UNITED STATES DISTRICT COURT EASTERN DISTRICT OF TEXAS MARSHALL DIVISION

UNITED STATES DISTRICT COURT EASTERN DISTRICT OF TEXAS MARSHALL DIVISION
OPTIWAVE SEMICONDUCTOR CORP.,
Plaintiff,
v.
LUMICORE TECHNOLOGIES, INC.,
Defendant.
Case No. 2:24-cv-00319-JRG
Hon. James R. Gilstrap
JURY TRIAL DEMANDED
PLAINTIFF OPTIWAVE SEMICONDUCTOR CORP.'S OPPOSITION TO DEFENDANT'S MOTION FOR PARTIAL SUMMARY JUDGMENT TO EXCLUDE ENTIRE-MODULE DAMAGES UNDER THE ENTIRE MARKET VALUE RULE
Alan Ngo (CA Bar No. 287341; admitted pro hac vice) Marcus Delgado (TX Bar No. 24098712) Wendy Schultz (CA Bar No. 302156; admitted pro hac vice) Omar Siddiqui (CA Bar No. 318904; admitted pro hac vice) CALDERWOOD & HARKNESS LLP 350 South Grand Avenue, 48th Floor Los Angeles, California 90071 Telephone: (213) 629-8400 Facsimile: (213) 629-8401
Email: ango@calderwoodharkness.com
Email: mdelgado@calderwoodharkness.com
Email: wschultz@calderwoodharkness.com
Email: osiddiqui@calderwoodharkness.com
James "Jim" Callahan (TX Bar No. 03791465) WHITMORE & CALLAHAN LLP 110 North College Avenue, Suite 800 Tyler, Texas 75702 Telephone: (903) 597-6200 Facsimile: (903) 597-6201
Email: jcallahan@whitmorecallahan.com
Attorneys for Plaintiff Optiwave Semiconductor Corp.
Dated: ____
TABLE OF CONTENTS
Right-click to update Table of Contents
TABLE OF AUTHORITIES
Cases
Statutes and Rules
Secondary Sources
I. INTRODUCTION
Defendant Lumicore Technologies, Inc. ("Lumicore") is anticipated to move for partial summary judgment seeking to exclude the damages analysis of Plaintiff Optiwave Semiconductor Corp.'s ("OWS") expert, Dr. Lawrence Keating, to the extent that Dr. Keating applies a royalty rate to the entire price of Lumicore's accused transceiver modules. Lumicore's anticipated argument rests on a fundamental misapplication of the Entire Market Value Rule ("EMVR"). The Court should reject it.
The patented technologies at the heart of this case — covered by U.S. Patent Nos. 10,234,847 (the "'847 Patent"), 10,891,193 (the "'193 Patent"), 9,678,562 (the "'562 Patent"), and 11,402,015 (the "'015 Patent") — are not minor, severable components buried within a larger multifunction product. They are the core technological innovations that define the accused LumiLink 400G-ZR+, LumiLink 800G-DR8, and LumiEdge 1600G-LR silicon-photonics transceiver modules. These patents cover the optical signal multiplexing, coherent transceiver integration, thermal management, and polarization-diverse light coupling subsystems that together constitute the very reason customers purchase these modules. Without the patented features, the accused products could not function as advertised and would have no commercial viability.
Critically, the transceiver module itself is the smallest salable patent-practicing unit ("SSPPU"). The patented waveguide multiplexer, coherent transceiver architecture, micro-channel heat sink, and polarization-diverse grating coupler are not separately sold sub-components — they are fabricated into the photonic integrated circuit ("PIC") at the core of each module and cannot be physically or commercially separated. Lumicore does not sell PICs, waveguide multiplexers, thermal management subsystems, or grating couplers as standalone products. No market participant does. The transceiver module is the indivisible commercial unit, and its price is the only economically meaningful royalty base.
Because the module is the SSPPU, the EMVR limitation does not apply — there is no "entire market" to apportion away from. But even if the Court were to apply the EMVR framework, OWS easily satisfies the demand-driving requirement. The record evidence — including Lumicore's own marketing materials, deposition testimony of Lumicore's Chief Technology Officer, independent teardown analyses, customer purchasing data, comparable license agreements, and expert testimony — overwhelmingly demonstrates that the patented features are the predominant basis for customer demand for the accused transceiver modules. Dr. Keating's application of an 8.5% royalty rate to the module-level price is methodologically rigorous, grounded in comparable licenses, and fully consistent with Federal Circuit precedent.
Lumicore's anticipated motion conflates the EMVR limitation — which prevents recovery of damages attributable to unpatented features when the patented feature is a minor component of a larger product — with a broader apportionment requirement that has no basis in law when the SSPPU is the module itself and the patented features drive the entirety of customer demand. The motion should be denied in its entirety.
II. STATEMENT OF FACTS
A. Optiwave Semiconductor Corp. and Its Patented Technologies
Optiwave Semiconductor Corp. is a Delaware corporation headquartered at 4810 Bayside Technology Parkway, Suite 300, Santa Clara, California 95054. Ex. 3 (Falk Decl.) ¶ 5. Founded in 2011, OWS has grown to approximately 1,250 employees and is the third-largest U.S. supplier of silicon-photonics transceiver modules for hyperscale data centers. Id. ¶ 6. For fiscal year 2023, OWS reported revenue of $487.3 million and net income of $61.4 million, as reflected in financial statements audited by Gresham & Whitfield LLP. Ex. 1 (Keating Decl.) ¶ 14. OWS holds 142 issued U.S. patents and 58 pending patent applications directed primarily to silicon-photonics technologies for data-center communications. Ex. 2 (Ashford Decl.) ¶ 8.
In 2019, Ridgeline Capital Partners, LP, a private equity firm headquartered at 200 Clarendon Street, 40th Floor, Boston, Massachusetts 02116, acquired a 55% ownership stake in OWS for $780 million — an investment that reflected, in significant part, the value of OWS's patent portfolio and its competitive position in the silicon-photonics market. Ex. 1 (Keating Decl.) ¶ 15. OWS's key commercial products — the OWS-400ZR, OWS-800XR, and OWS-1600DR transceiver module families — practice the four asserted patents and compete directly with the accused Lumicore products in the hyperscale data-center market. Id. ¶ 16; Ex. 3 (Falk Decl.) ¶ 11.
B. The Asserted Patents
The four patents asserted in this action collectively cover the critical functional subsystems of a high-speed silicon-photonics transceiver module: optical signal multiplexing, coherent transceiver integration, thermal management, and polarization-diverse light coupling. Ex. 2 (Ashford Decl.) ¶¶ 12–30.
