1014-00004 · Hawthorne Vega & Cruz LLP · Claim Construction

ashford-supplemental-declaration-reply.docx DOCUMENT Litigation

4 / 168 ← → ✕

UNITED STATES DISTRICT COURT EASTERN DISTRICT OF TEXAS MARSHALL DIVISION

SUPPLEMENTAL DECLARATION OF DR. CAROLINE ASHFORD, PH.D., IN SUPPORT OF PLAINTIFF OPTIWAVE SEMICONDUCTOR CORP.'S REPLY CLAIM CONSTRUCTION BRIEF

(Filed Concurrently with D.E. 52 — OWS's Reply Claim Construction Brief)

I. INTRODUCTION AND IDENTIFICATION OF DECLARANT

1. My name is Dr. Caroline Ashford. I am a Professor of Electrical Engineering & Photonics at the Northfield Institute of Technology in Evanston, Illinois, where I have held a tenured faculty appointment since 2009. I am over the age of twenty-one and am otherwise competent to make this declaration. I make this declaration based upon my personal knowledge and my professional opinions, formed as described herein.

2. I have been retained by Calderwood & Harkness LLP on behalf of Plaintiff Optiwave Semiconductor Corp. ("OWS" or "Plaintiff") as a technical and infringement expert in the above-captioned matter. In that capacity, I previously submitted a declaration in support of OWS's Opening Claim Construction Brief on November 15, 2024, which was filed as part of Docket Entry 44 (hereinafter, "my November 15 Declaration" or "my Original Declaration").

3. I submit this Supplemental Declaration in support of OWS's Reply Claim Construction Brief, filed concurrently herewith on December 20, 2024. The purpose of this Supplemental Declaration is to respond to certain opinions and arguments set forth in the declaration of Dr. Rajesh Sundaram, which was submitted in support of Defendant Lumicore Technologies, Inc.'s ("Lumicore" or "Defendant") Responsive Claim Construction Brief filed on December 6, 2024 (D.E. 48). This Supplemental Declaration supplements, but does not replace, my November 15 Declaration, and the opinions expressed in my Original Declaration remain unchanged and are incorporated herein by reference.

4. I make this declaration under penalty of perjury pursuant to 28 U.S.C. § 1746.

5. I have been compensated at a rate of $950 per hour for my work in this matter, including the time spent reviewing materials, forming opinions, and preparing both my Original Declaration and this Supplemental Declaration. My compensation is not contingent upon the outcome of this litigation or the substance of my opinions. My opinions are based solely on my professional judgment, my review of the relevant technical materials, and my expertise in the field of silicon photonics.

II. QUALIFICATIONS AND EXPERTISE

6. My qualifications are set forth in detail in my curriculum vitae, which was attached as Exhibit A to my November 15 Declaration (D.E. 44) and is incorporated herein by reference. I provide a brief summary of my relevant qualifications below.

7. I hold a Ph.D. in Electrical Engineering from the Massachusetts Institute of Technology, where my doctoral research focused on integrated photonics, specifically the design and characterization of silicon photonic waveguide structures for wavelength-division multiplexing applications. I also hold an M.S. in Electrical Engineering from Stanford University and a B.S. in Physics from the University of Michigan, Ann Arbor.

8. I have served as a Professor of Electrical Engineering & Photonics at the Northfield Institute of Technology since 2009. My research specializes in several areas directly relevant to the patents-in-suit, including silicon photonic waveguide design, optical transceiver architectures, thermal management of photonic integrated circuits ("PICs"), and grating coupler technologies. I have directed a research laboratory — the Northfield Integrated Photonics Laboratory — that has received over $18 million in external research funding from the National Science Foundation, the Defense Advanced Research Projects Agency, and private-sector sponsors during my tenure.

9. I have over twenty years of experience in the field of silicon photonics. Prior to my academic appointment, I worked as a senior design engineer at Granite Harbor Enterprises from 2003 to 2009, where I contributed to the design of photonic integrated circuits for long-haul optical communication systems.

10. I have published extensively in peer-reviewed journals, including IEEE Photonics Technology Letters, Journal of Lightwave Technology, Optics Express, Nature Photonics, and Optica. I have authored or co-authored over 140 peer-reviewed journal articles and conference proceedings papers relating to silicon photonics and related topics.

11. I am a named inventor on seven United States patents in the field of integrated photonic devices, covering innovations in grating coupler design, waveguide thermal tuning, and photonic transceiver architectures.

12. I have previously been qualified as an expert witness in patent infringement cases involving silicon photonics and optical communications technologies in the United States District Courts for the Northern District of California, the District of Delaware, and the Western District of Texas. In total, I have been retained as a technical expert in eight patent cases and have provided deposition testimony in six of those matters and trial testimony in two.

13. I am familiar with the four patents asserted by OWS in this litigation: U.S. Patent No. 10,234,847 ("the '847 Patent"), U.S. Patent No. 10,891,193 ("the '193 Patent"), U.S. Patent No. 9,678,562 ("the '562 Patent"), and U.S. Patent No. 11,402,015 ("the '015 Patent"). I have reviewed the complete specifications, claims, figures, and prosecution histories of each of these patents, as well as the prior art of record, in forming my opinions.

III. PURPOSE AND SCOPE OF SUPPLEMENTAL DECLARATION

14. This Supplemental Declaration responds specifically to opinions and arguments set forth in the declaration and report of Dr. Rajesh Sundaram, Professor of Electrical & Computer Engineering at Pacific Western University, San Diego, California, which was submitted in support of Lumicore's Responsive Claim Construction Brief filed December 6, 2024 (D.E. 48) (hereinafter, "Dr. Sundaram's Declaration" or "the Sundaram Declaration").

15. In particular, this Supplemental Declaration addresses three categories of issues:

(a) Dr. Sundaram's proposed definition of a person of ordinary skill in the art ("POSITA"), which is improperly narrow and results in an elevated baseline of knowledge that distorts the claim-construction analysis;

(b) Dr. Sundaram's improper reliance on extrinsic evidence — including post-filing-date journal articles, general-purpose technical dictionaries, industry whitepapers, and voluntary standards documents — to narrow the scope of disputed claim terms beyond what the intrinsic record of the asserted patents supports; and

(c) Dr. Sundaram's proposed constructions of four key disputed claim terms: (i) "waveguide multiplexer" (the '847 Patent), (ii) "integrated on a silicon substrate" (the '193 Patent), (iii) "micro-channel heat sink" (the '562 Patent), and (iv) "polarization-diverse grating coupler" (the '015 Patent).

16. This Supplemental Declaration supplements but does not replace my November 15 Declaration. All opinions expressed in my Original Declaration remain in full force, and I incorporate them by reference herein.

IV. MATERIALS REVIEWED

17. In forming the opinions expressed in this Supplemental Declaration, I reviewed the following materials:

(a) The Declaration of Dr. Rajesh Sundaram filed December 6, 2024, in support of Lumicore's Responsive Claim Construction Brief (D.E. 48);

(b) Lumicore's Responsive Claim Construction Brief (D.E. 48, filed December 6, 2024);

(c) The Joint Claim Construction Statement (D.E. 40, filed October 28, 2024), which identifies fourteen (14) disputed claim terms across the four asserted patents, of which two (2) terms have been agreed upon by the parties and twelve (12) remain in dispute;

(d) The asserted patents: U.S. Patent Nos. 10,234,847; 10,891,193; 9,678,562; and 11,402,015, including the complete specifications, all claims (independent and dependent), and all figures associated with each patent;

(e) The complete prosecution histories of all four asserted patents, including all office actions issued by the United States Patent and Trademark Office, all applicant responses and amendments, all declarations submitted during prosecution, all notices of allowance, and all records of examiner interviews;

(f) Provisional Application No. 62/499,117, filed April 7, 2017, which is the common provisional application to which the '193 Patent and the '015 Patent each claim priority;

(g) The extrinsic references cited by Dr. Sundaram in his declaration, including:

(i) Photonic Devices and Systems: A Comprehensive Reference (Predding & Knights, 3d ed. 2015);

(ii) IEEE Standard 802.3bs-2017 and associated technical supplements;

(iii) Selected passages from Silicon Photonics: An Introduction (Pavesi & Lockwood, 2d ed. 2016);

(iv) Bogaerts et al., "Silicon Microring Resonators," Tidewater Group, Vol. 6, No. 1 (2012);

(v) Chen et al., "Polarization-Diverse Grating Couplers on Silicon," Optics Express, Vol. 26, No. 12 (June 2018);

(vi) Tummidi et al., "Advanced Thermal Solutions for Silicon Photonic Transceivers," Journal of Electronic Packaging, Vol. 139, No. 4 (December 2017);

(vii) Additional references identified in the Sundaram Declaration at paragraphs 34–38 and 52–56;

(h) Benchmark Semiconductor Analysis, LLC teardown reports for the following accused products: (i) LumiLink 400G-ZR+ (teardown report dated December 12, 2022, lead analyst Kevin Ng); (ii) LumiLink 800G-DR8 (teardown report dated March 8, 2023, lead analyst Kevin Ng); and (iii) LumiEdge 1600G-LR (teardown report dated September 15, 2023, lead analyst Teresa Voss);

(i) My own prior declaration filed November 15, 2024 (D.E. 44);

(j) OWS's Opening Claim Construction Brief (D.E. 44, filed November 15, 2024);

(k) Relevant technical literature and standards in the silicon-photonics field as of each patent's effective filing date, including literature cited in my November 15 Declaration.

18. I reserve the right to supplement the materials reviewed and the opinions expressed herein as additional information becomes available in the course of this litigation.

V. RESPONSE TO DR. SUNDARAM'S PROPOSED POSITA DEFINITION

A. Dr. Sundaram's Proposed POSITA Definition

19. In paragraphs 14–18 of his declaration, Dr. Sundaram defines a person of ordinary skill in the art ("POSITA") for purposes of construing the claims of the asserted patents. Specifically, Dr. Sundaram proposes that a POSITA is:

"An individual holding a Ph.D. in electrical engineering, photonics, or physics, with at least five (5) years of direct industry experience specifically in the design and fabrication of silicon photonic integrated circuits at a commercial foundry or semiconductor company."

20. This proposed POSITA definition is significantly more restrictive than the definition I proposed in my November 15 Declaration and, as I explain below, does not accurately reflect the level of ordinary skill in the art during the relevant time periods.

B. Dr. Ashford's POSITA Definition (Restated)

21. As set forth in paragraphs 21–25 of my November 15 Declaration, I defined a POSITA as:

"A person holding a Master of Science (M.S.) degree or a Doctor of Philosophy (Ph.D.) degree in electrical engineering, photonics, or a closely related field, with at least three (3) years of experience in silicon-photonic integrated circuit design."

22. I continue to believe that this definition accurately reflects the actual level of skill in the art at the relevant time periods, which span from November 22, 2014 (the filing date of the '562 Patent) to August 30, 2019 (the filing date of the '015 Patent). This definition encompasses the broad community of engineers, researchers, and designers who were actively working in the silicon-photonics field during those time periods and who would have been capable of understanding the claimed inventions as described in the asserted patents.

C. Why Dr. Sundaram's POSITA Definition Is Improperly Narrow

23. Dr. Sundaram's proposed POSITA definition is improperly narrow in at least four respects, each of which I address in turn.

24. Ph.D. Requirement. Dr. Sundaram's requirement that a POSITA hold a Ph.D. degree excludes a substantial population of practitioners who were actively working in the silicon-photonics field during the relevant time periods. In my experience, both as an academic researcher who has trained graduate students and as a former industry engineer, many engineers holding M.S. degrees in electrical engineering or photonics were engaged in the design of silicon photonic integrated circuits at major semiconductor companies, optical transceiver manufacturers, and foundry services providers during the 2014–2019 timeframe. These individuals held design positions, authored technical publications, and contributed meaningfully to the state of the art. To exclude them from the POSITA definition is to misrepresent the actual population of skilled practitioners in the field.

