at each of these accounts. Therefore, OWS was fully positioned to fulfill additional orders from these customers in the but-for world.
Third, OWS's established position as the third-largest U.S.-based supplier of silicon-photonics transceiver modules (behind Hollowell Group/Mobileye Photonics and Drayton Group, based on FY 2023 revenue) demonstrates ample marketing capability. OWS maintains a dedicated sales force of approximately 85 sales professionals across North America, Europe, and Asia-Pacific, a robust channel partner network, and active participation in all major industry trade shows and technical conferences. There is no credible basis on which to argue that OWS lacked the marketing reach to capture the diverted sales.
Fourth, to the extent Dr. Volkov cites any specific capacity constraints (e.g., wafer allocation limitations at Drayton Holdings during the global semiconductor supply-chain disruptions of 2022–2023), Dr. Keating should be prepared to respond with evidence that OWS's capacity allocations were sufficient to serve existing demand plus incremental displaced demand, or that additional capacity could have been secured from alternative foundry partners (e.g., Thackeray Holdings, which OWS was in the process of qualifying as a secondary foundry source during the relevant period).
D. Anticipated Volkov Challenge: Amount of Profit (Panduit Factor 4)
Panduit Factor 4 requires OWS to demonstrate the amount of profit it would have earned on the diverted sales. Dr. Keating's opening report calculated OWS's lost profits using a 47.3% gross margin, applied to the displaced revenue derived from the market-share displacement analysis. Dr. Volkov is likely to challenge this margin figure or argue that incremental costs associated with fulfilling the additional demand would reduce the effective margin on displaced sales.
Keating Rebuttal Points:
First, the 47.3% gross margin figure is not an estimate or projection — it is sourced directly from OWS's audited internal financial statements for FY 2023, which were produced in discovery and verified by OWS's independent auditor, Gresham & Whitfield LLP (San Jose office). This figure reflects OWS's actual cost of revenue, including foundry costs, module assembly costs, test and qualification costs, and yield-related costs. It is the appropriate measure of the profit OWS would have earned on additional sales of its patented products.
Second, OWS's overall financial performance confirms the robustness of this margin. OWS reported FY 2023 net income of $61.4 million on revenue of $487.3 million, representing a net income margin of approximately 12.6%. The difference between the 47.3% gross margin and the 12.6% net margin reflects OWS's operating expenses (R&D, SG&A, depreciation, and amortization), which are predominantly fixed costs that would not increase proportionally with incremental displaced sales.
Third, Dr. Keating should present an incremental-cost analysis demonstrating that, because OWS is a fabless company, the incremental manufacturing costs associated with additional displaced sales would scale linearly with foundry orders (i.e., the cost per unit of additional wafer fabrication, assembly, and test would be approximately the same as the cost per unit reflected in OWS's existing financials). Fixed overhead — including R&D expenses, corporate overhead, executive compensation, and sales and marketing expenses — is already absorbed by OWS's existing revenue base and would not increase with incremental displaced sales. As a result, the margin on incremental displaced sales would be at or above 47.3%, making the lost-profits figure conservative.
Fourth, Dr. Keating should anticipate and address any attempt by Dr. Volkov to substitute Lumicore's margins as an alternative measure of profitability. Lumicore's gross margins on the accused products are higher than OWS's: 52.1% for the LumiLink 400G-ZR+, 58.3% for the LumiEdge 1600G-LR, and intermediate margins for the LumiLink 800G-DR8, per Lumicore's AEO-designated product-level financial data produced in discovery. However, Lumicore's margins are irrelevant to the lost-profits calculation — the question under Panduit Factor 4 is what profit OWS would have earned on the displaced sales, not what profit Lumicore actually earned. Any argument by Dr. Volkov that seeks to introduce Lumicore's margins into the lost-profits analysis confuses the legal standard and should be rejected.
Fifth, if Dr. Volkov argues that OWS would have needed to reduce prices to capture the displaced sales (i.e., that OWS would have had to undercut Lumicore's pricing to win the diverted customers), Dr. Keating should respond that the but-for world assumes the absence of Lumicore's infringing products from the market — in that world, OWS would not have faced the competitive pressure that might have required price reductions. To the contrary, as discussed in Section IX.A below regarding price erosion, OWS actually experienced downward pricing pressure during the infringement period because of Lumicore's market entry with infringing products.