The '847 Patent (U.S. Patent No. 10,234,847), titled "Multi-Channel Silicon Photonic Waveguide Multiplexer with Low Insertion Loss," was filed on September 14, 2016, and issued on March 19, 2019. OWS is the original assignee. The '847 Patent claims a waveguide multiplexer structure that enables multiple optical channels to be combined and separated on a silicon photonic chip with dramatically reduced insertion loss compared to prior-art approaches. Ex. 2 (Ashford Decl.) ¶¶ 13–16. Low insertion loss is critical because every decibel of loss in the multiplexer directly reduces the link budget available for data transmission over optical fiber, limiting the achievable reach and data rate of the transceiver module. Id. ¶ 17.
The '193 Patent (U.S. Patent No. 10,891,193), titled "High-Bandwidth Coherent Optical Transceiver Integrated on Silicon Substrate," was filed on April 7, 2017, claiming priority to U.S. Provisional Application No. 62/499,117, filed the same date. The '193 Patent issued on January 12, 2021, with OWS as the original assignee. This patent claims a coherent optical transceiver architecture fabricated on a silicon substrate, enabling high-bandwidth data communication using advanced modulation formats (such as DP-QPSK and DP-16QAM) within a compact, cost-effective silicon-photonics platform. Ex. 2 (Ashford Decl.) ¶¶ 18–21. Coherent detection is the defining technological capability that distinguishes the accused products from legacy direct-detect transceiver modules, enabling dramatically higher data rates over longer distances. Id. ¶ 22.
The '562 Patent (U.S. Patent No. 9,678,562), titled "Method for Thermal Management of Photonic Integrated Circuits Using Embedded Micro-Channel Heat Sinks," was filed on November 22, 2014, and issued on June 13, 2017. The original assignee was Solara Optics, Inc. OWS acquired the '562 Patent, along with 11 other patents, from Solara Optics pursuant to an asset purchase agreement dated June 15, 2018, for a total consideration of $34.5 million. Ex. 1 (Keating Decl.) ¶ 22. The '562 Patent claims a method for embedding micro-channel heat sinks directly within or adjacent to a photonic integrated circuit, providing efficient thermal dissipation for the high power densities generated by coherent transceiver PICs. Ex. 2 (Ashford Decl.) ¶¶ 23–25. Without effective thermal management, high-power PICs overheat, causing wavelength drift, increased bit-error rates, and ultimately device failure — rendering the transceiver module commercially unviable. Id. ¶ 26.
The '015 Patent (U.S. Patent No. 11,402,015), titled "Polarization-Diverse Grating Coupler for Silicon Photonic Transceivers," was filed on August 30, 2019, and issued on August 2, 2022, with OWS as the original assignee. The '015 Patent shares common provisional application No. 62/499,117 with the '193 Patent. It claims a polarization-diverse grating coupler structure that efficiently couples light between an optical fiber and the silicon photonic waveguide regardless of the polarization state of the incoming light. Ex. 2 (Ashford Decl.) ¶¶ 27–29. Polarization diversity is essential for reliable operation in data-center environments where the polarization state of light arriving through optical fiber is unpredictable and time-varying. Id. ¶ 30.
The '562 and '847 Patents share a continuation relationship, reflecting a unified inventive program in waveguide and thermal integration technologies. Ex. 2 (Ashford Decl.) ¶ 31. The '193 and '015 Patents share common provisional application No. 62/499,117, reflecting a coordinated effort to develop coherent transceiver and coupling technologies for silicon photonics platforms. Id. Taken together, these four patents cover essentially all of the core optoelectronic subsystems within a high-speed silicon-photonics transceiver module.
C. The Accused Products and Their Architecture
Lumicore Technologies, Inc. manufactures and sells the three accused product families: the LumiLink 400G-ZR+, the LumiLink 800G-DR8, and the LumiEdge 1600G-LR. Each is a pluggable silicon-photonics optical transceiver module sold to hyperscale data-center operators. Ex. 1 (Keating Decl.) ¶ 28.
LumiLink 400G-ZR+. Launched in the third quarter of 2022, the LumiLink 400G-ZR+ is a 400-gigabit coherent transceiver module designed for data-center interconnect applications. It is accused of infringing the '847, '193, and '562 Patents. For fiscal year 2023, the LumiLink 400G-ZR+ generated approximately $94.8 million in revenue for Lumicore. Ex. 1 (Keating Decl.) ¶ 29; Ex. 8 (Lumicore Financial Data Excerpts).
LumiLink 800G-DR8. Launched in the first quarter of 2023, the LumiLink 800G-DR8 is an 800-gigabit transceiver module employing eight parallel optical lanes at 100 gigabits per second each. It is accused of infringing all four asserted patents — the '847, '193, '562, and '015 Patents. For fiscal year 2023, the LumiLink 800G-DR8 generated approximately $128.4 million in revenue for Lumicore, making it Lumicore's highest-revenue product. Ex. 1 (Keating Decl.) ¶ 30; Ex. 8.
LumiEdge 1600G-LR. Launched in the fourth quarter of 2023, the LumiEdge 1600G-LR is a 1.6-terabit transceiver module designed for long-reach intra-data-center applications. It is accused of infringing the '193 and '015 Patents. For fiscal year 2023 (reflecting only one quarter of sales), the LumiEdge 1600G-LR generated approximately $41.7 million in revenue. Ex. 1 (Keating Decl.) ¶ 31; Ex. 8.
In total, the three accused products generated $264.9 million in revenue for Lumicore in fiscal year 2023 alone. Ex. 1 (Keating Decl.) ¶ 32. Over the projected infringement period from the third quarter of 2022 through the fourth quarter of 2025, Dr. Keating estimates total accused-product revenue of approximately $714.6 million. Id. ¶ 33. Lumicore's fiscal year 2023 total revenue was approximately $312.6 million (per its 10-K filing), meaning that accused-product revenue of $264.9 million constituted approximately 84.6% of Lumicore's total revenue — underscoring that the accused transceiver modules are Lumicore's core business, not a minor product line. Id. ¶ 34.
Each accused product is a self-contained optoelectronic device in a standard pluggable form factor (QSFP-DD or OSFP) that plugs into a network switch or router. Ex. 2 (Ashford Decl.) ¶ 33. The transceiver module is not a sub-component of a larger Lumicore product — Lumicore does not manufacture switches, routers, or any other system into which the modules are incorporated. Id. ¶ 34. Rather, Lumicore sells the modules as finished goods directly to hyperscale data-center operators and, in some cases, to original equipment manufacturers ("OEMs") who integrate them into networking equipment. Id.
Critically, each accused transceiver module incorporates a photonic integrated circuit fabricated on a silicon substrate. Ex. 2 (Ashford Decl.) ¶ 35. The patented waveguide multiplexer ('847), coherent transceiver architecture ('193), micro-channel heat sink ('562), and polarization-diverse grating coupler ('015) are all fabricated within the PIC or integrated into the module housing as part of a monolithic or hybrid-integrated assembly. Id. ¶¶ 36–39. There is no separately sold sub-component corresponding to any individual patent — the PIC and surrounding module are manufactured and sold as a single, indivisible unit. Id. ¶ 40. The transceiver module is thus the smallest salable patent-practicing unit. Id. ¶ 41; Ex. 1 (Keating Decl.) ¶ 38.