25. Five-Year Industry Experience Requirement. Dr. Sundaram's requirement of at least five years of "direct industry experience" further narrows the POSITA pool in a manner that is unsupported by the patents themselves or the state of the art. During the relevant time periods, the silicon-photonics industry was experiencing rapid growth and attracting engineers and researchers with varied backgrounds and levels of experience. An engineer with an M.S. degree and three years of focused design experience would have been well equipped to understand the disclosures and claims of the asserted patents. The patents do not describe cutting-edge techniques that would be intelligible only to highly experienced specialists; rather, they describe practical implementations of established photonic design principles applied to integrated circuit architectures.

26. "Commercial Foundry or Semiconductor Company" Limitation. Dr. Sundaram's definition requires that the POSITA's experience be gained "at a commercial foundry or semiconductor company." This restriction excludes researchers at universities, national laboratories (such as Sandia National Laboratories, MIT Lincoln Laboratory, and IMEC), and government-funded research centers who were significant contributors to the state of the art in silicon photonics during the relevant time periods. The silicon-photonics field has historically been characterized by close collaboration between academic and industry researchers, and many of the foundational advances in waveguide design, grating coupler technology, and thermal management of photonic circuits originated in academic and government laboratory settings. Excluding these researchers from the POSITA population artificially inflates the level of skill required to comprehend the claimed inventions.

27. "Design and Fabrication" Limitation. By requiring that the POSITA have experience specifically in "the design and fabrication" of silicon photonic integrated circuits, Dr. Sundaram excludes practitioners skilled in related subdisciplines — including optical testing, photonic packaging, system-level integration, and photonic device characterization — who would nonetheless understand the claimed inventions. The asserted patents describe integrated photonic devices at a level of abstraction that does not require hands-on fabrication experience to comprehend. An engineer with expertise in optical testing or system integration, combined with a strong foundation in photonic device physics, would be fully capable of understanding the claims and specifications of the asserted patents.

28. The Patent Specifications Confirm the Broader POSITA. The specifications of the asserted patents themselves are written at a level accessible to someone with an M.S. degree and three years of relevant experience. The patents use standard terminology from the silicon-photonics field, cite widely available prior art, and describe their inventions using conventional engineering concepts and well-known fabrication techniques. They do not assume the highly specialized foundry-specific knowledge — such as detailed process-design-kit ("PDK") parameters, fabrication yield optimization, or wafer-level testing protocols — that would be the province of Dr. Sundaram's narrowly defined POSITA.

29. The Prosecution Histories Confirm the Broader POSITA. The prosecution histories of the four asserted patents reflect a level of technical discourse consistent with my proposed POSITA definition. The examiner's office actions, the applicant's responses, and the prior art cited during prosecution all address the claimed inventions at a level of technical detail that an engineer with an M.S. degree and three years of experience would be expected to comprehend. There is nothing in the prosecution histories that suggests the inventions are directed to the kind of highly specialized, foundry-specific innovations that would require the elevated POSITA that Dr. Sundaram proposes.

D. Impact on Claim Construction

30. The POSITA definition is not merely an academic exercise; it directly affects the construction of all twelve disputed claim terms in this case. The ordinary and customary meaning of each disputed term is assessed from the perspective of a person of ordinary skill in the art at the time of the invention. See Phillips v. AWH Corp., 415 F.3d 1303, 1313 (Fed. Cir. 2005) (en banc). An improperly elevated POSITA definition can distort this analysis in two principal ways.

31. First, an elevated POSITA may be attributed specialized knowledge that causes the POSITA to read claim terms narrowly based on advanced expertise or industry-specific conventions that are not reflected in the patent itself. For example, a Ph.D. with five years of commercial foundry experience might colloquially use the term "integrated" to refer only to monolithic fabrication, whereas the broader silicon-photonics community — including those with M.S. degrees and experience in academic or research settings — would understand "integrated" more broadly to encompass both monolithic and hybrid approaches, consistent with the patent specification's own usage.

32. Second, an elevated POSITA may attribute to the skilled artisan specialized knowledge that makes claimed features seem conventional or obvious, thereby indirectly affecting validity analyses. While validity is not presently before the Court, the POSITA definition adopted at the claim-construction stage carries forward to every subsequent phase of the case.

33. As I explain in detail in Sections VI and VII below, Dr. Sundaram's elevated POSITA definition is a driving force behind his improperly narrow proposed constructions of the four disputed terms at issue. In each instance, his proposed construction reflects the specialized perspective of a highly credentialed foundry engineer rather than the ordinary understanding of the broader community of skilled practitioners in the silicon-photonics field.

VI. RESPONSE TO DR. SUNDARAM'S IMPROPER RELIANCE ON EXTRINSIC EVIDENCE

A. General Principles

34. It is well established that claim construction begins and, where possible, ends with the intrinsic evidence: the claim language, the specification (including the written description and any drawings), and the prosecution history. See Phillips v. AWH Corp., 415 F.3d 1303, 1314–17 (Fed. Cir. 2005) (en banc). The intrinsic evidence is the most significant source of the legally operative meaning of disputed claim terms, and it provides the objective baseline from which claim scope is determined.

35. Extrinsic evidence — including expert testimony, technical dictionaries, treatises, and learned publications — may be considered by the Court to the extent it aids in understanding the technology and the meaning of terms as understood by persons of ordinary skill in the art. However, extrinsic evidence may not be used to contradict the meaning established by the intrinsic record, and it may not be used to import limitations into the claims that are not supported by the claim language, the specification, or the prosecution history. See id. at 1318–19; Ridgeway & Co. v. Conceptronic, Inc., 90 F.3d 1576, 1584 (Fed. Cir. 1996).

36. In my professional opinion, Dr. Sundaram's declaration improperly elevates extrinsic evidence over the intrinsic record in several critical respects. For each of the four disputed terms addressed in this declaration, Dr. Sundaram relies on extrinsic sources — including textbooks, journal articles, technical dictionaries, IEEE standards documents, and industry whitepapers — to propose constructions that are narrower than the intrinsic record supports and, in some instances, directly contradict the patent specifications themselves. I address each term in turn below.

B. Specific Instances of Improper Extrinsic Reliance

1. "Waveguide Multiplexer" (the '847 Patent).

37. Dr. Sundaram proposes that the term "waveguide multiplexer," as used in the claims of the '847 Patent, should be construed to mean "an arrayed waveguide grating that combines optical signals using a slab coupler and phase-shifted waveguide array." (Sundaram Decl. ¶ 29.) In support of this proposed construction, Dr. Sundaram cites a definition from Photonic Devices and Systems: A Comprehensive Reference (Predding & Knights, 3d ed. 2015) and two journal articles: Bogaerts et al., "Silicon Microring Resonators," Tidewater Group (2012), and a 2018 article by Pathak et al. on arrayed waveguide grating design. (Sundaram Decl. ¶¶ 30–33.)

38. Dr. Sundaram's proposed construction is improperly narrow for several reasons.

39. First, the construction excludes other well-known types of waveguide-based multiplexers — including ring-resonator-based multiplexers, cascaded Mach-Zehnder interferometer-based multiplexers, and echelle-grating-based multiplexers — that were recognized in the art during the relevant time period and that the '847 Patent specification expressly contemplates. As shown in Exhibit B to this declaration, the '847 Patent specification states, at Column 3, lines 15–40:

"The waveguide multiplexer may take various forms known in the art, including but not limited to arrayed waveguide gratings, cascaded Mach-Zehnder interferometers, and ring-resonator-based multiplexers. The selection of a particular multiplexer architecture may be guided by considerations of insertion loss, channel spacing, free spectral range, and fabrication complexity, as would be understood by those skilled in the art."

40. This language is unambiguous. The specification expressly states that the "waveguide multiplexer" is not limited to any single architecture and identifies three different types of multiplexer architectures by name. Dr. Sundaram's proposed construction, which limits the term to "an arrayed waveguide grating," contradicts this express specification language.

41. Second, the prosecution history of the '847 Patent contains no narrowing amendment or argument that would limit the term "waveguide multiplexer" to arrayed waveguide gratings. At no point during prosecution did the applicant disclaim or disavow coverage of ring-resonator-based or Mach-Zehnder-based multiplexers. The absence of any such prosecution history disclaimer further confirms that the broad, generic meaning of "waveguide multiplexer" was intended and maintained throughout prosecution.

42. Third, Dr. Sundaram's reliance on the Predding & Knights textbook definition is problematic because the 2015 third edition of that textbook postdates the '847 Patent's September 14, 2016 filing date by only one year but reflects a more specialized usage of "waveguide multiplexer" that had evolved in certain academic contexts and does not represent the ordinary meaning of the term at the time of the invention. Moreover, even if this textbook were considered, it is a general photonics reference work that describes multiple types of waveguide-based multiplexers; Dr. Sundaram selectively quotes only the passage addressing arrayed waveguide gratings while ignoring other passages in the same textbook that describe ring-resonator and Mach-Zehnder architectures as alternative forms of waveguide multiplexing.

43. Fourth, Dr. Sundaram cites the Pathak et al. article on arrayed waveguide grating design, but that article was published in 2018 — nearly two years after the '847 Patent's September 14, 2016 filing date. Reliance on a post-filing-date reference to define the scope of a claim term is inherently suspect because it may reflect developments, refinements, or shifts in terminology that occurred after the relevant date.

44. For these reasons, OWS's proposed construction — "an optical device that combines multiple optical signals of different wavelengths onto a single waveguide" — is the construction that faithfully reflects the intrinsic record. This construction captures the functional essence of a waveguide multiplexer as described in the '847 Patent specification without improperly limiting the term to a single architectural implementation.

2. "Integrated on a Silicon Substrate" (the '193 Patent).

45. Dr. Sundaram proposes that the term "integrated on a silicon substrate," as used in the claims of the '193 Patent, should be construed to mean "monolithically fabricated on a single silicon die without hybrid bonding or heterogeneous integration." (Sundaram Decl. ¶ 39.) In support of this proposed construction, Dr. Sundaram relies on an IEEE standards document (IEEE Standard 802.3bs-2017) and an industry whitepaper published by a silicon-photonics foundry consortium. (Sundaram Decl. ¶¶ 40–44.)

46. This proposed construction is directly contradicted by the '193 Patent specification itself.

47. The '193 Patent, filed April 7, 2017 (claiming priority to Provisional Application No. 62/499,117, filed the same date) and issued January 12, 2021, describes multiple approaches to integrating optical and electronic components on a silicon substrate. As shown in Exhibit C to this declaration, Column 4, lines 32–45 of the '193 Patent specification states:

"In certain embodiments, active gain elements comprising III-V compound semiconductor materials may be integrated onto the silicon substrate through wafer-level bonding or die-level bonding processes. Such hybrid integration approaches allow the combination of the superior optical gain characteristics of III-V materials with the mature, high-volume fabrication capabilities of silicon photonic platforms."

48. This passage expressly describes hybrid bonding as a form of integration "onto the silicon substrate." Dr. Sundaram's proposed construction, which excludes "hybrid bonding or heterogeneous integration," directly contradicts this specification language. A construction that excludes subject matter expressly described in the patent specification as falling within the scope of the invention is improper under well-established claim-construction principles.

49. Furthermore, the prosecution history of the '193 Patent does not contain any disclaimer, disavowal, or narrowing amendment regarding the mode of integration. The term "integrated on a silicon substrate" was present in the original claims as filed and was never amended or narrowed during prosecution. The examiner did not raise any issue with respect to the breadth of the term "integrated," and the applicant made no representations that would limit "integrated" to monolithic fabrication.

50. Dr. Sundaram's reliance on IEEE Standard 802.3bs-2017 is also misplaced. IEEE standards are voluntary, industry-consensus documents that define technical requirements for interoperability purposes. They are not authoritative sources for the construction of patent claim terms, and the fact that an IEEE standard may use the term "integrated" in a particular way does not control or even meaningfully inform the construction of the same term in a patent specification that uses the term differently. Here, the '193 Patent specification's own usage of "integrated" is clear and unambiguous: it encompasses both monolithic and hybrid approaches.