E. Market Share Displacement Figure (38.2%)
A critical input to Dr. Keating's lost-profits model is the 38.2% market-share displacement assumption, which represents Dr. Keating's estimate of the percentage of Lumicore's accused-product sales that would have been captured by OWS in the but-for world (i.e., a world in which Lumicore's infringing products were not available). Applying this figure to total projected accused-product revenue of $714.6 million, and then applying OWS's 47.3% gross margin, yields the approximately $128.9 million lost-profits figure ($714.6M × 38.2% × 47.3% ≈ $128.9M).
Dr. Volkov is likely to challenge the 38.2% displacement figure as overstated, arguing either that OWS would have captured a smaller share of Lumicore's diverted sales or that a larger portion of those sales would have gone to other non-infringing competitors.
Keating Rebuttal Points:
First, the 38.2% market-share displacement figure is based on a rigorous analysis of customer overlap data, demonstrating that OWS and Lumicore compete head-to-head at 12 of the top 20 hyperscale data-center operators. This analysis was conducted using customer-by-customer purchase data produced in discovery by both OWS and Lumicore, corroborated by publicly available industry reports from Dunmore Consulting Group and Granite Harbor Management Consulting. The 38.2% figure was adjusted downward from the raw customer overlap percentage (approximately 60%) to account for (a) customers where OWS is not currently an approved vendor and would have needed to complete a qualification process, (b) customers with contractual commitments to Lumicore or other suppliers that would have limited their ability to switch to OWS in the short term, and (c) a reasonable allocation of some displaced sales to other non-infringing competitors in the market.
Second, the 38.2% figure is conservative. It does not assume that OWS would have captured all of Lumicore's displaced sales — far from it. The figure explicitly acknowledges that approximately 61.8% of Lumicore's sales would have gone to other competitors or would not have been made at all in the but-for world. This is a more conservative approach than the method accepted by some courts, which have allowed patentees to claim a larger share of displaced sales where the patentee and infringer are the two dominant competitors in a market. Here, OWS and Lumicore together represent approximately 55–60% of the U.S. silicon-photonics transceiver market for hyperscale data-center applications, which supports a displacement figure at or above 38.2%.
Third, any alternative market-share displacement figure proposed by Dr. Volkov must account for the "but for" world — a hypothetical marketplace in which Lumicore's infringing products do not exist — and the absence of acceptable non-infringing substitutes. If Dr. Volkov's analysis assumes that customers would have switched to other products that she identifies as non-infringing alternatives, her displacement figure is only as strong as her Panduit Factor 2 analysis. If those purported non-infringing alternatives fail the acceptability and availability tests described in Section III.B above, then the displacement figure must be adjusted upward, not downward.
Fourth, Dr. Keating should prepare sensitivity analyses showing the impact of alternative displacement assumptions on the lost-profits figure. For example, if the displacement figure were reduced to 30% (below any figure supportable by the evidence), lost profits would still be approximately $101.4 million; if increased to 45% (arguably still conservative given the two-firm dominance of the market), lost profits would be approximately $151.3 million. These sensitivity analyses demonstrate that the lost-profits figure is robust across a range of reasonable displacement assumptions.
IV. REBUTTAL SECTION 2: DEFENDING THE REASONABLE ROYALTY ANALYSIS
Dr. Keating's opening report presented an alternative reasonable royalty analysis under the Georgia-Pacific framework, opining that the appropriate reasonable royalty rate is 8.5% of accused-product revenue. Applied to total accused-product revenue of $714.6 million over the infringement period, this yields a reasonable royalty of approximately $37.5 million (adjusted to account for the different base period used in the reasonable royalty analysis compared to the lost-profits analysis, as certain periods overlap). We anticipate that Dr. Volkov will challenge the royalty rate, the Georgia-Pacific factor analysis, and the hypothetical negotiation framework.
A. Anticipated Volkov Challenge: Royalty Rate
Dr. Volkov is likely to propose a significantly lower royalty rate, estimated at 1.5%–3.0%, based on what we anticipate will be three primary lines of argument:
(a) Reliance on license agreements from different technology fields that are not comparable to the silicon-photonics transceiver market;
(b) A selective or incomplete Georgia-Pacific factor analysis that underweights or ignores factors favorable to a higher royalty rate; and
(c) A failure to account for the full value of the patented technology as reflected in OWS's licensing history and the commercial success of the patented products.