D. The Patented Features Drive Customer Demand for the Accused Modules
The record evidence overwhelmingly demonstrates that the patented features are the primary basis upon which customers purchase the accused transceiver modules.
Independent Teardown Analysis. Benchmark Semiconductor Analysis, LLC, an independent semiconductor reverse-engineering firm located at 2244 Mission College Boulevard, Santa Clara, California 95054, conducted physical teardowns of all three accused products under the direction of Kevin Ng, Lead Engineer. Ex. 5 (Benchmark Teardown Report Excerpts). The LumiLink 400G-ZR+ was torn down on December 12, 2022; the LumiLink 800G-DR8 in June 2023; and the LumiEdge 1600G-LR in February 2024. Id. These teardowns confirmed that the patented structures — the multi-channel waveguide multiplexer, coherent transceiver circuits, embedded micro-channel heat sink, and polarization-diverse grating coupler — are central to each module's architecture and cannot be physically separated from the PIC without destroying the device. Id. ¶¶ 8–15.
Lumicore's Marketing Materials. Lumicore's own marketing materials for all three accused products prominently feature the precise capabilities enabled by the patented technologies. Ex. 6 (Lumicore Marketing Materials). Product brochures for the LumiLink 400G-ZR+ advertise "ultra-low insertion loss multiplexing" — the capability enabled by the '847 Patent. Id. at OWS0014523. Materials for the LumiLink 800G-DR8 tout a "fully integrated coherent transceiver on silicon" — the architecture claimed in the '193 Patent — and "advanced thermal management for high-power density" — the innovation covered by the '562 Patent. Id. at OWS0014587. The LumiEdge 1600G-LR brochure highlights a "polarization-independent grating coupler" — the structure disclosed and claimed in the '015 Patent. Id. at OWS0014612. These are not incidental specifications buried in fine-print datasheets; they are the headline capabilities Lumicore uses to compete for hyperscale data-center business. Ex. 3 (Falk Decl.) ¶ 22.
Lumicore CTO Deposition Testimony. Dr. Sandra Michaud, Lumicore's Chief Technology Officer, testified at her deposition on November 19, 2024, that Lumicore's LumiLink 400G-ZR+ "was designed around the need for a low-loss silicon waveguide multiplexer and a thermally robust PIC." Ex. 4 (Michaud Dep. Tr.) at 87:14–19. Dr. Michaud further admitted that during the design phase, she and her team reviewed OWS's published patent application No. 2018/0123456 (corresponding to the '193 Patent). Id. at 112:4–16. When asked to identify specific non-infringing design alternatives that would achieve comparable performance to the accused products, Dr. Michaud was unable to do so. Id. at 134:7–135:22. She also testified that Lumicore estimated a redesign around the patented technologies would require approximately 18 to 24 months — testimony consistent with Lumicore's own injunction-impact assessment produced in discovery. Id. at 141:3–18; see also Dkt. 187-12 (Lumicore Injunction Impact Assessment) at LTC0034891.
Technical Expert Opinion. Dr. Caroline Ashford, Professor of Electrical Engineering and Photonics at the Northfield Institute of Technology in Evanston, Illinois, and OWS's technical expert, opines in her declaration that the patented features are not ancillary add-ons but rather fundamental architectural elements without which the accused transceiver modules could not achieve their advertised performance specifications. Ex. 2 (Ashford Decl.) ¶¶ 42–51. Dr. Ashford explains that the waveguide multiplexer determines the module's channel count and optical efficiency; the coherent transceiver architecture determines its data rate and modulation format; the thermal management subsystem determines its reliability and power-handling capability; and the grating coupler determines its coupling efficiency and polarization robustness. Id. ¶¶ 44–47. Remove any one of these subsystems, and the module cannot function at its rated specifications. Id. ¶ 48.
Industry Expert Opinion. Gregory Falk, an independent consultant and formerly Vice President of Engineering at Solara Optics, Inc., confirms in his declaration that customer purchasing decisions for silicon-photonics transceivers are driven primarily by optical performance metrics — insertion loss, bandwidth, power efficiency, and polarization handling — the very metrics addressed by the four asserted patents. Ex. 3 (Falk Decl.) ¶¶ 18–25. Mr. Falk explains that hyperscale data-center operators evaluate transceiver modules based on a set of key performance indicators ("KPIs") that map directly onto the capabilities enabled by OWS's patents: link budget (driven by insertion loss, addressed by the '847 Patent), data rate and reach (driven by coherent detection, addressed by the '193 Patent), thermal reliability (addressed by the '562 Patent), and coupling efficiency (addressed by the '015 Patent). Id. ¶ 20.
Mr. Falk's competitive analysis identifies seven major hyperscale customers that purchase from both OWS and Lumicore, and confirms that OWS and Lumicore compete head-to-head at 12 of the top 20 hyperscale data-center operators. Id. ¶¶ 26–28. The competition between the two companies centers on precisely the optical performance and thermal management metrics enabled by the patented inventions. Id. ¶ 29.
Price Erosion Data. The demand-driving nature of the patented features is further confirmed by price erosion data. Average selling prices for 400G-ZR transceiver modules declined approximately 18.7% from the third quarter of 2022 to the fourth quarter of 2023, attributable in significant part to Lumicore's entry into the market with products employing the patented technologies. Ex. 9 (Price Erosion Analysis); Ex. 1 (Keating Decl.) ¶ 47. This price erosion is direct evidence that Lumicore's accused products — leveraging the patented innovations — have competitively displaced OWS's products in the market.
E. Dr. Keating's Damages Methodology
Dr. Lawrence Keating, Managing Director of Meridian Economic Consulting Group, LLC, located at 233 South Wacker Drive, Suite 4400, Chicago, Illinois 60606, has prepared a comprehensive damages analysis presenting two damages theories. Ex. 1 (Keating Decl.) ¶¶ 50–95.
Theory 1 — Lost Profits (Panduit). Dr. Keating estimates OWS's lost profits at approximately $128.9 million, calculated as follows: projected accused-product revenue over the infringement period of approximately $714.6 million, multiplied by a 38.2% market-share displacement factor (reflecting the portion of Lumicore's sales that OWS would have captured but for the infringement), multiplied by OWS's 47.3% gross margin on competing products (the OWS-400ZR and OWS-800XR product families). Ex. 1 (Keating Decl.) ¶¶ 55–68. The 38.2% market-share displacement figure is derived from customer overlap data (12 of the top 20 hyperscale operators purchase from both OWS and Lumicore), adjusted downward for customers at which OWS is not an approved vendor. Id. ¶ 62. Dr. Keating also seeks enhanced damages for willful infringement: post-notice lost profits (calculated from January 15, 2023, the date of the notice letter sent by FairPoint Licensing Group on behalf of OWS) of approximately $109.2 million, with enhancement sought pursuant to 35 U.S.C. § 284. Id. ¶¶ 69–72.