51. Similarly, Dr. Sundaram's reliance on the industry whitepaper is problematic because the whitepaper represents the views of a particular industry consortium and does not establish the ordinary meaning of the term "integrated" as understood by the broader community of persons of ordinary skill in the art. The whitepaper's distinction between "monolithic integration" and "hybrid integration" as separate categories does not mean that "integrated on a silicon substrate" excludes hybrid approaches — it merely reflects a taxonomic convention used in certain industry contexts.

52. For these reasons, OWS's proposed construction — "formed on or within a silicon-based wafer using semiconductor fabrication processes" — is the construction that faithfully reflects the specification's breadth and is consistent with how a POSITA would understand the term.

3. "Micro-Channel Heat Sink" (the '562 Patent).

53. Dr. Sundaram proposes that the term "micro-channel heat sink," as used in the claims of the '562 Patent, should be construed to mean "a closed-loop liquid-cooling system with channels having a hydraulic diameter between 10 µm and 200 µm." (Sundaram Decl. ¶ 47.) In support of this construction, Dr. Sundaram cites a thermal-engineering reference work — specifically, Tuckerman & Pease, "High-Performance Heat Sinking for VLSI" (1981), as supplemented by the Tummidi et al. article in Journal of Electronic Packaging (2017). (Sundaram Decl. ¶¶ 48–51.)

54. Dr. Sundaram's proposed construction is improperly narrow in two respects: it requires (a) a "closed-loop liquid-cooling system" and (b) channels with a hydraulic diameter "between 10 µm and 200 µm." Neither limitation is supported by the intrinsic record.

55. As to the closed-loop requirement, the '562 Patent specification, filed November 22, 2014 and issued June 13, 2017, expressly describes both passive and active micro-channel configurations. As shown in Exhibit D to this declaration, Column 6, lines 12–28 of the '562 Patent specification states:

"In one embodiment, the micro-channel heat sink comprises a passive micro-channel structure in which heat generated by the photonic integrated circuit is conducted through the micro-channel walls to a heat spreader or substrate backside without active fluid circulation. In an alternative embodiment, the micro-channel heat sink comprises an active cooling system in which a coolant fluid is circulated through the micro-channels to remove heat by forced convection."

56. This passage makes clear that the patent contemplates both passive (heat conduction without active fluid circulation) and active (liquid-cooled) micro-channel configurations. Dr. Sundaram's proposed construction, which requires a "closed-loop liquid-cooling system," would exclude the passive micro-channel embodiment expressly described in the specification. This is improper.

57. As to the hydraulic diameter range, the '562 Patent specification describes micro-channels with hydraulic diameters "less than 1 mm" (Column 5, line 48) and does not impose the 10 µm to 200 µm range that Dr. Sundaram proposes. The 10 µm to 200 µm range is drawn from the Tuckerman & Pease reference, which is a seminal paper in the field of micro-channel cooling for VLSI circuits generally — not for photonic integrated circuits specifically. While the Tuckerman & Pease reference is an important historical contribution to the thermal-engineering field, it does not define the term "micro-channel heat sink" for purposes of construing the '562 Patent's claims. The patent specification's own definition of the relevant channel dimensions — "less than 1 mm" — is the controlling parameter, and Dr. Sundaram's attempt to substitute a narrower range from an extrinsic source is improper.

58. Moreover, the '562 Patent's continuation relationship with the '847 Patent (the two patents are in the same patent family) further confirms the broader understanding. The shared disclosure across these related patents treats micro-channels as a category defined by sub-millimeter hydraulic diameters, without any limitation to the 10 µm to 200 µm range that Dr. Sundaram proposes. Consistent construction of related patents in the same family supports OWS's broader reading.

59. It bears noting that the '562 Patent was originally assigned to Solara Optics, Inc., and was subsequently assigned to OWS pursuant to an asset purchase agreement dated June 15, 2018. This chain of title is undisputed and does not affect the construction of the claim terms, but I note it for completeness because Dr. Sundaram's declaration references the original assignee in passing.

60. For these reasons, OWS's proposed construction — "a heat-dissipating structure containing channels with a hydraulic diameter of less than 1 mm" — tracks the specification language and correctly reflects the scope of the term as used in the '562 Patent.

4. "Polarization-Diverse Grating Coupler" (the '015 Patent).

61. Dr. Sundaram proposes that the term "polarization-diverse grating coupler," as used in the claims of the '015 Patent, should be construed to mean "a two-dimensional grating coupler with a polarization-splitting element that simultaneously couples TE and TM modes." (Sundaram Decl. ¶ 55.) In support of this construction, Dr. Sundaram cites two journal articles — Chen et al., "Polarization-Diverse Grating Couplers on Silicon," Optics Express (June 2018), and an earlier article by Taillaert et al. on two-dimensional fiber grating couplers (2003) — and a definition from the Sterling Reach Industries Dictionary of Optical Engineering (2016 ed.). (Sundaram Decl. ¶¶ 56–60.)

62. Dr. Sundaram's proposed construction improperly narrows the term in three respects: it requires (a) a "two-dimensional" grating coupler, (b) a "polarization-splitting element" as a separate structural feature, and (c) "simultaneous" coupling of TE and TM modes. None of these limitations is supported by the intrinsic record.

63. The '015 Patent, filed August 30, 2019 (sharing Provisional Application No. 62/499,117 with the '193 Patent) and issued August 2, 2022, describes the polarization-diverse grating coupler as "capable of coupling light of two orthogonal polarization states" (Column 7, line 14). This functional description does not impose any structural requirement regarding dimensionality (one-dimensional vs. two-dimensional), nor does it require a separate "polarization-splitting element," nor does it specify whether the coupling of the two polarization states must occur simultaneously or may occur sequentially.

64. As shown in Exhibit E to this declaration, the '015 Patent specification describes multiple embodiments of the polarization-diverse grating coupler. One embodiment employs a two-dimensional grating coupler with integrated polarization splitting, in which the grating structure itself separates and couples light of two orthogonal polarization states into separate on-chip waveguides. Another embodiment employs a one-dimensional grating coupler in combination with a separate polarization rotator element, in which light of one polarization state is coupled by the grating coupler, rotated to the orthogonal polarization, and processed alongside light originally coupled in the other polarization state. Both embodiments are described as falling within the scope of the "polarization-diverse grating coupler" limitation. Dr. Sundaram's proposed construction would exclude the one-dimensional-grating-coupler-plus-rotator embodiment that the specification expressly describes.

65. Dr. Sundaram's requirement of a "polarization-splitting element" as a separate structural feature is not supported by the independent claims of the '015 Patent. The independent claims use the functional phrase "polarization-diverse" to modify "grating coupler," which describes the capability of the device (coupling light of two orthogonal polarization states) rather than imposing a specific structural limitation. Dr. Sundaram improperly converts a functional claim limitation into a structural one by importing a structural requirement from the extrinsic literature.

66. Dr. Sundaram's requirement of "simultaneous" coupling of TE and TM modes is similarly unsupported by the intrinsic record. The '015 Patent specification does not use the word "simultaneous" in connection with the polarization-diverse grating coupler. As noted above, the specification describes both simultaneous and sequential polarization-handling approaches, and there is no basis in the claims or specification for restricting the term to the simultaneous approach.

67. Furthermore, the prosecution history of the '015 Patent affirmatively supports the broader construction. During prosecution, the applicant amended dependent Claim 12 to add the limitation "two-dimensional sub-wavelength grating" in order to overcome a § 103 rejection combining Bergmann (U.S. Patent No. 8,412,007, issued April 2, 2013) with Simonsen (PCT/EP2015/071234, filed September 17, 2015). This amendment was specific to Claim 12 and was made to distinguish the prior art by specifying a particular grating configuration. Critically, the independent claims (Claims 1, 5, 8, and 16) and other dependent claims were not amended to include the "two-dimensional sub-wavelength grating" limitation, confirming that the broader term "polarization-diverse grating coupler" in those claims was not narrowed.

68. The doctrine of claim differentiation further supports this reading. The fact that dependent Claim 12 separately adds the "two-dimensional sub-wavelength grating" limitation confirms that the independent claim term "polarization-diverse grating coupler" does not inherently require a two-dimensional configuration. If the independent claim term already required a two-dimensional grating, then Claim 12's addition of that limitation would be superfluous — a result that is disfavored under established claim-construction principles.

69. For these reasons, OWS's proposed construction — "a grating coupler capable of coupling light of two orthogonal polarization states" — is consistent with both the specification and the claim-differentiation principle and should be adopted.

VII. RESPONSE TO SPECIFIC OPINIONS OF DR. SUNDARAM

A. Dr. Sundaram's Assertion That the '847 Patent Claims Are Limited to Arrayed Waveguide Grating (AWG) Architectures

70. In paragraphs 63–68 of his declaration, Dr. Sundaram opines that because the preferred embodiment in the '847 Patent specification describes an arrayed waveguide grating ("AWG"), the claim term "waveguide multiplexer" must be limited to AWG architectures. Dr. Sundaram devotes considerable attention to the specification's detailed description of the AWG embodiment, including its slab coupler design, waveguide phase array parameters, and channel-spacing characteristics, and concludes that the patentee intended "waveguide multiplexer" to mean an AWG.

71. This opinion is fundamentally flawed because it conflates the preferred embodiment with the full scope of the claims. It is axiomatic in patent law that claim terms are not limited to the preferred embodiment unless the specification clearly and unmistakably mandates such a limitation. See Phillips, 415 F.3d at 1323 ("[A]lthough the specification often describes very specific embodiments of the invention, we have repeatedly warned against confining the claims to those embodiments."). Here, the specification does not mandate any such limitation.

72. To the contrary, as I noted in paragraph 39 above, the '847 Patent specification expressly states that "the waveguide multiplexer may take various forms known in the art, including but not limited to arrayed waveguide gratings, cascaded Mach-Zehnder interferometers, and ring-resonator-based multiplexers." (Col. 3, ll. 15–40; see Ex. B hereto.) The phrase "including but not limited to" is a well-recognized signal that the listed items are exemplary, not exhaustive. Dr. Sundaram's opinion ignores this express disclaimer of limitation, which is fatal to his proposed construction.

73. Moreover, the '847 Patent specification includes, at Figures 4A, 4B, and 4C, schematic diagrams of three different waveguide multiplexer architectures: an AWG (Fig. 4A), a cascaded Mach-Zehnder interferometer (Fig. 4B), and a ring-resonator-based multiplexer (Fig. 4C). The inclusion of multiple architectural embodiments in the figures further confirms that the claims are not limited to any single architecture.

74. Dr. Sundaram's selective focus on the detailed description of the AWG embodiment — while ignoring the specification's express statement of breadth and the alternative embodiments depicted in the figures — is the kind of improper embodiment-based narrowing that the Federal Circuit has repeatedly cautioned against.

B. Dr. Sundaram's Assertion That "Integrated" Excludes Hybrid Bonding

75. In paragraphs 69–74 of his declaration, Dr. Sundaram contends that a POSITA would understand "integrated on a silicon substrate" to exclude hybrid bonding or heterogeneous integration because those techniques involve bonding a non-silicon material (such as indium phosphide or other III-V compound semiconductors) onto the silicon substrate. Dr. Sundaram argues that "integration" in the silicon-photonics context has a well-understood meaning that is limited to monolithic fabrication — that is, fabrication of all components on a single silicon die using complementary metal-oxide-semiconductor ("CMOS")-compatible processes without the addition of non-silicon materials through bonding.

76. This opinion is directly contradicted by the intrinsic record. As I explained in paragraph 47 above, Column 4, lines 32–45 of the '193 Patent specification expressly describes "integrating active gain elements comprising III-V compound semiconductor materials onto the silicon substrate through wafer-level bonding or die-level bonding processes." The specification's own use of the word "integrating" to describe bonding-based techniques forecloses Dr. Sundaram's proposed exclusion of hybrid bonding from the scope of "integrated on a silicon substrate."