Keating Rebuttal Points:
Dr. Keating's proposed 8.5% royalty rate is supported by five comparable license agreements, each of which involves the same or closely related technology (silicon photonics and optical transceivers) and provides a reliable benchmark for the hypothetical negotiation:
1. OWS–Terralink Semiconductor License (2020). A running royalty of 6.5% on approximately $45 million in licensed revenue, covering a portfolio of OWS patents including the '847 Patent (waveguide multiplexer) and the '562 Patent (thermal management). This license was negotiated at arm's length between OWS and Terralink, a smaller competitor in the silicon-photonics market, and represents a floor for the royalty rate in a license between direct competitors of comparable size (OWS and Lumicore are larger and more directly competitive than OWS and Terralink).
2. OWS–NovaWave Corp. License (2021). A running royalty of 9.0% on approximately $22 million in licensed revenue, covering the same patent portfolio. This license was negotiated after NovaWave launched a transceiver product that OWS believed infringed the asserted patents. The 9.0% rate reflects the premium associated with licensing after product launch, which is analogous to the hypothetical negotiation in this case (the hypothetical negotiation is assumed to occur just before Lumicore's launch of the first accused product in Q3 2022).
3. OWS–Helix Integrated Photonics License (2022). A running royalty of 7.75% plus a $2.5 million upfront payment on approximately $68 million in licensed revenue. This license, the most recent in OWS's licensing program, represents the best evidence of the royalty rate that OWS has been able to command for licenses to the asserted patents. The combined effective rate (running royalty plus amortized upfront payment) is approximately 11.4%, which is above Dr. Keating's proposed 8.5% rate.
4. Industry Comparable: *Stratos Photonics v. ArcLight Systems* Settlement (2019). A royalty rate of 8.0%, negotiated in settlement of a silicon-photonics patent infringement case involving transceiver technology. While this license involved different patents, the technology field, market conditions, and economic dynamics are closely comparable.
5. Industry Comparable: *Coreline Technologies v. Prisma Semiconductor* Cross-License (2021). An effective royalty rate of 10.2%, reflecting the net payment terms of a cross-license agreement between two silicon-photonics competitors. This agreement is relevant because it demonstrates that royalty rates in the 8%–12% range are standard in this industry for comparable patent portfolios.
The range of comparable transactions in the silicon-photonics industry is approximately 4.0%–12.5%. Dr. Keating's proposed 8.5% rate falls squarely within this established range and is supported by the specific characteristics of the hypothetical negotiation in this case (direct competitors, valuable and commercially proven patents, no acceptable non-infringing alternatives, and a licensee that was aware of the patents before launching its products).
Any licenses that Dr. Volkov relies upon to support a lower rate must be scrutinized for comparability. To be probative, a comparable license must involve: (a) the same or closely related technology field (silicon photonics and optical transceivers); (b) a comparable patent portfolio in terms of quality, scope, and commercial significance; (c) comparable market conditions (including the competitive dynamics of the hyperscale data-center transceiver market); and (d) comparable temporal proximity to the hypothetical negotiation date (Q2 2022). Licenses from other technology fields — such as general semiconductor, RF/wireless, digital IC, or MEMS technologies — are not probative because the silicon-photonics transceiver market has materially different economic characteristics, including higher barriers to entry, a smaller number of competitors, higher per-unit prices, longer design-in cycles, and greater R&D intensity.
B. Anticipated Volkov Challenge: Georgia-Pacific Factor Analysis
Dr. Volkov is likely to present her own Georgia-Pacific factor analysis, which we anticipate will be incomplete or selective — either addressing fewer than all 15 factors, or weighting certain factors in a manner that systematically favors a lower royalty rate while ignoring or underweighting factors that support a higher rate.
Keating Rebuttal Points:
Dr. Keating conducted a comprehensive analysis of all 15 Georgia-Pacific factors in his opening report, as detailed in Appendix B thereto. His rebuttal should address any omissions or mischaracterizations in Dr. Volkov's analysis, with particular attention to the following factors that we anticipate Dr. Volkov will underweight or ignore:
Factor 1 — Royalties received by the patentee for licensing the patent-in-suit. This factor strongly favors a higher royalty rate. OWS has entered into three existing license agreements directly on the asserted patents (the Terralink, NovaWave, and Helix licenses described above), with royalty rates ranging from 6.5% to 9.0% (and an effective combined rate as high as 11.4% for the Helix license). These are not hypothetical or speculative benchmarks; they are actual, arm's-length transactions involving the same patents at issue in this case. Dr. Volkov cannot credibly propose a 1.5%–3.0% rate without explaining why the hypothetical negotiation with Lumicore would have produced a rate 50%–80% below what OWS has actually received from other licensees for the same patents. Any attempt to distinguish these licenses (e.g., by arguing that they involved smaller licensees or different market conditions) must be weighed against the fact that a license with a larger, more directly competitive licensee like Lumicore would command a higher rate, not a lower one. See Factor 5 below.