Theory 2 — Reasonable Royalty (Georgia-Pacific). For the portion of accused-product sales not subject to the lost-profits theory — approximately $441.6 million in revenue — Dr. Keating applies a reasonable royalty rate of 8.5%, yielding approximately $37.5 million in royalty damages. Ex. 1 (Keating Decl.) ¶¶ 73–90. Dr. Keating conducted a comprehensive analysis of all 15 Georgia-Pacific factors and identified five comparable license agreements that support the 8.5% royalty rate:
(1) OWS-Terralink Semiconductor license (2020): 6.5% running royalty on $45 million in licensed revenue, applied at the module level. Ex. 7 (Comparable License Agreements) at OWS-LIC-00001.
(2) OWS-NovaWave Corp. license (2021): 9.0% running royalty on $22 million in licensed revenue, applied at the module level. Id. at OWS-LIC-00047.
(3) OWS-Helix Integrated Photonics license (2022): 7.75% running royalty plus a $2.5 million upfront payment on $68 million in licensed revenue, applied at the module level. Id. at OWS-LIC-00098.
(4) Stratos Photonics v. ArcLight Systems settlement (2019): 8.0% royalty applied at the module level. Id. at OWS-LIC-00156.
(5) Coreline Technologies v. Prisma Semiconductor cross-license (2021): 10.2% effective royalty rate, applied at the module level. Id. at OWS-LIC-00201.
Ex. 1 (Keating Decl.) ¶¶ 78–84. All five comparable licenses apply royalty rates to module-level revenue — not to any disaggregated sub-component value — confirming that the industry standard practice for licensing silicon-photonics patents is to apply the royalty at the transceiver-module level. Id. ¶ 85. The hypothetical negotiation date for purposes of Dr. Keating's analysis is the second quarter of 2022, immediately preceding the launch of Lumicore's first accused product. Id. ¶ 74.
For both theories, Dr. Keating uses the module-level price as the royalty base. Id. ¶ 91. He does so because the transceiver module is the SSPPU — there is no smaller commercially available sub-unit that practices the asserted patents — and because the patented features are the predominant driver of demand for the entire module. Id. ¶¶ 92–95.
F. Lumicore's Financial Data Confirms Module-Level Pricing
Lumicore's product-level financial data, produced in discovery and designated "Highly Confidential — Attorneys' Eyes Only" under the Court's three-tier confidentiality order, further corroborates that the transceiver module is the relevant commercial unit. Ex. 8 (Lumicore Financial Data Excerpts). Lumicore's internal records show accused-product gross margins ranging from 52.1% for the LumiLink 400G-ZR+ to 58.3% for the LumiEdge 1600G-LR. Id. Revenue, cost of goods sold, average selling prices, and gross margins are all reported at the module level. Id. Lumicore does not separately price or sell the PIC, waveguide multiplexer, thermal management subsystem, or grating coupler components — because these components have no independent commercial existence. Id.; Ex. 1 (Keating Decl.) ¶ 40. Average selling prices are set and tracked at the transceiver-module level, confirming that the module is the unit of commerce. Ex. 8.
III. LEGAL STANDARD
A. Summary Judgment Standard Under Federal Rule of Civil Procedure 56
Summary judgment is appropriate "if the movant shows that there is no genuine dispute as to any material fact and the movant is entitled to judgment as a matter of law." Fed. R. Civ. P. 56(a). A dispute is "genuine" if "the evidence is such that a reasonable jury could return a verdict for the nonmoving party." Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 248 (1986). A fact is "material" if it "might affect the outcome of the suit under the governing law." Id. The moving party bears the initial burden of demonstrating the absence of a genuine dispute of material fact. Celotex Corp. v. Catrett, 477 U.S. 317, 323 (1986).
In evaluating a motion for summary judgment, the Court must view the evidence in the light most favorable to the nonmoving party and draw all reasonable inferences in that party's favor. Matsushita Elec. Indus. Co. v. Zenith Radio Corp., 475 U.S. 574, 587 (1986). The Court does not weigh evidence or make credibility determinations at the summary judgment stage — those functions are reserved for the jury. Anderson, 477 U.S. at 255. Where, as here, the issues involve competing expert opinions and factual questions about customer demand, market dynamics, and the proper scope of damages, summary judgment is particularly disfavored. See id. at 249–50 (summary judgment inappropriate where reasonable minds could differ on the inferences to be drawn from the evidence).
B. The Entire Market Value Rule — Governing Legal Framework
Under 35 U.S.C. § 284, a patentee is entitled to "damages adequate to compensate for the infringement, but in no event less than a reasonable royalty." The Federal Circuit has developed two interrelated doctrines to ensure that damages awards are properly tied to the value of the patented invention, as opposed to unpatented features: the SSPPU doctrine and the EMVR.
The smallest salable patent-practicing unit ("SSPPU") doctrine requires that, as a general rule, the royalty base for a reasonable-royalty damages calculation should be the smallest salable unit that practices the patented invention, rather than the price of a larger product that incorporates both patented and unpatented features. LaserDynamics, Inc. v. Quanta Computer, Inc., 694 F.3d 51, 67 (Fed. Cir. 2012). The purpose of this doctrine is to prevent the damages calculation from being based on the value of features that are unrelated to the patent. Id. at 68.
The Entire Market Value Rule is a corollary to the apportionment requirement. It provides an exception: a patentee may base its damages on the entire market value of the accused product — even when the product contains unpatented features — where "the patent-related feature is the basis for customer demand." Rite-Hite Corp. v. Kelley, Inc., 56 F.3d 1538, 1549 (Fed. Cir. 1995) (en banc). The EMVR thus "permits 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.; see also Commonwealth Sci. & Indus. Research Organisation v. Cisco Sys., Inc., 809 F.3d 1295, 1302 (Fed. Cir. 2015).
The EMVR is a limitation on damages — it prevents overcompensation by ensuring that the royalty base reflects the value attributable to the patented invention. But it is not an absolute bar on using the price of a multi-feature product as the royalty base. When the patented feature is the predominant driver of customer demand, the entire product price is an appropriate base because the patent accounts for the product's commercial value. See VirnetX, Inc. v. Cisco Sys., Inc., 767 F.3d 1308, 1326 (Fed. Cir. 2014); Lucent Techs., Inc. v. Gateway, Inc., 580 F.3d 1301, 1336 (Fed. Cir. 2009); Uniloc USA, Inc. v. Kestrel Partners LLC, 632 F.3d 1292, 1318 (Fed. Cir. 2011). As the Federal Circuit explained in Ericsson, Inc. v. D-Link Systems, Inc., the critical question is whether "the patented feature drives the demand for an entire multi-component product." 773 F.3d 1201, 1226 (Fed. Cir. 2014).