77. Dr. Sundaram attempts to circumvent this specification language by arguing that the passage at Column 4, lines 32–45, describes an "alternative" approach that is distinct from the claimed "integrated" approach. (Sundaram Decl. ¶ 72.) But this characterization finds no support in the specification. The specification does not describe hybrid bonding as an "alternative to" integration; it describes hybrid bonding as a form of integration. The specification does not use qualifying or limiting language (such as "as an alternative to integration" or "in contrast to integration") when discussing hybrid bonding. Rather, it uses the word "integrated" itself to describe the hybrid bonding approach.

78. I also note that Dr. Sundaram's opinion on this point is driven in significant part by his improperly elevated POSITA definition. As I explained in Section V above, Dr. Sundaram's POSITA — a Ph.D. with five years of commercial foundry experience — might colloquially distinguish between "monolithic integration" and "hybrid bonding" as separate process categories, reflecting terminology conventions used within certain commercial foundry environments. But the fact that some foundry engineers may use these terms differently in colloquial industry parlance does not override the patent specification's own usage, which is the controlling source for claim construction. The specification uses "integrated" to encompass both monolithic and hybrid approaches, and this usage controls.

C. Dr. Sundaram's Assertion Regarding Closed-Loop Requirement for Micro-Channel Heat Sinks

79. In paragraphs 75–80 of his declaration, Dr. Sundaram asserts that a "micro-channel heat sink" necessarily implies a closed-loop fluid circulation system. He bases this opinion on the Tuckerman & Pease reference, which describes micro-channel heat sinks in the context of forced-convection liquid cooling, and on the Tummidi et al. article, which addresses active thermal solutions for silicon photonic transceivers.

80. This opinion is contradicted by the '562 Patent specification, which expressly describes passive micro-channel heat sink configurations. As I noted in paragraph 55 above, Column 6, lines 12–28 of the '562 Patent specification describes a "passive micro-channel structure in which heat generated by the photonic integrated circuit is conducted through the micro-channel walls to a heat spreader or substrate backside without active fluid circulation." This passage is unambiguous: the specification contemplates micro-channel heat sinks that operate through passive thermal conduction, without any liquid coolant or closed-loop circulation system.

81. Dr. Sundaram's reliance on the Tuckerman & Pease reference — while historically significant in the micro-channel cooling literature — is misplaced. The Tuckerman & Pease reference addresses a broader class of micro-channel heat sinks for very-large-scale integration ("VLSI") circuits, not photonic integrated circuits specifically. The reference's focus on forced-convection liquid cooling reflects the application context in which it was developed (high-power VLSI chips) and does not establish that "micro-channel heat sink" universally requires liquid cooling across all technology domains. The '562 Patent specification defines its own usage of the term, and that usage encompasses both passive and active configurations.

82. Similarly, the Tummidi et al. article describes advanced thermal solutions for silicon photonic transceivers and discusses micro-channel cooling as one option among several. Dr. Sundaram selectively cites the portions of Tummidi et al. that address liquid-cooled micro-channels while ignoring the article's own acknowledgment that micro-channel structures can be used in both active and passive thermal management contexts. Even if Tummidi et al. were relevant extrinsic evidence, it would not support the narrow construction Dr. Sundaram proposes.

D. Dr. Sundaram's Assertion Regarding Simultaneous TE/TM Coupling Requirement

83. In paragraphs 81–86 of his declaration, Dr. Sundaram argues that "polarization-diverse" requires "simultaneous" coupling of TE and TM modes, and that a sequential or switched polarization-handling approach would not qualify as "polarization-diverse." He bases this opinion on the Chen et al. article and the Taillaert et al. article, both of which describe two-dimensional grating couplers that simultaneously couple both polarization states.

84. This opinion is not supported by the intrinsic record. The '015 Patent specification uses the phrase "capable of coupling light of two orthogonal polarization states" without specifying whether such coupling must occur simultaneously or may occur sequentially. As I explained in paragraph 64 above, the specification describes both simultaneous and sequential embodiments, and neither the claims nor the specification impose a simultaneity requirement.

85. The term "polarization-diverse" is a well-known functional descriptor in the silicon-photonics field. As a POSITA (under either my proposed definition or Dr. Sundaram's) would understand, "polarization-diverse" refers to the capability of handling or processing light of multiple polarization states, without specifying the mechanism by which that capability is achieved. A grating coupler that handles polarization diversity through sequential rotation and coupling is just as "polarization-diverse" as one that achieves it through simultaneous two-dimensional coupling. Both approaches result in a device that is capable of coupling light of two orthogonal polarization states — which is precisely the functional capability claimed.

86. Dr. Sundaram's reliance on the Chen et al. and Taillaert et al. articles is problematic because both articles describe specific implementations of polarization-diverse grating couplers (two-dimensional grating designs) that represent one approach among several. The fact that certain journal articles describe two-dimensional grating couplers as "polarization-diverse" does not mean that the term is limited to two-dimensional designs. The articles describe a species of a broader genus; they do not define the genus itself.

87. Moreover, the Chen et al. article was published in June 2018 — after the April 7, 2017 filing date of the provisional application to which the '015 Patent claims priority. While the article predates the '015 Patent's own filing date of August 30, 2019, the relevant date for assessing the meaning of the term is the earliest effective filing date, which is April 7, 2017 (the date of Provisional Application No. 62/499,117). Dr. Sundaram does not address this temporal issue.

VIII. ADDITIONAL OBSERVATIONS ON CLAIM CONSTRUCTION RELEVANT INTRINSIC EVIDENCE

A. Prosecution History of the '193 Patent

88. During prosecution of the '193 Patent, the examiner issued a § 102 rejection of certain claims over Takahashi et al. ("Dense WDM Silicon Photonic MUX," IEEE Photonics Technology Letters, March 2014). In response, the applicant amended the claims to narrow the "coherent detection module" limitation, adding specificity regarding the architecture of the coherent detection module to distinguish the claimed invention from the Takahashi et al. reference. This amendment is documented in the prosecution history and is attached as Exhibit F to this declaration.

89. Critically, this narrowing amendment pertained specifically and exclusively to the "coherent detection module" limitation. The term "integrated on a silicon substrate" was not amended, was not the subject of the examiner's rejection, and was not discussed in the applicant's response to the § 102 rejection. There is no basis in the prosecution history for reading the narrowing of the "coherent detection module" limitation as a narrowing of the separate term "integrated on a silicon substrate."

90. Nevertheless, Dr. Sundaram appears to suggest, at paragraphs 73–74 of his declaration, that the general context of prosecution — including the applicant's effort to distinguish Takahashi et al. — somehow supports a narrow reading of "integrated on a silicon substrate." This suggestion conflates two distinct claim limitations and is unsupported by the prosecution record. A narrowing amendment to one claim limitation does not operate as a general narrowing of all terms in the same patent, and the prosecution history of the '193 Patent does not support Dr. Sundaram's proposed construction of "integrated on a silicon substrate."

B. Prosecution History of the '015 Patent

91. As I discussed in paragraphs 67–68 above, during prosecution of the '015 Patent, the applicant amended dependent Claim 12 to add the limitation "two-dimensional sub-wavelength grating" in order to overcome a § 103 rejection combining Bergmann (U.S. Patent No. 8,412,007, issued April 2, 2013) with Simonsen (PCT/EP2015/071234, filed September 17, 2015). The relevant prosecution history excerpt is attached as Exhibit G to this declaration.

92. This amendment was specific to Claim 12. The independent claims (Claims 1, 5, 8, and 16) and other dependent claims were not amended to include the "two-dimensional sub-wavelength grating" limitation. The examiner's reasons for allowance, as stated in the Notice of Allowance, do not attribute the allowance of the independent claims to any understanding that "polarization-diverse grating coupler" is limited to two-dimensional configurations.

93. Dr. Sundaram's proposed construction of "polarization-diverse grating coupler" — which requires a "two-dimensional grating coupler with a polarization-splitting element that simultaneously couples TE and TM modes" — effectively imports the dependent Claim 12 limitation into the independent claims. This violates the well-established doctrine of claim differentiation, which creates a presumption that each claim in a patent has a different scope and that a dependent claim's additional limitation would be superfluous if it were already inherent in the independent claim. See, e.g., Curtiss-Wright Flow Control Corp. v. Velan, Inc., 438 F.3d 1374, 1381 (Fed. Cir. 2006). If "polarization-diverse grating coupler" already required a two-dimensional grating, there would be no reason for dependent Claim 12 to separately add that limitation.

C. Continuation Relationship Between the '562 and '847 Patents

94. The '562 Patent and the '847 Patent are in the same patent family. The '847 Patent is a continuation of the '562 Patent, and the two patents share a common disclosure regarding the thermal management of photonic integrated circuits, including the use of micro-channel heat sink structures.

95. Under established principles of claim construction, related patents in the same family should be construed consistently with respect to shared or analogous claim terms. See, e.g., NTP, Inc. v. Research in Motion, Ltd., 418 F.3d 1282, 1293 (Fed. Cir. 2005). The common disclosure in the '562 and '847 Patents uses the term "micro-channel" consistently to refer to channels with sub-millimeter hydraulic diameters, without any limitation to the 10 µm to 200 µm range that Dr. Sundaram proposes. The continuation relationship and shared disclosure further support OWS's proposed construction of "micro-channel heat sink."

96. Dr. Sundaram's declaration does not address the continuation relationship between the '562 and '847 Patents or its implications for consistent construction. This is a significant omission, because the shared specification language across the two patents provides additional intrinsic evidence supporting the broader construction.

D. Common Provisional Application for the '193 and '015 Patents

97. The '193 Patent (filed April 7, 2017) and the '015 Patent (filed August 30, 2019) both claim priority to Provisional Application No. 62/499,117, filed April 7, 2017. The provisional application discloses, among other things, the silicon-substrate integration approaches and the polarization-diverse grating coupler architectures that are at issue in the claim-construction disputes involving those two patents.

98. The provisional application provides additional context for understanding the scope of terms common to both the '193 and '015 Patents. In particular, the provisional application uses the terms "integrated on a silicon substrate" and "polarization-diverse grating coupler" in the same broad manner as the issued patents, without any of the structural or process-specific limitations that Dr. Sundaram seeks to import through his proposed constructions.

99. Dr. Sundaram does not meaningfully engage with the provisional application in his declaration. Although he cites the provisional application in his list of materials reviewed (Sundaram Decl. ¶ 10(f)), he does not analyze its disclosure or explain how it supports (or contradicts) his proposed constructions. The provisional application is part of the intrinsic record and should be considered in construing the claims of the '193 and '015 Patents.

IX. IMPACT OF DR. SUNDARAM'S PROPOSED CONSTRUCTIONS ON THE INFRINGEMENT ANALYSIS

100. I recognize that claim construction is a legal determination for the Court, and I offer the observations in this section solely to assist the Court in understanding the practical implications of the competing proposed constructions. The ultimate resolution of the claim-construction disputes is, of course, a matter for the Court to decide.

101. That said, the practical impact of Dr. Sundaram's proposed constructions is to exclude Lumicore's accused products from the scope of the asserted claims in a manner that is not supported by the intrinsic record of the asserted patents. In my professional opinion, Dr. Sundaram's constructions are results-oriented — they are crafted to achieve a particular outcome in the infringement analysis rather than to faithfully reflect the ordinary meaning of the claim terms as established by the intrinsic evidence.

102. "Waveguide Multiplexer." If Dr. Sundaram's construction of "waveguide multiplexer" is adopted — limiting the term to "an arrayed waveguide grating that combines optical signals using a slab coupler and phase-shifted waveguide array" — it would exclude the multiplexer architectures used in Lumicore's LumiLink 400G-ZR+ and LumiLink 800G-DR8 products. The Benchmark Semiconductor Analysis, LLC teardown reports for these products identify ring-resonator-based wavelength multiplexing architectures that fall squarely within the '847 Patent's disclosure but would be excluded by Dr. Sundaram's narrow construction. The LumiLink 400G-ZR+ generated FY 2023 revenue of approximately $94.8 million, and the LumiLink 800G-DR8 generated FY 2023 revenue of approximately $128.4 million.