Factor 5 — Commercial relationship between the licensor and licensee. OWS and Lumicore are direct competitors in the silicon-photonics transceiver market. They compete head-to-head at 12 of the top 20 hyperscale data-center operators. A license between direct competitors commands a premium over a license between a patentee and a non-competitor (or a smaller, non-overlapping competitor) because the licensor is effectively enabling a direct rival to compete using the licensor's own technology. This factor drives the royalty rate upward, as the Federal Circuit has repeatedly recognized. See, e.g., Rite-Hite Corp. v. Kelley Co., 56 F.3d 1538, 1554 (Fed. Cir. 1995) (en banc). Dr. Volkov's analysis, if it proposes a 1.5%–3.0% rate, implicitly ignores or drastically underweights the competitive premium.
Factor 8 — Established profitability of the patented product, its commercial success, and its current popularity. OWS's products practicing the asserted patents have generated approximately $310 million in FY 2023 revenue with a gross margin of 47.3%. The patented products have been commercially successful since their initial launch and have maintained strong demand throughout the infringement period. These products are among the most popular transceiver modules in the hyperscale data-center market. This factor supports a high royalty rate because it demonstrates that the patented technology is commercially proven and valuable.
Factor 9 — Advantages of the patented device over old modes or devices, if any, that had been used for working out similar results. The patented innovations represent significant technical advances over prior approaches to transceiver module design. The low-insertion-loss waveguide multiplexing of the '847 Patent provides a 2.5–3.8 dB improvement over conventional arrayed waveguide gratings. The embedded micro-channel thermal management of the '562 Patent enables significantly better thermal performance in a smaller form factor compared to conventional heat sinks and thermoelectric coolers. The coherent transceiver integration architecture of the '193 Patent enables higher bandwidth density and lower power consumption compared to prior coherent receiver designs. The polarization-diverse grating coupling of the '015 Patent improves coupling efficiency and signal quality compared to prior edge-coupling approaches. These technical advantages are documented in Dr. Ashford's infringement report and in the prosecution histories of the asserted patents. This factor strongly supports a high royalty rate.
Factor 11 — Extent to which the infringer has made use of the invention, and any evidence probative of the value of that use. This factor is particularly powerful in this case. Lumicore's accused products generated $264.9 million in FY 2023 revenue, representing approximately 84.7% of Lumicore's total FY 2023 revenue of $312.6 million. In other words, nearly 85% of Lumicore's entire business is built on products that practice OWS's patents. This extraordinary level of use demonstrates that the patented technology is of central importance to Lumicore's business and justifies a commensurately high royalty rate. A willing licensee in Lumicore's position — knowing that the licensed technology would constitute the foundation of nearly its entire product portfolio — would be willing to pay a premium for access to that technology.
Factor 13 — Portion of the realizable profit that should be credited to the invention as distinguished from non-patented elements, manufacturing processes, business risks, or significant features or improvements added by the infringer. The patented features are core to the transceiver module's value proposition. As discussed in Section V below regarding apportionment, the patented innovations are not peripheral features that can be separated from the overall product — they are the fundamental technology that enables the accused products' competitive performance. Customer demand is driven by the optical performance characteristics (bandwidth, insertion loss, thermal stability, polarization handling) that are directly provided by the patented innovations. While Lumicore has contributed its own engineering and manufacturing capabilities to the accused products, the patented features represent the critical differentiating technology that makes Lumicore's products competitive. This factor supports attributing a significant portion of the realizable profit to the patented invention, which in turn supports a high royalty rate.
C. Anticipated Volkov Challenge: Hypothetical Negotiation Framework
Dr. Volkov may attempt to recast the hypothetical negotiation in a manner that assumes a weaker bargaining position for OWS or introduces factors not properly part of the Georgia-Pacific framework. For example, she may argue that OWS would have been a willing licensor at a lower rate because of competitive pressure from other market participants, or that Lumicore's willingness to pay was constrained by the availability of non-infringing alternatives, or that the parties would have agreed to a lump-sum payment rather than a running royalty.