The Federal Circuit has cautioned that applying a royalty to the entire product price creates "a risk of skewing the damages horizon" when the patented feature is "but a tiny component of the product." LaserDynamics, 694 F.3d at 67. But where the patented technology is central to the product's function and is the reason customers purchase the product, there is no such risk. See Finjan, Inc. v. Blue Coat Sys., Inc., 879 F.3d 1299, 1310–11 (Fed. Cir. 2018) (affirming entire-product royalty base where patented security technology was the "reason the customer purchases" the accused product). Indeed, requiring artificial apportionment below the SSPPU level when no smaller commercial unit exists would be to mandate speculation rather than precision — the opposite of what the apportionment doctrine demands. See 7 Donald S. Chisum, Chisum on Patents § 20.03 (2024).
C. Apportionment Is Not Required Below the SSPPU When the SSPPU Is the Product Sold
When the accused product itself is the smallest salable patent-practicing unit — that is, when there is no smaller commercially available sub-unit that practices the patent — the product-level price is the proper royalty base, and no further apportionment of the royalty base is required. See VLSI Tech. LLC v. Bramwell Partners LLC, 87 F.4th 1332, 1351–52 (Fed. Cir. 2024); Power Integrations, Inc. v. Fairchild Semiconductor Int'l, Inc., 904 F.3d 965, 977 (Fed. Cir. 2018). In such circumstances, the EMVR limitation is inapplicable because the royalty base already reflects the SSPPU — there is no "entire market" to which the patentee is improperly reaching. Apportionment, to the extent necessary, is accomplished through the royalty rate, not by artificially disaggregating the royalty base into hypothetical sub-components that have no independent commercial existence.
The burden at summary judgment rests squarely on the movant. Lumicore must demonstrate that there is no genuine dispute of material fact and that, as a matter of law, Dr. Keating's methodology is impermissible. If there is any genuine factual dispute about whether the patented features drive demand for the accused products, or about whether the module is the SSPPU, the question must go to the jury. See Anderson, 477 U.S. at 248.
IV. ARGUMENT
A. The Transceiver Module Is the Smallest Salable Patent-Practicing Unit
1. The Module Is a Single Integrated Product, Not an Assembly of Separable Sub-Components.  The fundamental predicate of Lumicore's anticipated motion — that Dr. Keating should have applied his royalty rate to some sub-component of the transceiver module rather than to the module itself — fails at the threshold because no such sub-component exists as a separately salable product. Each accused transceiver module is sold as a single pluggable unit in a standard QSFP-DD or OSFP form factor. Ex. 2 (Ashford Decl.) ¶ 33. The photonic integrated circuit within each module is fabricated as a monolithic or hybrid-integrated chip on a silicon substrate. Id. ¶ 35. The patented waveguide multiplexer ('847 Patent), coherent transceiver architecture ('193 Patent), micro-channel heat sink ('562 Patent), and polarization-diverse grating coupler ('015 Patent) are physically integrated into the PIC and module package — they are not discrete components that can be separated, replaced, or sold independently. Id. ¶¶ 36–40.
The Benchmark Semiconductor Analysis teardowns of all three accused products confirm this physical integration. Ex. 5 (Benchmark Teardown Report Excerpts). Kevin Ng, Lead Engineer at Benchmark, concluded that "the waveguide multiplexer structures are fabricated as part of the silicon photonic die and cannot be physically separated without destroying the device." Id. ¶ 10. The same is true of the coherent transceiver circuits, the micro-channel thermal management structures, and the grating coupler — all are integrated at the chip or package level and have no independent physical or commercial existence. Id. ¶¶ 11–14.
Dr. Caroline Ashford corroborates this conclusion from a technical perspective. She explains that the architecture of a silicon-photonics transceiver module is fundamentally different from, for example, a laptop computer containing a discrete disk drive. Ex. 2 (Ashford Decl.) ¶ 42. In a laptop, individual components (the disk drive, the display, the keyboard) can be separately sourced, replaced, and sold. Id. In a silicon-photonics transceiver module, by contrast, the optical and electronic subsystems are fabricated on and into the same semiconductor substrate using lithographic and deposition processes — they cannot be disaggregated into separately functional components any more than a transistor can be separated from the integrated circuit of which it forms a part. Id. ¶ 43.
There is no separate commercial market for the waveguide multiplexer, the thermal management subsystem, the coherent transceiver circuit, or the grating coupler standing alone. Ex. 3 (Falk Decl.) ¶¶ 30–33. These elements are not independently priced, sold, or marketed by Lumicore, by OWS, or by any other market participant. Id. Mr. Falk, who has more than 25 years of experience in the silicon-photonics industry, confirms that "no company in the silicon-photonics transceiver market sells waveguide multiplexers, grating couplers, or embedded thermal management subsystems as standalone commercial products. These are integrated functional blocks within a monolithic or hybrid-integrated photonic chip, not separately available components." Id. ¶ 32.
2. Lumicore's Own Sales and Accounting Practices Confirm the Module Is the Relevant Unit.  Lumicore's own business records confirm that the transceiver module is the unit of commerce. Lumicore sells transceiver modules as complete, finished products. Ex. 8 (Lumicore Financial Data Excerpts). Its revenue reporting, pricing, cost-of-goods-sold accounting, and margin analysis are all conducted at the module level. Id. Lumicore does not maintain separate revenue line items, pricing schedules, or cost breakdowns for PICs, waveguide multiplexers, thermal management subsystems, or grating couplers — because these are not separate products that Lumicore offers for sale. Id.
Lumicore's product datasheets, customer purchase orders, and invoices — all produced in discovery — uniformly reference the transceiver module as the unit of sale. See Dkt. 187-5 (Sample Lumicore Purchase Orders); Dkt. 187-6 (Sample Lumicore Invoices). Customers order "LumiLink 400G-ZR+ Transceiver Module," not "LumiLink Waveguide Multiplexer" or "LumiLink Thermal Management Subsystem." Id. Average selling prices reported in Lumicore's internal financial documents are module-level prices. Ex. 8. Lumicore's financial data, produced under the "Attorneys' Eyes Only" designation, reports gross margins ranging from 52.1% to 58.3% — all calculated at the module level. Id.
This evidence is precisely the type of commercial reality that the SSPPU doctrine is designed to capture. The SSPPU is not an abstract theoretical construct — it is the smallest unit that is actually sold in commerce and that practices the asserted patents. Here, that unit is the transceiver module.
3. There Is No Smaller Salable Unit to Which Apportionment Could Be Applied.  If the SSPPU is the transceiver module — and the undisputed evidence shows that it is — then the EMVR limitation is inapplicable. The EMVR concern arises when a patentee seeks to apply a royalty to a product that is larger than the SSPPU, thereby capturing value attributable to unpatented features. See LaserDynamics, 694 F.3d at 67. That concern is absent here. OWS is not seeking to apply a royalty to the price of an entire data-center switch or router in which the transceiver module is installed. OWS is seeking to apply the royalty to the module itself — the SSPPU.