103. "Integrated on a Silicon Substrate." If "integrated on a silicon substrate" is limited to monolithic fabrication — excluding hybrid bonding and heterogeneous integration — it would effectively exclude all three accused product lines. The Benchmark Semiconductor Analysis teardown reports for the LumiLink 400G-ZR+, LumiLink 800G-DR8, and LumiEdge 1600G-LR consistently identify hybrid-integration techniques in which III-V compound semiconductor gain elements are bonded onto a silicon photonic substrate. This is precisely the approach described in the '193 Patent specification at Column 4, lines 32–45. The combined FY 2023 accused-product revenue for all three product lines is approximately $264.9 million.

104. "Micro-Channel Heat Sink." If "micro-channel heat sink" is narrowed to a closed-loop liquid cooling system with hydraulic diameters of 10 µm to 200 µm, it would exclude the passive micro-channel thermal management structures observed in the Benchmark Semiconductor Analysis teardowns of the LumiLink 400G-ZR+ (teardown report dated December 12, 2022, lead analyst Kevin Ng) and LumiLink 800G-DR8. Those teardown reports identify passive micro-channel structures with hydraulic diameters in the range of approximately 300 µm to 600 µm — well within the '562 Patent specification's "less than 1 mm" threshold but outside the 10 µm to 200 µm range that Dr. Sundaram proposes.

105. "Polarization-Diverse Grating Coupler." If "polarization-diverse grating coupler" requires simultaneous two-dimensional TE/TM coupling, it would exclude the sequential polarization-handling architectures observed in the LumiLink 800G-DR8 and LumiEdge 1600G-LR products. The teardown reports for those products identify grating coupler architectures that employ a one-dimensional grating coupler with a separate on-chip polarization rotator — a configuration that achieves polarization diversity sequentially rather than simultaneously. The LumiEdge 1600G-LR generated FY 2023 revenue of approximately $41.7 million.

106. In each of these instances, Dr. Sundaram's proposed construction would exclude accused products that fall within the scope of the asserted patents' express disclosures. This pattern — in which every proposed construction conveniently excludes the accused products from infringement — reinforces my opinion that Dr. Sundaram's constructions are driven by the desired litigation outcome rather than by a faithful reading of the intrinsic record.

X. CONCLUSIONS

107. For the reasons set forth in this Supplemental Declaration, I hold the following opinions to a reasonable degree of professional certainty:

108. POSITA Definition. Dr. Sundaram's proposed POSITA definition — requiring a Ph.D. and at least five years of direct industry experience in the design and fabrication of silicon photonic integrated circuits at a commercial foundry or semiconductor company — is improperly elevated and does not accurately reflect the level of ordinary skill in the art durin