Keating Rebuttal Points:
Dr. Keating should respond to any such arguments by reaffirming the legal framework governing the hypothetical negotiation. The hypothetical negotiation must assume that both parties are willing: OWS as a willing licensor and Lumicore as a willing licensee, each acting rationally with full knowledge of the patent's validity and the fact of infringement. Georgia-Pacific, 318 F. Supp. at 1120. The negotiation is not a real-world bargaining scenario in which the patentee faces uncertainty about validity or the infringer can threaten to walk away — it is a construct designed to determine the price a willing buyer would pay and a willing seller would accept for a license to the patented technology, knowing that the patents are valid and infringed.
Several aspects of Lumicore's actual conduct are probative of the value it placed on the patented technology and the price it would have been willing to pay in the hypothetical negotiation:
First, Lumicore's pre-suit conduct demonstrates the high value Lumicore placed on the technology. Lumicore launched the LumiLink 400G-ZR+ in Q3 2022 despite awareness of OWS's published patent applications. Dr. Michaud admitted in her November 19, 2024 deposition that she personally reviewed OWS's Published Application No. 2018/0123456 (which matured into the '847 Patent) before the LumiLink design was finalized. This evidence demonstrates that Lumicore made a deliberate decision to incorporate the patented technology into its products, knowing the risk of infringement, because the technology was essential to achieving competitive performance. A willing licensee that valued the technology highly enough to take that risk would have been willing to pay a commensurately high royalty rate.
Second, Lumicore's post-notice conduct further demonstrates the value of the patented technology to Lumicore's business. After receiving FairPoint intellectual Property Group PLLC's first licensing letter on January 15, 2023, and a second letter on May 15, 2023, Lumicore did not cease manufacturing or selling the accused products, did not commission a design-around analysis (per Mr. Bautista's deposition testimony), and did not seek a license. Instead, Lumicore launched an additional accused product — the LumiEdge 1600G-LR — in Q4 2023, expanding the scope of infringement. This conduct demonstrates that Lumicore regarded the patented technology as indispensable to its product line and was willing to accept the risk of patent infringement damages rather than forego the technology. This is strong evidence that, in a hypothetical negotiation, Lumicore would have been willing to pay a high royalty rate to secure a license.
Third, Dr. Keating should rebut any attempt by Dr. Volkov to introduce the availability of non-infringing alternatives into the hypothetical negotiation framework in a manner that artificially reduces Lumicore's willingness to pay. In the hypothetical negotiation, both parties are assumed to know that the patents are valid and infringed; the question is what price would be agreed upon for a license. While the availability of non-infringing alternatives is relevant under Georgia-Pacific Factor 6, it must be assessed based on the actual facts — and, as discussed in Section III.B above, no acceptable non-infringing substitutes were available during the infringement period.
Fourth, the hypothetical negotiation date in this case is Q2 2022, just before Lumicore's launch of the first accused product (LumiLink 400G-ZR+) in Q3 2022. At that date, a willing licensee in Lumicore's position would have known: (a) OWS's patents were valid and enforceable; (b) the planned LumiLink product line would infringe those patents; (c) the patented technology was essential to achieving competitive performance in the transceiver market; (d) no acceptable non-infringing alternatives were available; and (e) OWS's existing licensing program had established royalty rates in the range of 6.5%–9.0% for the same patents. A rational willing licensee with this knowledge would have been willing to pay at least 8.5% to secure a license.
V. REBUTTAL SECTION 3: APPORTIONMENT / ROYALTY BASE
A. Anticipated Volkov Challenge: Entire Market Value Rule (EMVR)
The apportionment argument is expected to be one of Dr. Volkov's strongest and most heavily litigated points. We anticipate that she will contend that the entire accused transceiver module price cannot serve as the royalty base because the patented features constitute only a component of the module's overall value. She will likely argue that the EMVR does not apply and that the smallest salable patent-practicing unit ("SSPPU") is something smaller than the transceiver module — potentially the photonic integrated circuit ("PIC") die alone, or even a specific functional block within the PIC (e.g., the waveguide multiplexer array or the grating coupler structure).
By reducing the royalty base from the full module price (average selling prices ranging from approximately $350 for the LumiLink 400G-ZR+ to approximately $2,200 for the LumiEdge 1600G-LR) to a sub-component value (potentially as low as $40–$80 for the PIC die alone), Dr. Volkov can dramatically reduce the royalty damages figure even while appearing to concede a reasonable royalty rate. This is a common defense strategy in semiconductor patent cases and one that Hawthorne Vega & Cruz LLP has employed effectively in prior representations.