Lumicore cannot manufacture a legal fiction of a "sub-component" that does not exist in commerce simply to artificially reduce the damages base. The Federal Circuit has repeatedly held that the SSPPU must be an actual product sold in the marketplace. See VLSI Tech., 87 F.4th at 1351. A hypothetical "value allocation" to non-existent sub-components is not a proper SSPPU — it is an exercise in speculation that the apportionment doctrine was designed to prevent, not require.
To the extent any apportionment is warranted to account for the relative contribution of patented versus unpatented features within the module, that apportionment is properly accomplished through the royalty rate, not through an artificial reduction of the royalty base. Dr. Keating's 8.5% royalty rate already reflects this apportionment — it is derived from comparable licenses and a comprehensive Georgia-Pacific analysis that accounts for the full range of factors bearing on the value of the patented technology relative to the overall product. Ex. 1 (Keating Decl.) ¶¶ 73–90. Requiring a further reduction of the royalty base below the SSPPU would result in impermissible double-counting of apportionment — once in the base and again in the rate — producing a damages figure that is inadequate to compensate for the infringement in violation of 35 U.S.C. § 284.
B. Even If the EMVR Applied, the Patented Features Are the Basis for Customer Demand
Even assuming, arguendo, that the Court applies the EMVR framework to assess whether Dr. Keating's use of the module-level price is permissible, the record evidence overwhelmingly satisfies the demand-driving requirement. The patented technologies are not incidental add-ons to the accused products; they are the core innovations that define what the products are and why customers buy them.
1. The Patented Technologies Define the Core Performance Characteristics.  The four asserted patents, taken individually and collectively, cover the fundamental optoelectronic capabilities that make the accused transceiver modules commercially valuable.
The '847 Patent (waveguide multiplexer). This patent covers a multi-channel silicon photonic waveguide multiplexer with low insertion loss. Ex. 2 (Ashford Decl.) ¶¶ 13–17. Wavelength-division multiplexing — combining multiple data channels onto different wavelengths of light and then separating them at the receiver — is the fundamental function of an optical transceiver module. Without the multiplexer, there is no multi-channel transceiver. And without low insertion loss in the multiplexer, the transceiver's link budget is consumed by internal optical losses, degrading performance to the point of commercial unviability. Id. ¶ 17. The '847 Patent's innovation is not a nice-to-have feature — it is the enabling technology for the module's core function.
The '193 Patent (coherent transceiver on silicon). This patent covers the integration of a high-bandwidth coherent optical transceiver on a silicon substrate. Ex. 2 (Ashford Decl.) ¶¶ 18–22. Coherent detection — using both the amplitude and phase of the optical signal to encode data — is the defining technological capability that separates the accused products from legacy direct-detect transceiver modules. Coherent detection enables dramatically higher data rates (400G, 800G, 1600G) over longer distances than direct detection. Id. ¶ 22. The '193 Patent's coherent transceiver architecture is not a peripheral feature of the accused modules — it is the very technology that makes 400G-ZR+, 800G-DR8, and 1600G-LR operation possible.
The '562 Patent (micro-channel heat sink). This patent covers a method for thermal management of photonic integrated circuits using embedded micro-channel heat sinks. Ex. 2 (Ashford Decl.) ¶¶ 23–26. High-speed coherent transceivers generate significant heat within the PIC — heat that, if not effectively managed, causes wavelength drift in the laser source, increased bit-error rates, and premature device failure. Id. ¶ 25. The '562 Patent's micro-channel heat sink technology is what enables the accused modules to operate at their rated power levels without overheating. Without it, the modules would be commercially unviable — customers would not purchase transceiver modules that overheat and fail in operation. Id. ¶ 26.
The '015 Patent (polarization-diverse grating coupler). This patent covers a polarization-diverse grating coupler for silicon photonic transceivers. Ex. 2 (Ashford Decl.) ¶¶ 27–30. In a data-center fiber-optic link, the polarization state of the light arriving at the receiver varies unpredictably over time due to environmental factors such as fiber stress, bending, and temperature changes. Id. ¶ 28. A conventional grating coupler is polarization-sensitive, meaning that its coupling efficiency varies with the polarization state — an unacceptable performance limitation for a data-center transceiver that must operate reliably under all conditions. Id. ¶ 29. The '015 Patent's polarization-diverse coupler eliminates this limitation, enabling efficient and reliable coupling regardless of polarization state. Id. ¶ 30.
Together, these four patents cover the optical signal path (multiplexing), the detection architecture (coherent transceiver), the thermal stability (micro-channel heat sink), and the fiber-to-chip interface (grating coupler). Ex. 2 (Ashford Decl.) ¶ 48. As Dr. Ashford explains, these are not four isolated improvements to four separate sub-components — they are four tightly integrated innovations that collectively define the core optoelectronic functionality of the transceiver module. Id. Remove any one of them, and the module cannot achieve its rated specifications; remove all four, and there is no commercially viable product. Id. ¶ 49.
2. Lumicore's Marketing Materials Confirm the Patented Features Drive Demand.  If there were any doubt that the patented features drive customer demand, Lumicore's own marketing materials resolve it. Lumicore does not bury the capabilities enabled by OWS's patents in appendices or technical footnotes. Instead, Lumicore places them front and center in its advertising, using them as the primary competitive differentiators to win hyperscale data-center business.
The LumiLink 400G-ZR+ product brochure features the headline: "Ultra-Low Insertion Loss Multiplexing for Maximum Link Budget" — a direct reference to the capability enabled by the '847 Patent. Ex. 6 at OWS0014523. The same brochure advertises "Silicon-Integrated Coherent Detection" as a key feature — the capability enabled by the '193 Patent. Id. at OWS0014524.
The LumiLink 800G-DR8 product brochure prominently advertises a "Fully Integrated Coherent Transceiver on Silicon" and "Advanced Thermal Management for High-Power Density" — capabilities enabled by the '193 and '562 Patents, respectively. Id. at OWS0014587–88. The brochure also touts "Polarization-Independent Grating Coupler for Maximum Coupling Efficiency" — the capability enabled by the '015 Patent. Id. at OWS0014589.
The LumiEdge 1600G-LR product brochure similarly highlights "Next-Generation Coherent Silicon Photonics" and "Polarization-Diverse Fiber-Chip Interface" — capabilities enabled by the '193 and '015 Patents. Id. at OWS0014612–13.
These marketing materials are powerful evidence that the patented features are not incidental to the accused products' commercial success — they are the commercial appeal. As the Federal Circuit has recognized, a defendant's marketing materials highlighting the patented feature constitute probative evidence that the feature drives customer demand. See Finjan, 879 F.3d at 1311. Here, the alignment between the patented features and the headline advertising claims is essentially one-to-one.