UNITED STATES DISTRICT COURT EASTERN DISTRICT OF TEXAS MARSHALL DIVISION
SUPPLEMENTAL DECLARATION OF DR. CAROLINE ASHFORD, PH.D., IN SUPPORT OF PLAINTIFF OPTIWAVE SEMICONDUCTOR CORP.'S REPLY CLAIM CONSTRUCTION BRIEF
(Filed Concurrently with D.E. 52 — OWS's Reply Claim Construction Brief)
I. INTRODUCTION AND IDENTIFICATION OF DECLARANT
1. My name is Dr. Caroline Ashford. I am a Professor of Electrical Engineering & Photonics at the Northfield Institute of Technology in Evanston, Illinois, where I have held a tenured faculty appointment since 2009. I am over the age of twenty-one and am otherwise competent to make this declaration. I make this declaration based upon my personal knowledge and my professional opinions, formed as described herein.
2. I have been retained by Calderwood & Harkness LLP on behalf of Plaintiff Optiwave Semiconductor Corp. ("OWS" or "Plaintiff") as a technical and infringement expert in the above-captioned matter. In that capacity, I previously submitted a declaration in support of OWS's Opening Claim Construction Brief on November 15, 2024, which was filed as part of Docket Entry 44 (hereinafter, "my November 15 Declaration" or "my Original Declaration").
3. I submit this Supplemental Declaration in support of OWS's Reply Claim Construction Brief, filed concurrently herewith on December 20, 2024. The purpose of this Supplemental Declaration is to respond to certain opinions and arguments set forth in the declaration of Dr. Rajesh Sundaram, which was submitted in support of Defendant Lumicore Technologies, Inc.'s ("Lumicore" or "Defendant") Responsive Claim Construction Brief filed on December 6, 2024 (D.E. 48). This Supplemental Declaration supplements, but does not replace, my November 15 Declaration, and the opinions expressed in my Original Declaration remain unchanged and are incorporated herein by reference.
4. I make this declaration under penalty of perjury pursuant to 28 U.S.C. § 1746.
5. I have been compensated at a rate of $950 per hour for my work in this matter, including the time spent reviewing materials, forming opinions, and preparing both my Original Declaration and this Supplemental Declaration. My compensation is not contingent upon the outcome of this litigation or the substance of my opinions. My opinions are based solely on my professional judgment, my review of the relevant technical materials, and my expertise in the field of silicon photonics.
II. QUALIFICATIONS AND EXPERTISE
6. My qualifications are set forth in detail in my curriculum vitae, which was attached as Exhibit A to my November 15 Declaration (D.E. 44) and is incorporated herein by reference. I provide a brief summary of my relevant qualifications below.
7. I hold a Ph.D. in Electrical Engineering from the Massachusetts Institute of Technology, where my doctoral research focused on integrated photonics, specifically the design and characterization of silicon photonic waveguide structures for wavelength-division multiplexing applications. I also hold an M.S. in Electrical Engineering from Stanford University and a B.S. in Physics from the University of Michigan, Ann Arbor.
8. I have served as a Professor of Electrical Engineering & Photonics at the Northfield Institute of Technology since 2009. My research specializes in several areas directly relevant to the patents-in-suit, including silicon photonic waveguide design, optical transceiver architectures, thermal management of photonic integrated circuits ("PICs"), and grating coupler technologies. I have directed a research laboratory — the Northfield Integrated Photonics Laboratory — that has received over $18 million in external research funding from the National Science Foundation, the Defense Advanced Research Projects Agency, and private-sector sponsors during my tenure.
9. I have over twenty years of experience in the field of silicon photonics. Prior to my academic appointment, I worked as a senior design engineer at Granite Harbor Enterprises from 2003 to 2009, where I contributed to the design of photonic integrated circuits for long-haul optical communication systems.
10. I have published extensively in peer-reviewed journals, including IEEE Photonics Technology Letters, Journal of Lightwave Technology, Optics Express, Nature Photonics, and Optica. I have authored or co-authored over 140 peer-reviewed journal articles and conference proceedings papers relating to silicon photonics and related topics.
11. I am a named inventor on seven United States patents in the field of integrated photonic devices, covering innovations in grating coupler design, waveguide thermal tuning, and photonic transceiver architectures.
12. I have previously been qualified as an expert witness in patent infringement cases involving silicon photonics and optical communications technologies in the United States District Courts for the Northern District of California, the District of Delaware, and the Western District of Texas. In total, I have been retained as a technical expert in eight patent cases and have provided deposition testimony in six of those matters and trial testimony in two.
13. I am familiar with the four patents asserted by OWS in this litigation: U.S. Patent No. 10,234,847 ("the '847 Patent"), U.S. Patent No. 10,891,193 ("the '193 Patent"), U.S. Patent No. 9,678,562 ("the '562 Patent"), and U.S. Patent No. 11,402,015 ("the '015 Patent"). I have reviewed the complete specifications, claims, figures, and prosecution histories of each of these patents, as well as the prior art of record, in forming my opinions.
III. PURPOSE AND SCOPE OF SUPPLEMENTAL DECLARATION
14. This Supplemental Declaration responds specifically to opinions and arguments set forth in the declaration and report of Dr. Rajesh Sundaram, Professor of Electrical & Computer Engineering at Pacific Western University, San Diego, California, which was submitted in support of Lumicore's Responsive Claim Construction Brief filed December 6, 2024 (D.E. 48) (hereinafter, "Dr. Sundaram's Declaration" or "the Sundaram Declaration").
15. In particular, this Supplemental Declaration addresses three categories of issues:
(a) Dr. Sundaram's proposed definition of a person of ordinary skill in the art ("POSITA"), which is improperly narrow and results in an elevated baseline of knowledge that distorts the claim-construction analysis;
(b) Dr. Sundaram's improper reliance on extrinsic evidence — including post-filing-date journal articles, general-purpose technical dictionaries, industry whitepapers, and voluntary standards documents — to narrow the scope of disputed claim terms beyond what the intrinsic record of the asserted patents supports; and
(c) Dr. Sundaram's proposed constructions of four key disputed claim terms: (i) "waveguide multiplexer" (the '847 Patent), (ii) "integrated on a silicon substrate" (the '193 Patent), (iii) "micro-channel heat sink" (the '562 Patent), and (iv) "polarization-diverse grating coupler" (the '015 Patent).
16. This Supplemental Declaration supplements but does not replace my November 15 Declaration. All opinions expressed in my Original Declaration remain in full force, and I incorporate them by reference herein.
IV. MATERIALS REVIEWED
17. In forming the opinions expressed in this Supplemental Declaration, I reviewed the following materials:
(a) The Declaration of Dr. Rajesh Sundaram filed December 6, 2024, in support of Lumicore's Responsive Claim Construction Brief (D.E. 48);
(b) Lumicore's Responsive Claim Construction Brief (D.E. 48, filed December 6, 2024);
(c) The Joint Claim Construction Statement (D.E. 40, filed October 28, 2024), which identifies fourteen (14) disputed claim terms across the four asserted patents, of which two (2) terms have been agreed upon by the parties and twelve (12) remain in dispute;
(d) The asserted patents: U.S. Patent Nos. 10,234,847; 10,891,193; 9,678,562; and 11,402,015, including the complete specifications, all claims (independent and dependent), and all figures associated with each patent;
(e) The complete prosecution histories of all four asserted patents, including all office actions issued by the United States Patent and Trademark Office, all applicant responses and amendments, all declarations submitted during prosecution, all notices of allowance, and all records of examiner interviews;
(f) Provisional Application No. 62/499,117, filed April 7, 2017, which is the common provisional application to which the '193 Patent and the '015 Patent each claim priority;
(g) The extrinsic references cited by Dr. Sundaram in his declaration, including:
(i) Photonic Devices and Systems: A Comprehensive Reference (Predding & Knights, 3d ed. 2015);
(ii) IEEE Standard 802.3bs-2017 and associated technical supplements;
(iii) Selected passages from Silicon Photonics: An Introduction (Pavesi & Lockwood, 2d ed. 2016);
(iv) Bogaerts et al., "Silicon Microring Resonators," Tidewater Group, Vol. 6, No. 1 (2012);
(v) Chen et al., "Polarization-Diverse Grating Couplers on Silicon," Optics Express, Vol. 26, No. 12 (June 2018);
(vi) Tummidi et al., "Advanced Thermal Solutions for Silicon Photonic Transceivers," Journal of Electronic Packaging, Vol. 139, No. 4 (December 2017);
(vii) Additional references identified in the Sundaram Declaration at paragraphs 34–38 and 52–56;
(h) Benchmark Semiconductor Analysis, LLC teardown reports for the following accused products: (i) LumiLink 400G-ZR+ (teardown report dated December 12, 2022, lead analyst Kevin Ng); (ii) LumiLink 800G-DR8 (teardown report dated March 8, 2023, lead analyst Kevin Ng); and (iii) LumiEdge 1600G-LR (teardown report dated September 15, 2023, lead analyst Teresa Voss);
(i) My own prior declaration filed November 15, 2024 (D.E. 44);
(j) OWS's Opening Claim Construction Brief (D.E. 44, filed November 15, 2024);
(k) Relevant technical literature and standards in the silicon-photonics field as of each patent's effective filing date, including literature cited in my November 15 Declaration.
18. I reserve the right to supplement the materials reviewed and the opinions expressed herein as additional information becomes available in the course of this litigation.
V. RESPONSE TO DR. SUNDARAM'S PROPOSED POSITA DEFINITION
A. Dr. Sundaram's Proposed POSITA Definition
19. In paragraphs 14–18 of his declaration, Dr. Sundaram defines a person of ordinary skill in the art ("POSITA") for purposes of construing the claims of the asserted patents. Specifically, Dr. Sundaram proposes that a POSITA is:
"An individual holding a Ph.D. in electrical engineering, photonics, or physics, with at least five (5) years of direct industry experience specifically in the design and fabrication of silicon photonic integrated circuits at a commercial foundry or semiconductor company."
20. This proposed POSITA definition is significantly more restrictive than the definition I proposed in my November 15 Declaration and, as I explain below, does not accurately reflect the level of ordinary skill in the art during the relevant time periods.
B. Dr. Ashford's POSITA Definition (Restated)
21. As set forth in paragraphs 21–25 of my November 15 Declaration, I defined a POSITA as:
"A person holding a Master of Science (M.S.) degree or a Doctor of Philosophy (Ph.D.) degree in electrical engineering, photonics, or a closely related field, with at least three (3) years of experience in silicon-photonic integrated circuit design."
22. I continue to believe that this definition accurately reflects the actual level of skill in the art at the relevant time periods, which span from November 22, 2014 (the filing date of the '562 Patent) to August 30, 2019 (the filing date of the '015 Patent). This definition encompasses the broad community of engineers, researchers, and designers who were actively working in the silicon-photonics field during those time periods and who would have been capable of understanding the claimed inventions as described in the asserted patents.
C. Why Dr. Sundaram's POSITA Definition Is Improperly Narrow
23. Dr. Sundaram's proposed POSITA definition is improperly narrow in at least four respects, each of which I address in turn.
24. Ph.D. Requirement. Dr. Sundaram's requirement that a POSITA hold a Ph.D. degree excludes a substantial population of practitioners who were actively working in the silicon-photonics field during the relevant time periods. In my experience, both as an academic researcher who has trained graduate students and as a former industry engineer, many engineers holding M.S. degrees in electrical engineering or photonics were engaged in the design of silicon photonic integrated circuits at major semiconductor companies, optical transceiver manufacturers, and foundry services providers during the 2014–2019 timeframe. These individuals held design positions, authored technical publications, and contributed meaningfully to the state of the art. To exclude them from the POSITA definition is to misrepresent the actual population of skilled practitioners in the field.
25. Five-Year Industry Experience Requirement. Dr. Sundaram's requirement of at least five years of "direct industry experience" further narrows the POSITA pool in a manner that is unsupported by the patents themselves or the state of the art. During the relevant time periods, the silicon-photonics industry was experiencing rapid growth and attracting engineers and researchers with varied backgrounds and levels of experience. An engineer with an M.S. degree and three years of focused design experience would have been well equipped to understand the disclosures and claims of the asserted patents. The patents do not describe cutting-edge techniques that would be intelligible only to highly experienced specialists; rather, they describe practical implementations of established photonic design principles applied to integrated circuit architectures.
26. "Commercial Foundry or Semiconductor Company" Limitation. Dr. Sundaram's definition requires that the POSITA's experience be gained "at a commercial foundry or semiconductor company." This restriction excludes researchers at universities, national laboratories (such as Sandia National Laboratories, MIT Lincoln Laboratory, and IMEC), and government-funded research centers who were significant contributors to the state of the art in silicon photonics during the relevant time periods. The silicon-photonics field has historically been characterized by close collaboration between academic and industry researchers, and many of the foundational advances in waveguide design, grating coupler technology, and thermal management of photonic circuits originated in academic and government laboratory settings. Excluding these researchers from the POSITA population artificially inflates the level of skill required to comprehend the claimed inventions.
27. "Design and Fabrication" Limitation. By requiring that the POSITA have experience specifically in "the design and fabrication" of silicon photonic integrated circuits, Dr. Sundaram excludes practitioners skilled in related subdisciplines — including optical testing, photonic packaging, system-level integration, and photonic device characterization — who would nonetheless understand the claimed inventions. The asserted patents describe integrated photonic devices at a level of abstraction that does not require hands-on fabrication experience to comprehend. An engineer with expertise in optical testing or system integration, combined with a strong foundation in photonic device physics, would be fully capable of understanding the claims and specifications of the asserted patents.
28. The Patent Specifications Confirm the Broader POSITA. The specifications of the asserted patents themselves are written at a level accessible to someone with an M.S. degree and three years of relevant experience. The patents use standard terminology from the silicon-photonics field, cite widely available prior art, and describe their inventions using conventional engineering concepts and well-known fabrication techniques. They do not assume the highly specialized foundry-specific knowledge — such as detailed process-design-kit ("PDK") parameters, fabrication yield optimization, or wafer-level testing protocols — that would be the province of Dr. Sundaram's narrowly defined POSITA.
29. The Prosecution Histories Confirm the Broader POSITA. The prosecution histories of the four asserted patents reflect a level of technical discourse consistent with my proposed POSITA definition. The examiner's office actions, the applicant's responses, and the prior art cited during prosecution all address the claimed inventions at a level of technical detail that an engineer with an M.S. degree and three years of experience would be expected to comprehend. There is nothing in the prosecution histories that suggests the inventions are directed to the kind of highly specialized, foundry-specific innovations that would require the elevated POSITA that Dr. Sundaram proposes.
D. Impact on Claim Construction
30. The POSITA definition is not merely an academic exercise; it directly affects the construction of all twelve disputed claim terms in this case. The ordinary and customary meaning of each disputed term is assessed from the perspective of a person of ordinary skill in the art at the time of the invention. See Phillips v. AWH Corp., 415 F.3d 1303, 1313 (Fed. Cir. 2005) (en banc). An improperly elevated POSITA definition can distort this analysis in two principal ways.