Keating Rebuttal Points:
Dr. Keating's rebuttal on apportionment should be comprehensive and address both the legal and economic foundations of the EMVR analysis:
First, the smallest salable patent-practicing unit for the accused products is the transceiver module itself. The patented features — the waveguide multiplexer (U.S. Patent No. [the '847 Patent]), the coherent transceiver integration architecture (U.S. Patent No. [the '193 Patent]), the embedded micro-channel thermal management (U.S. Patent No. [the '562 Patent]), and the polarization-diverse grating coupler (U.S. Patent No. [the '015 Patent]) — are not separately sold or separately salable components. They are integrated structures fabricated directly into the photonic IC die, which is itself integral to and inseparable from the transceiver module. Lumicore does not sell PIC dice separately; it does not sell waveguide multiplexer arrays separately; it does not sell grating coupler structures separately. The transceiver module is the unit of commerce — the product that Lumicore manufactures, markets, sells, and imports. It is the smallest unit that customers can purchase and that Lumicore can sell. Accordingly, the transceiver module is the appropriate SSPPU and royalty base.
Second, even if the SSPPU analysis were not dispositive, the EMVR applies because the patented features drive demand for the entire transceiver module. The Federal Circuit has held that the entire product price may serve as the royalty base where the patented feature is the basis for customer demand. See Rite-Hite Corp. v. Kelley Co., 56 F.3d 1538 (Fed. Cir. 1995) (en banc); Lucent Techs., Inc. v. Gateway, Inc., 580 F.3d 1301 (Fed. Cir. 2009). Here, customers purchase transceiver modules because of their optical performance characteristics — bandwidth, insertion loss, thermal stability, and polarization handling — which are directly provided by the patented innovations. The evidence supporting this demand nexus includes:
(a) Gregory Falk's industry report, which establishes that the patented features are the key differentiators in the hyperscale data-center transceiver market and the primary drivers of customer purchasing decisions;
(b) Dr. Caroline Ashford's infringement report, which demonstrates that the patented features are integral to the functionality of the accused products and cannot be removed or replaced without fundamentally altering the product's performance;
(c) Customer testimony and design-in evidence from OWS's sales organization, documenting that hyperscale customers specifically require the performance attributes provided by the patented features;
(d) Lumicore's own marketing materials for the accused products, which prominently feature the performance characteristics enabled by the patented innovations (low insertion loss, superior thermal performance, high bandwidth density, polarization-diverse operation) as key selling points; and
(e) Lumicore's pricing data, which shows that the accused products command premium prices compared to non-infringing transceiver modules that do not offer comparable performance, further demonstrating that customers are willing to pay for the value provided by the patented features.
Third, Dr. Keating's rebuttal should incorporate the independent teardown reports prepared by Benchmark Semiconductor Analysis, LLC (2244 Mission College Boulevard, Santa Clara, CA 95054; Lead Engineer: Kevin Ng). Benchmark conducted detailed physical teardowns of all three accused products:
• LumiLink 400G-ZR+ teardown report, dated December 12, 2022;
• LumiLink 800G-DR8 teardown report, dated June 2023; and
• LumiEdge 1600G-LR teardown report, dated February 2024.
These teardown reports confirm that the patented structures — waveguide multiplexer arrays, embedded micro-channel thermal management structures, coherent transceiver integration circuitry, and polarization-diverse grating couplers — are physically integrated into the PIC die and the transceiver module. They are fabricated as part of the semiconductor manufacturing process and cannot be physically separated from the rest of the device. This physical integration further supports the conclusion that the transceiver module is the appropriate SSPPU and that the patented features are inseparable from the product as sold.
Fourth, Dr. Keating should directly address any attempt by Dr. Volkov to use a disaggregated royalty base. If Dr. Volkov proposes a royalty base limited to the PIC die value (or a functional block within the PIC), Dr. Keating should demonstrate that such disaggregation lacks economic foundation. The accused transceiver modules are not sold as component parts — they are sold as fully integrated, tested, and qualified modules. Customers do not purchase PIC dice separately, and there is no established market price for the PIC die as a standalone component. Any value Dr. Volkov assigns to the PIC die (or a sub-component thereof) would necessarily be an artificial construct derived from internal cost accounting, bill-of-materials data, or arbitrary allocation — not a market-based measure of value. Such disaggregation would violate the economic principle that the royalty base should reflect the value of the infringing product as it is actually sold in the marketplace.
Fifth, even if the Court were to reject the EMVR and require apportionment below the module level, Dr. Keating should demonstrate that a properly apportioned analysis — applying a higher royalty rate to a smaller base — yields a substantially similar total damages figure. This is because the paten