3. Customer Purchasing Decisions Are Driven by the Patented Features.  Beyond Lumicore's self-serving marketing representations, the objective evidence of customer purchasing behavior confirms that the patented features are the basis for demand.
Gregory Falk's competitive analysis identifies seven major hyperscale customers that purchase transceiver modules from both OWS and Lumicore, and confirms that OWS and Lumicore compete head-to-head at 12 of the top 20 hyperscale data-center operators. Ex. 3 (Falk Decl.) ¶¶ 26–28. Mr. Falk explains that at these customers, purchasing decisions are made through formal qualification and evaluation processes in which the key selection criteria are optical performance (insertion loss, bandwidth, spectral efficiency), thermal reliability (operating temperature range, mean time between failures), and coupling efficiency (fiber-to-chip loss, polarization sensitivity). Id. ¶ 29. These are precisely the performance metrics addressed by the four asserted patents. Id. ¶ 30.
Customer declarations anticipated from Karen Whitfield, OWS's Vice President of Sales, and OWS's regional sales directors will attest to specific competitive encounters in which customers cited superior optical performance and thermal management as reasons for selecting Lumicore's accused products over OWS's competing offerings and over other alternatives. Id. ¶ 31; Ex. 1 (Keating Decl.) ¶ 46. These are not abstract assertions — they are evidence of the actual purchasing decisions of the specific customers who buy the accused products.
The price erosion data further corroborates the demand-driving nature of the patented features. Average selling prices for 400G-ZR transceiver modules declined 18.7% from the third quarter of 2022 to the fourth quarter of 2023. Ex. 9 (Price Erosion Analysis). This price decline is attributable in significant part to Lumicore's entry into the market with products that employ the patented technologies to achieve competitive — and in some cases superior — optical performance. Ex. 1 (Keating Decl.) ¶ 47. If the patented features were merely incidental to the accused products, Lumicore's market entry would not have produced the dramatic price erosion that OWS has experienced. The price erosion is direct, quantifiable evidence that the patented features are driving competitive dynamics — and hence customer demand — in the silicon-photonics transceiver market.
4. Lumicore's CTO Testimony Confirms the Centrality of the Patented Features.  Lumicore's own Chief Technology Officer has effectively conceded that the patented features are central to the accused products' design and commercial viability. At her deposition on November 19, 2024, Dr. Sandra Michaud testified that the LumiLink 400G-ZR+ was "designed around the need for a low-loss silicon waveguide multiplexer and a thermally robust PIC." Ex. 4 (Michaud Dep. Tr.) at 87:14–19. This testimony is a direct admission that the capabilities claimed in the '847 Patent (low-loss waveguide multiplexer) and the '562 Patent (thermal management) were not afterthoughts — they were the design objectives that drove the product's development.
Dr. Michaud further admitted that during the design phase for the LumiLink product family, she and her engineering team reviewed OWS's published patent application No. 2018/0123456, which corresponds to the '193 Patent. Id. at 112:4–16. This admission is significant not only for willfulness purposes but also because it demonstrates that Lumicore was aware of OWS's patented technology and affirmatively looked to it during product development — further evidence that the patented features were central to Lumicore's product design.
When pressed to identify specific non-infringing design alternatives that would achieve performance comparable to the accused products, Dr. Michaud was unable to do so. Id. at 134:7–135:22. Her testimony on this point was revealing:
Q: Can you identify a specific design alternative for the waveguide multiplexer used in the LumiLink 400G-ZR+ that does not practice the claims of the '847 Patent and that achieves comparable insertion loss?
A: Not off the top of my head. It would require significant engineering work.
Q: Have you or anyone at Lumicore attempted such a redesign?
A: We have looked at alternatives, but nothing that achieves comparable insertion loss on a silicon platform.
Id. at 134:7–22. This inability to identify non-infringing alternatives that achieve comparable performance is compelling evidence that the patented features are not merely one of many possible design choices but are essential to the products' competitive position.
Moreover, Dr. Michaud testified that Lumicore estimated a redesign around the patented technologies would require approximately 18 to 24 months. Id. at 141:3–18. This extended redesign timeline — nearly two years — reflects how deeply embedded the patented features are in the product architecture. If the patented features were truly peripheral or easily separable, one would expect a redesign to be achievable in a matter of weeks, not years. The 18-to-24-month timeline is itself evidence that the patented features are the architectural foundation of the accused products — exactly the type of integral, demand-driving technology for which module-level damages are appropriate.
5. Industry Licensing Practice Supports a Module-Level Royalty Base.  The EMVR analysis does not occur in an economic vacuum. The way the industry actually licenses silicon-photonics patent rights provides powerful evidence of the appropriate royalty base. Here, industry practice overwhelmingly confirms that the transceiver module is the economically relevant unit for patent licensing purposes.
All five comparable license agreements identified by Dr. Keating apply royalty rates to module-level revenue — not to any hypothetical sub-component value:
(1) The OWS-Terralink Semiconductor license (executed in 2020) applies a 6.5% running royalty to $45 million in licensed revenue, calculated at the module level. Ex. 7 at OWS-LIC-00001. Terralink manufactures and sells silicon-photonics transceiver modules that practice certain claims of the '847 and '193 Patents. Id. The license expressly defines the royalty base as "Net Sales of Licensed Products," where "Licensed Products" are defined as transceiver modules. Id. at OWS-LIC-00005.
(2) The OWS-NovaWave Corp. license (executed in 2021) applies a 9.0% running royalty to $22 million in licensed revenue, calculated at the module level. Id. at OWS-LIC-00047. NovaWave's licensed products include coherent transceiver modules employing technology covered by the '193 and '562 Patents. Id. at OWS-LIC-00049.
(3) The OWS-Helix Integrated Photonics license (executed in 2022) applies a 7.75% running royalty plus a $2.5 million upfront payment to $68 million in licensed revenue, calculated at the module level. Id. at OWS-LIC-00098. This license covers all four asserted patents and is the most technologically comparable agreement. Id. at OWS-LIC-00101.
(4) The Stratos Photonics v. ArcLight Systems settlement (2019) resulted in a license with an 8.0% royalty rate applied to module-level sales. Id. at OWS-LIC-00156. Although this license arose in the context of litigation between third parties, the technologies at issue (silicon-photonics waveguide and thermal management) are closely analogous to the patents asserted here. Id. at OWS-LIC-00158.
(5) The Coreline Technologies v. Prisma Semiconductor cross-license (2021) resulted in an effective royalty rate of 10.2%, also applied at the module level. Id. at OWS-LIC-00201. The patents licensed in this agreement include coherent transceiver and polarization management technologies similar to the '193 and '015 Patents. Id. at OWS-LIC-00204.
This consistent pattern of module-level licensing is not coincidental. As Mr. Falk explains, the silicon-photonics industry licenses at the module level because the module is the unit of commerce — there is no meaningful way to license at a "sub-component" level when the sub-components are not separately sold. Ex. 3 (Falk Decl.) ¶ 35. Dr. Keating reaches the same conclusion: module-level licensing is the established industry practice because the module is the economically relevant unit and the patented features are inseparable from the module's commercial value. Ex. 1 (Keating Decl.) ¶ 85.