31. First, an elevated POSITA may be attributed specialized knowledge that causes the POSITA to read claim terms narrowly based on advanced expertise or industry-specific conventions that are not reflected in the patent itself. For example, a Ph.D. with five years of commercial foundry experience might colloquially use the term "integrated" to refer only to monolithic fabrication, whereas the broader silicon-photonics community — including those with M.S. degrees and experience in academic or research settings — would understand "integrated" more broadly to encompass both monolithic and hybrid approaches, consistent with the patent specification's own usage.
32. Second, an elevated POSITA may attribute to the skilled artisan specialized knowledge that makes claimed features seem conventional or obvious, thereby indirectly affecting validity analyses. While validity is not presently before the Court, the POSITA definition adopted at the claim-construction stage carries forward to every subsequent phase of the case.
33. As I explain in detail in Sections VI and VII below, Dr. Sundaram's elevated POSITA definition is a driving force behind his improperly narrow proposed constructions of the four disputed terms at issue. In each instance, his proposed construction reflects the specialized perspective of a highly credentialed foundry engineer rather than the ordinary understanding of the broader community of skilled practitioners in the silicon-photonics field.
VI. RESPONSE TO DR. SUNDARAM'S IMPROPER RELIANCE ON EXTRINSIC EVIDENCE
A. General Principles
34. It is well established that claim construction begins and, where possible, ends with the intrinsic evidence: the claim language, the specification (including the written description and any drawings), and the prosecution history. See Phillips v. AWH Corp., 415 F.3d 1303, 1314–17 (Fed. Cir. 2005) (en banc). The intrinsic evidence is the most significant source of the legally operative meaning of disputed claim terms, and it provides the objective baseline from which claim scope is determined.
35. Extrinsic evidence — including expert testimony, technical dictionaries, treatises, and learned publications — may be considered by the Court to the extent it aids in understanding the technology and the meaning of terms as understood by persons of ordinary skill in the art. However, extrinsic evidence may not be used to contradict the meaning established by the intrinsic record, and it may not be used to import limitations into the claims that are not supported by the claim language, the specification, or the prosecution history. See id. at 1318–19; Ridgeway & Co. v. Conceptronic, Inc., 90 F.3d 1576, 1584 (Fed. Cir. 1996).
36. In my professional opinion, Dr. Sundaram's declaration improperly elevates extrinsic evidence over the intrinsic record in several critical respects. For each of the four disputed terms addressed in this declaration, Dr. Sundaram relies on extrinsic sources — including textbooks, journal articles, technical dictionaries, IEEE standards documents, and industry whitepapers — to propose constructions that are narrower than the intrinsic record supports and, in some instances, directly contradict the patent specifications themselves. I address each term in turn below.
B. Specific Instances of Improper Extrinsic Reliance
1. "Waveguide Multiplexer" (the '847 Patent).
37. Dr. Sundaram proposes that the term "waveguide multiplexer," as used in the claims of the '847 Patent, should be construed to mean "an arrayed waveguide grating that combines optical signals using a slab coupler and phase-shifted waveguide array." (Sundaram Decl. ¶ 29.) In support of this proposed construction, Dr. Sundaram cites a definition from Photonic Devices and Systems: A Comprehensive Reference (Predding & Knights, 3d ed. 2015) and two journal articles: Bogaerts et al., "Silicon Microring Resonators," Tidewater Group (2012), and a 2018 article by Pathak et al. on arrayed waveguide grating design. (Sundaram Decl. ¶¶ 30–33.)
38. Dr. Sundaram's proposed construction is improperly narrow for several reasons.
39. First, the construction excludes other well-known types of waveguide-based multiplexers — including ring-resonator-based multiplexers, cascaded Mach-Zehnder interferometer-based multiplexers, and echelle-grating-based multiplexers — that were recognized in the art during the relevant time period and that the '847 Patent specification expressly contemplates. As shown in Exhibit B to this declaration, the '847 Patent specification states, at Column 3, lines 15–40:
"The waveguide multiplexer may take various forms known in the art, including but not limited to arrayed waveguide gratings, cascaded Mach-Zehnder interferometers, and ring-resonator-based multiplexers. The selection of a particular multiplexer architecture may be guided by considerations of insertion loss, channel spacing, free spectral range, and fabrication complexity, as would be understood by those skilled in the art."
40. This language is unambiguous. The specification expressly states that the "waveguide multiplexer" is not limited to any single architecture and identifies three different types of multiplexer architectures by name. Dr. Sundaram's proposed construction, which limits the term to "an arrayed waveguide grating," contradicts this express specification language.
41. Second, the prosecution history of the '847 Patent contains no narrowing amendment or argument that would limit the term "waveguide multiplexer" to arrayed waveguide gratings. At no point during prosecution did the applicant disclaim or disavow coverage of ring-resonator-based or Mach-Zehnder-based multiplexers. The absence of any such prosecution history disclaimer further confirms that the broad, generic meaning of "waveguide multiplexer" was intended and maintained throughout prosecution.
42. Third, Dr. Sundaram's reliance on the Predding & Knights textbook definition is problematic because the 2015 third edition of that textbook postdates the '847 Patent's September 14, 2016 filing date by only one year but reflects a more specialized usage of "waveguide multiplexer" that had evolved in certain academic contexts and does not represent the ordinary meaning of the term at the time of the invention. Moreover, even if this textbook were considered, it is a general photonics reference work that describes multiple types of waveguide-based multiplexers; Dr. Sundaram selectively quotes only the passage addressing arrayed waveguide gratings while ignoring other passages in the same textbook that describe ring-resonator and Mach-Zehnder architectures as alternative forms of waveguide multiplexing.
43. Fourth, Dr. Sundaram cites the Pathak et al. article on arrayed waveguide grating design, but that article was published in 2018 — nearly two years after the '847 Patent's September 14, 2016 filing date. Reliance on a post-filing-date reference to define the scope of a claim term is inherently suspect because it may reflect developments, refinements, or shifts in terminology that occurred after the relevant date.
44. For these reasons, OWS's proposed construction — "an optical device that combines multiple optical signals of different wavelengths onto a single waveguide" — is the construction that faithfully reflects the intrinsic record. This construction captures the functional essence of a waveguide multiplexer as described in the '847 Patent specification without improperly limiting the term to a single architectural implementation.
2. "Integrated on a Silicon Substrate" (the '193 Patent).
45. Dr. Sundaram proposes that the term "integrated on a silicon substrate," as used in the claims of the '193 Patent, should be construed to mean "monolithically fabricated on a single silicon die without hybrid bonding or heterogeneous integration." (Sundaram Decl. ¶ 39.) In support of this proposed construction, Dr. Sundaram relies on an IEEE standards document (IEEE Standard 802.3bs-2017) and an industry whitepaper published by a silicon-photonics foundry consortium. (Sundaram Decl. ¶¶ 40–44.)
46. This proposed construction is directly contradicted by the '193 Patent specification itself.
47. The '193 Patent, filed April 7, 2017 (claiming priority to Provisional Application No. 62/499,117, filed the same date) and issued January 12, 2021, describes multiple approaches to integrating optical and electronic components on a silicon substrate. As shown in Exhibit C to this declaration, Column 4, lines 32–45 of the '193 Patent specification states:
"In certain embodiments, active gain elements comprising III-V compound semiconductor materials may be integrated onto the silicon substrate through wafer-level bonding or die-level bonding processes. Such hybrid integration approaches allow the combination of the superior optical gain characteristics of III-V materials with the mature, high-volume fabrication capabilities of silicon photonic platforms."
48. This passage expressly describes hybrid bonding as a form of integration "onto the silicon substrate." Dr. Sundaram's proposed construction, which excludes "hybrid bonding or heterogeneous integration," directly contradicts this specification language. A construction that excludes subject matter expressly described in the patent specification as falling within the scope of the invention is improper under well-established claim-construction principles.
49. Furthermore, the prosecution history of the '193 Patent does not contain any disclaimer, disavowal, or narrowing amendment regarding the mode of integration. The term "integrated on a silicon substrate" was present in the original claims as filed and was never amended or narrowed during prosecution. The examiner did not raise any issue with respect to the breadth of the term "integrated," and the applicant made no representations that would limit "integrated" to monolithic fabrication.
50. Dr. Sundaram's reliance on IEEE Standard 802.3bs-2017 is also misplaced. IEEE standards are voluntary, industry-consensus documents that define technical requirements for interoperability purposes. They are not authoritative sources for the construction of patent claim terms, and the fact that an IEEE standard may use the term "integrated" in a particular way does not control or even meaningfully inform the construction of the same term in a patent specification that uses the term differently. Here, the '193 Patent specification's own usage of "integrated" is clear and unambiguous: it encompasses both monolithic and hybrid approaches.
51. Similarly, Dr. Sundaram's reliance on the industry whitepaper is problematic because the whitepaper represents the views of a particular industry consortium and does not establish the ordinary meaning of the term "integrated" as understood by the broader community of persons of ordinary skill in the art. The whitepaper's distinction between "monolithic integration" and "hybrid integration" as separate categories does not mean that "integrated on a silicon substrate" excludes hybrid approaches — it merely reflects a taxonomic convention used in certain industry contexts.
52. For these reasons, OWS's proposed construction — "formed on or within a silicon-based wafer using semiconductor fabrication processes" — is the construction that faithfully reflects the specification's breadth and is consistent with how a POSITA would understand the term.
3. "Micro-Channel Heat Sink" (the '562 Patent).
53. Dr. Sundaram proposes that the term "micro-channel heat sink," as used in the claims of the '562 Patent, should be construed to mean "a closed-loop liquid-cooling system with channels having a hydraulic diameter between 10 µm and 200 µm." (Sundaram Decl. ¶ 47.) In support of this construction, Dr. Sundaram cites a thermal-engineering reference work — specifically, Tuckerman & Pease, "High-Performance Heat Sinking for VLSI" (1981), as supplemented by the Tummidi et al. article in Journal of Electronic Packaging (2017). (Sundaram Decl. ¶¶ 48–51.)
54. Dr. Sundaram's proposed construction is improperly narrow in two respects: it requires (a) a "closed-loop liquid-cooling system" and (b) channels with a hydraulic diameter "between 10 µm and 200 µm." Neither limitation is supported by the intrinsic record.
55. As to the closed-loop requirement, the '562 Patent specification, filed November 22, 2014 and issued June 13, 2017, expressly describes both passive and active micro-channel configurations. As shown in Exhibit D to this declaration, Column 6, lines 12–28 of the '562 Patent specification states:
"In one embodiment, the micro-channel heat sink comprises a passive micro-channel structure in which heat generated by the photonic integrated circuit is conducted through the micro-channel walls to a heat spreader or substrate backside without active fluid circulation. In an alternative embodiment, the micro-channel heat sink comprises an active cooling system in which a coolant fluid is circulated through the micro-channels to remove heat by forced convection."
56. This passage makes clear that the patent contemplates both passive (heat conduction without active fluid circulation) and active (liquid-cooled) micro-channel configurations. Dr. Sundaram's proposed construction, which requires a "closed-loop liquid-cooling system," would exclude the passive micro-channel embodiment expressly described in the specification. This is improper.
57. As to the hydraulic diameter range, the '562 Patent specification describes micro-channels with hydraulic diameters "less than 1 mm" (Column 5, line 48) and does not impose the 10 µm to 200 µm range that Dr. Sundaram proposes. The 10 µm to 200 µm range is drawn from the Tuckerman & Pease reference, which is a seminal paper in the field of micro-channel cooling for VLSI circuits generally — not for photonic integrated circuits specifically. While the Tuckerman & Pease reference is an important historical contribution to the thermal-engineering field, it does not define the term "micro-channel heat sink" for purposes of construing the '562 Patent's claims. The patent specification's own definition of the relevant channel dimensions — "less than 1 mm" — is the controlling parameter, and Dr. Sundaram's attempt to substitute a narrower range from an extrinsic source is improper.
58. Moreover, the '562 Patent's continuation relationship with the '847 Patent (the two patents are in the same patent family) further confirms the broader understanding. The shared disclosure across these related patents treats micro-channels as a category defined by sub-millimeter hydraulic diameters, without any limitation to the 10 µm to 200 µm range that Dr. Sundaram proposes. Consistent construction of related patents in the same family supports OWS's broader reading.
59. It bears noting that the '562 Patent was originally assigned to Solara Optics, Inc., and was subsequently assigned to OWS pursuant to an asset purchase agreement dated June 15, 2018. This chain of title is undisputed and does not affect the construction of the claim terms, but I note it for completeness because Dr. Sundaram's declaration references the original assignee in passing.
60. For these reasons, OWS's proposed construction — "a heat-dissipating structure containing channels with a hydraulic diameter of less than 1 mm" — tracks the specification language and correctly reflects the scope of the term as used in the '562 Patent.
4. "Polarization-Diverse Grating Coupler" (the '015 Patent).
61. Dr. Sundaram proposes that the term "polarization-diverse grating coupler," as used in the claims of the '015 Patent, should be construed to mean "a two-dimensional grating coupler with a polarization-splitting element that simultaneously couples TE and TM modes." (Sundaram Decl. ¶ 55.) In support of this construction, Dr. Sundaram cites two journal articles — Chen et al., "Polarization-Diverse Grating Couplers on Silicon," Optics Express (June 2018), and an earlier article by Taillaert et al. on two-dimensional fiber grating couplers (2003) — and a definition from the Sterling Reach Industries Dictionary of Optical Engineering (2016 ed.). (Sundaram Decl. ¶¶ 56–60.)
62. Dr. Sundaram's proposed construction improperly narrows the term in three respects: it requires (a) a "two-dimensional" grating coupler, (b) a "polarization-splitting element" as a separate structural feature, and (c) "simultaneous" coupling of TE and TM modes. None of these limitations is supported by the intrinsic record.
63. The '015 Patent, filed August 30, 2019 (sharing Provisional Application No. 62/499,117 with the '193 Patent) and issued August 2, 2022, describes the polarization-diverse grating coupler as "capable of coupling light of two orthogonal polarization states" (Column 7, line 14). This functional description does not impose any structural requirement regarding dimensionality (one-dimensional vs. two-dimensional), nor does it require a separate "polarization-splitting element," nor does it specify whether the coupling of the two polarization states must occur simultaneously or may occur sequentially.
64. As shown in Exhibit E to this declaration, the '015 Patent specification describes multiple embodiments of the polarization-diverse grating coupler. One embodiment employs a two-dimensional grating coupler with integrated polarization splitting, in which the grating structure itself separates and couples light of two orthogonal polarization states into separate on-chip waveguides. Another embodiment employs a one-dimensional grating coupler in combination with a separate polarization rotator element, in which light of one polarization state is coupled by the grating coupler, rotated to the orthogonal polarization, and processed alongside light originally coupled in the other polarization state. Both embodiments are described as falling within the scope of the "polarization-diverse grating coupler" limitation. Dr. Sundaram's proposed construction would exclude the one-dimensional-grating-coupler-plus-rotator embodiment that the specification expressly describes.