The Federal Circuit has recognized that comparable license agreements are among the most probative evidence of a reasonable royalty, including the appropriate royalty base. See Ericsson, 773 F.3d at 1227. Here, the unanimous evidence from comparable licenses is that the transceiver module is the appropriate royalty base. Lumicore can point to no license agreement in the silicon-photonics industry that applies a royalty to a hypothetical sub-component value rather than the module price.
C. Lumicore's Anticipated Apportionment Argument Is Legally and Factually Flawed
1. Lumicore Cannot Manufacture a Sub-Module Royalty Base That Does Not Exist in Commerce.  Lumicore is expected to argue — through its damages expert, Dr. Natasha Volkov of Pinnacle Analytics Partners LLP, Houston, Texas — that the royalty should be applied to a disaggregated "value" of the specific patented sub-components rather than to the module price. This approach fundamentally misapplies the SSPPU doctrine and contravenes Federal Circuit precedent.
The SSPPU doctrine requires that the royalty base be the "smallest salable patent-practicing unit." LaserDynamics, 694 F.3d at 67 (emphasis added). The key word is "salable" — meaning a product that is actually sold in the marketplace. The doctrine does not permit or require a party to construct a hypothetical "value" for a non-existent sub-component and then use that hypothetical value as the royalty base. To do so would be to substitute speculation for the market-based analysis that the SSPPU doctrine was designed to promote.
Here, the waveguide multiplexer, coherent transceiver circuit, micro-channel heat sink, and grating coupler are not salable units. They are not products. They are not items of commerce. They are integrated functional blocks within a monolithic semiconductor device. Ex. 2 (Ashford Decl.) ¶¶ 40–41; Ex. 3 (Falk Decl.) ¶ 32. There is no market price for a "waveguide multiplexer" because waveguide multiplexers are not bought and sold. There is no invoice, no purchase order, no price list, and no catalog entry for any of these sub-components — because they do not exist as separate commercial products.
Dr. Volkov's anticipated approach of constructing a hypothetical sub-component value is unreliable for precisely this reason. It requires the expert to engage in speculative allocation of costs, margins, and value to functional blocks that have never been priced in any arm's-length transaction. Such an allocation is not grounded in market evidence — it is an exercise in economic fiction that violates the fundamental premise of the reasonable-royalty framework: that damages should be based on what willing parties would have agreed to in a real-world negotiation informed by real-world market data. See Georgia-Pacific Corp. v. U.S. Plywood Corp., 318 F. Supp. 1116, 1120 (S.D.N.Y. 1970). No willing licensor would negotiate a royalty based on a hypothetical "value" of a non-existent sub-component when the product is sold as a module and priced as a module.
2. Lumicore's Expert's Methodology Is the Subject of a Separate Daubert Challenge.  OWS has separately prepared a motion pursuant to Federal Rule of Evidence 702 and Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993), challenging the reliability and admissibility of Dr. Volkov's damages opinions. Without previewing the full scope of that motion, OWS notes that its Daubert challenge targets, among other things, Dr. Volkov's disaggregated royalty base methodology, her failure to conduct a comprehensive analysis of all 15 Georgia-Pacific factors, and her reliance on license agreements that are not economically or technologically comparable to the hypothetical negotiation at issue in this case.
The existence of a pending Daubert challenge to Dr. Volkov's methodology is itself an independent ground for denying Lumicore's anticipated summary judgment motion. Lumicore's motion necessarily depends on the proposition that Dr. Keating's module-level approach is legally impermissible and that Dr. Volkov's disaggregated approach is the only permissible methodology. But if there is a genuine dispute about the reliability of Dr. Volkov's approach — as there manifestly is — then the parties present competing damages methodologies, and the question of which approach is correct is for the jury, not for resolution on summary judgment.
Moreover, even setting aside the Daubert challenge, the existence of competing expert opinions on the proper damages methodology is itself a genuine dispute of material fact that precludes summary judgment. See Anderson, 477 U.S. at 248–49. Dr. Keating opines, based on extensive analysis, that the module-level price is the appropriate royalty base. Dr. Volkov is expected to opine otherwise. A reasonable jury could credit either expert. That factual dispute cannot be resolved as a matter of law.
3. Dr. Keating's Methodology Is Reliable and Consistent with Federal Circuit Precedent.  Dr. Keating's damages methodology is rigorous, well-supported, and fully consistent with Federal Circuit precedent. His approach proceeds in three analytically distinct steps, each grounded in established law:
First, Dr. Keating identifies the transceiver module as the SSPPU based on the undisputed facts that (a) the patented features are integrated into the module and cannot be separately purchased, (b) Lumicore sells the module as a complete product, and (c) the module is the smallest unit in commerce that practices the asserted patents. Ex. 1 (Keating Decl.) ¶¶ 35–42.
Second, Dr. Keating confirms that the patented features drive customer demand for the module, based on (a) Lumicore's marketing materials highlighting the patented capabilities, (b) the deposition testimony of Dr. Michaud, (c) the Benchmark teardown reports showing the centrality of the patented structures, (d) competitive analysis showing that purchasing decisions are driven by the performance metrics enabled by the patents, and (e) price erosion data showing competitive displacement. Id. ¶¶ 43–49.
Third, Dr. Keating applies a reasonable royalty rate of 8.5% to the module-level price, derived from a comprehensive Georgia-Pacific analysis of all 15 factors and calibrated against five comparable license agreements — all of which apply royalties at the module level. Id. ¶¶ 73–90. The 8.5% rate reflects the full range of apportionment considerations: the value contributed by the patented features relative to unpatented features, the existing licensing landscape, the parties' relative bargaining positions, and the profitability of the accused products. Id. ¶¶ 86–90.
Dr. Keating's lost-profits analysis under Panduit Corp. v. Stahlin Bros. Fibre Works, Inc., 575 F.2d 1152 (6th Cir. 1978), similarly uses module-level pricing. OWS's 47.3% gross margin on its competing products (the OWS-400ZR and OWS-800XR families) is a module-level margin. Ex. 1 (Keating Decl.) ¶ 58. The 38.2% market-share displacement factor is based on detailed customer overlap data — 12 of the top 20 hyperscale data-center operators purchase from both OWS and Lumicore — adjusted downward for customers at which OWS is not an approved vendor and would not have captured the sale regardless of infringement. Id. ¶¶ 60–65. This is a rigorous, fact-based methodology that appropriately measures OWS's actual competitive harm at the module level — the level at which the competition occurs.
This methodology is precisely the type of analysis the Federal Circuit has endorsed. In Commonwealth Scientific, the Federal Circuit upheld a module-level royalty base where the patented technology was integral to the accused wireless chipset. 809 F.3d