65. Dr. Sundaram's requirement of a "polarization-splitting element" as a separate structural feature is not supported by the independent claims of the '015 Patent. The independent claims use the functional phrase "polarization-diverse" to modify "grating coupler," which describes the capability of the device (coupling light of two orthogonal polarization states) rather than imposing a specific structural limitation. Dr. Sundaram improperly converts a functional claim limitation into a structural one by importing a structural requirement from the extrinsic literature.
66. Dr. Sundaram's requirement of "simultaneous" coupling of TE and TM modes is similarly unsupported by the intrinsic record. The '015 Patent specification does not use the word "simultaneous" in connection with the polarization-diverse grating coupler. As noted above, the specification describes both simultaneous and sequential polarization-handling approaches, and there is no basis in the claims or specification for restricting the term to the simultaneous approach.
67. Furthermore, the prosecution history of the '015 Patent affirmatively supports the broader construction. During prosecution, the applicant amended dependent Claim 12 to add the limitation "two-dimensional sub-wavelength grating" in order to overcome a § 103 rejection combining Bergmann (U.S. Patent No. 8,412,007, issued April 2, 2013) with Simonsen (PCT/EP2015/071234, filed September 17, 2015). This amendment was specific to Claim 12 and was made to distinguish the prior art by specifying a particular grating configuration. Critically, the independent claims (Claims 1, 5, 8, and 16) and other dependent claims were not amended to include the "two-dimensional sub-wavelength grating" limitation, confirming that the broader term "polarization-diverse grating coupler" in those claims was not narrowed.
68. The doctrine of claim differentiation further supports this reading. The fact that dependent Claim 12 separately adds the "two-dimensional sub-wavelength grating" limitation confirms that the independent claim term "polarization-diverse grating coupler" does not inherently require a two-dimensional configuration. If the independent claim term already required a two-dimensional grating, then Claim 12's addition of that limitation would be superfluous — a result that is disfavored under established claim-construction principles.
69. For these reasons, OWS's proposed construction — "a grating coupler capable of coupling light of two orthogonal polarization states" — is consistent with both the specification and the claim-differentiation principle and should be adopted.
VII. RESPONSE TO SPECIFIC OPINIONS OF DR. SUNDARAM
A. Dr. Sundaram's Assertion That the '847 Patent Claims Are Limited to Arrayed Waveguide Grating (AWG) Architectures
70. In paragraphs 63–68 of his declaration, Dr. Sundaram opines that because the preferred embodiment in the '847 Patent specification describes an arrayed waveguide grating ("AWG"), the claim term "waveguide multiplexer" must be limited to AWG architectures. Dr. Sundaram devotes considerable attention to the specification's detailed description of the AWG embodiment, including its slab coupler design, waveguide phase array parameters, and channel-spacing characteristics, and concludes that the patentee intended "waveguide multiplexer" to mean an AWG.
71. This opinion is fundamentally flawed because it conflates the preferred embodiment with the full scope of the claims. It is axiomatic in patent law that claim terms are not limited to the preferred embodiment unless the specification clearly and unmistakably mandates such a limitation. See Phillips, 415 F.3d at 1323 ("[A]lthough the specification often describes very specific embodiments of the invention, we have repeatedly warned against confining the claims to those embodiments."). Here, the specification does not mandate any such limitation.
72. To the contrary, as I noted in paragraph 39 above, the '847 Patent specification expressly states that "the waveguide multiplexer may take various forms known in the art, including but not limited to arrayed waveguide gratings, cascaded Mach-Zehnder interferometers, and ring-resonator-based multiplexers." (Col. 3, ll. 15–40; see Ex. B hereto.) The phrase "including but not limited to" is a well-recognized signal that the listed items are exemplary, not exhaustive. Dr. Sundaram's opinion ignores this express disclaimer of limitation, which is fatal to his proposed construction.
73. Moreover, the '847 Patent specification includes, at Figures 4A, 4B, and 4C, schematic diagrams of three different waveguide multiplexer architectures: an AWG (Fig. 4A), a cascaded Mach-Zehnder interferometer (Fig. 4B), and a ring-resonator-based multiplexer (Fig. 4C). The inclusion of multiple architectural embodiments in the figures further confirms that the claims are not limited to any single architecture.
74. Dr. Sundaram's selective focus on the detailed description of the AWG embodiment — while ignoring the specification's express statement of breadth and the alternative embodiments depicted in the figures — is the kind of improper embodiment-based narrowing that the Federal Circuit has repeatedly cautioned against.
B. Dr. Sundaram's Assertion That "Integrated" Excludes Hybrid Bonding
75. In paragraphs 69–74 of his declaration, Dr. Sundaram contends that a POSITA would understand "integrated on a silicon substrate" to exclude hybrid bonding or heterogeneous integration because those techniques involve bonding a non-silicon material (such as indium phosphide or other III-V compound semiconductors) onto the silicon substrate. Dr. Sundaram argues that "integration" in the silicon-photonics context has a well-understood meaning that is limited to monolithic fabrication — that is, fabrication of all components on a single silicon die using complementary metal-oxide-semiconductor ("CMOS")-compatible processes without the addition of non-silicon materials through bonding.
76. This opinion is directly contradicted by the intrinsic record. As I explained in paragraph 47 above, Column 4, lines 32–45 of the '193 Patent specification expressly describes "integrating active gain elements comprising III-V compound semiconductor materials onto the silicon substrate through wafer-level bonding or die-level bonding processes." The specification's own use of the word "integrating" to describe bonding-based techniques forecloses Dr. Sundaram's proposed exclusion of hybrid bonding from the scope of "integrated on a silicon substrate."
77. Dr. Sundaram attempts to circumvent this specification language by arguing that the passage at Column 4, lines 32–45, describes an "alternative" approach that is distinct from the claimed "integrated" approach. (Sundaram Decl. ¶ 72.) But this characterization finds no support in the specification. The specification does not describe hybrid bonding as an "alternative to" integration; it describes hybrid bonding as a form of integration. The specification does not use qualifying or limiting language (such as "as an alternative to integration" or "in contrast to integration") when discussing hybrid bonding. Rather, it uses the word "integrated" itself to describe the hybrid bonding approach.
78. I also note that Dr. Sundaram's opinion on this point is driven in significant part by his improperly elevated POSITA definition. As I explained in Section V above, Dr. Sundaram's POSITA — a Ph.D. with five years of commercial foundry experience — might colloquially distinguish between "monolithic integration" and "hybrid bonding" as separate process categories, reflecting terminology conventions used within certain commercial foundry environments. But the fact that some foundry engineers may use these terms differently in colloquial industry parlance does not override the patent specification's own usage, which is the controlling source for claim construction. The specification uses "integrated" to encompass both monolithic and hybrid approaches, and this usage controls.
C. Dr. Sundaram's Assertion Regarding Closed-Loop Requirement for Micro-Channel Heat Sinks
79. In paragraphs 75–80 of his declaration, Dr. Sundaram asserts that a "micro-channel heat sink" necessarily implies a closed-loop fluid circulation system. He bases this opinion on the Tuckerman & Pease reference, which describes micro-channel heat sinks in the context of forced-convection liquid cooling, and on the Tummidi et al. article, which addresses active thermal solutions for silicon photonic transceivers.
80. This opinion is contradicted by the '562 Patent specification, which expressly describes passive micro-channel heat sink configurations. As I noted in paragraph 55 above, Column 6, lines 12–28 of the '562 Patent specification describes a "passive micro-channel structure in which heat generated by the photonic integrated circuit is conducted through the micro-channel walls to a heat spreader or substrate backside without active fluid circulation." This passage is unambiguous: the specification contemplates micro-channel heat sinks that operate through passive thermal conduction, without any liquid coolant or closed-loop circulation system.
81. Dr. Sundaram's reliance on the Tuckerman & Pease reference — while historically significant in the micro-channel cooling literature — is misplaced. The Tuckerman & Pease reference addresses a broader class of micro-channel heat sinks for very-large-scale integration ("VLSI") circuits, not photonic integrated circuits specifically. The reference's focus on forced-convection liquid cooling reflects the application context in which it was developed (high-power VLSI chips) and does not establish that "micro-channel heat sink" universally requires liquid cooling across all technology domains. The '562 Patent specification defines its own usage of the term, and that usage encompasses both passive and active configurations.
82. Similarly, the Tummidi et al. article describes advanced thermal solutions for silicon photonic transceivers and discusses micro-channel cooling as one option among several. Dr. Sundaram selectively cites the portions of Tummidi et al. that address liquid-cooled micro-channels while ignoring the article's own acknowledgment that micro-channel structures can be used in both active and passive thermal management contexts. Even if Tummidi et al. were relevant extrinsic evidence, it would not support the narrow construction Dr. Sundaram proposes.
D. Dr. Sundaram's Assertion Regarding Simultaneous TE/TM Coupling Requirement
83. In paragraphs 81–86 of his declaration, Dr. Sundaram argues that "polarization-diverse" requires "simultaneous" coupling of TE and TM modes, and that a sequential or switched polarization-handling approach would not qualify as "polarization-diverse." He bases this opinion on the Chen et al. article and the Taillaert et al. article, both of which describe two-dimensional grating couplers that simultaneously couple both polarization states.
84. This opinion is not supported by the intrinsic record. The '015 Patent specification uses the phrase "capable of coupling light of two orthogonal polarization states" without specifying whether such coupling must occur simultaneously or may occur sequentially. As I explained in paragraph 64 above, the specification describes both simultaneous and sequential embodiments, and neither the claims nor the specification impose a simultaneity requirement.
85. The term "polarization-diverse" is a well-known functional descriptor in the silicon-photonics field. As a POSITA (under either my proposed definition or Dr. Sundaram's) would understand, "polarization-diverse" refers to the capability of handling or processing light of multiple polarization states, without specifying the mechanism by which that capability is achieved. A grating coupler that handles polarization diversity through sequential rotation and coupling is just as "polarization-diverse" as one that achieves it through simultaneous two-dimensional coupling. Both approaches result in a device that is capable of coupling light of two orthogonal polarization states — which is precisely the functional capability claimed.
86. Dr. Sundaram's reliance on the Chen et al. and Taillaert et al. articles is problematic because both articles describe specific implementations of polarization-diverse grating couplers (two-dimensional grating designs) that represent one approach among several. The fact that certain journal articles describe two-dimensional grating couplers as "polarization-diverse" does not mean that the term is limited to two-dimensional designs. The articles describe a species of a broader genus; they do not define the genus itself.
87. Moreover, the Chen et al. article was published in June 2018 — after the April 7, 2017 filing date of the provisional application to which the '015 Patent claims priority. While the article predates the '015 Patent's own filing date of August 30, 2019, the relevant date for assessing the meaning of the term is the earliest effective filing date, which is April 7, 2017 (the date of Provisional Application No. 62/499,117). Dr. Sundaram does not address this temporal issue.
VIII. ADDITIONAL OBSERVATIONS ON CLAIM CONSTRUCTION RELEVANT INTRINSIC EVIDENCE
A. Prosecution History of the '193 Patent
88. During prosecution of the '193 Patent, the examiner issued a § 102 rejection of certain claims over Takahashi et al. ("Dense WDM Silicon Photonic MUX," IEEE Photonics Technology Letters, March 2014). In response, the applicant amended the claims to narrow the "coherent detection module" limitation, adding specificity regarding the architecture of the coherent detection module to distinguish the claimed invention from the Takahashi et al. reference. This amendment is documented in the prosecution history and is attached as Exhibit F to this declaration.
89. Critically, this narrowing amendment pertained specifically and exclusively to the "coherent detection module" limitation. The term "integrated on a silicon substrate" was not amended, was not the subject of the examiner's rejection, and was not discussed in the applicant's response to the § 102 rejection. There is no basis in the prosecution history for reading the narrowing of the "coherent detection module" limitation as a narrowing of the separate term "integrated on a silicon substrate."
90. Nevertheless, Dr. Sundaram appears to suggest, at paragraphs 73–74 of his declaration, that the general context of prosecution — including the applicant's effort to distinguish Takahashi et al. — somehow supports a narrow reading of "integrated on a silicon substrate." This suggestion conflates two distinct claim limitations and is unsupported by the prosecution record. A narrowing amendment to one claim limitation does not operate as a general narrowing of all terms in the same patent, and the prosecution history of the '193 Patent does not support Dr. Sundaram's proposed construction of "integrated on a silicon substrate."
B. Prosecution History of the '015 Patent
91. As I discussed in paragraphs 67–68 above, during prosecution of the '015 Patent, the applicant amended dependent Claim 12 to add the limitation "two-dimensional sub-wavelength grating" in order to overcome a § 103 rejection combining Bergmann (U.S. Patent No. 8,412,007, issued April 2, 2013) with Simonsen (PCT/EP2015/071234, filed September 17, 2015). The relevant prosecution history excerpt is attached as Exhibit G to this declaration.
92. This amendment was specific to Claim 12. The independent claims (Claims 1, 5, 8, and 16) and other dependent claims were not amended to include the "two-dimensional sub-wavelength grating" limitation. The examiner's reasons for allowance, as stated in the Notice of Allowance, do not attribute the allowance of the independent claims to any understanding that "polarization-diverse grating coupler" is limited to two-dimensional configurations.
93. Dr. Sundaram's proposed construction of "polarization-diverse grating coupler" — which requires a "two-dimensional grating coupler with a polarization-splitting element that simultaneously couples TE and TM modes" — effectively imports the dependent Claim 12 limitation into the independent claims. This violates the well-established doctrine of claim differentiation, which creates a presumption that each claim in a patent has a different scope and that a dependent claim's additional limitation would be superfluous if it were already inherent in the independent claim. See, e.g., Curtiss-Wright Flow Control Corp. v. Velan, Inc., 438 F.3d 1374, 1381 (Fed. Cir. 2006). If "polarization-diverse grating coupler" already required a two-dimensional grating, there would be no reason for dependent Claim 12 to separately add that limitation.
C. Continuation Relationship Between the '562 and '847 Patents
94. The '562 Patent and the '847 Patent are in the same patent family. The '847 Patent is a continuation of the '562 Patent, and the two patents share a common disclosure regarding the thermal management of photonic integrated circuits, including the use of micro-channel heat sink structures.
95. Under established principles of claim construction, related patents in the same family should be construed consistently with respect to shared or analogous claim terms. See, e.g., NTP, Inc. v. Research in Motion, Ltd., 418 F.3d 1282, 1293 (Fed. Cir. 2005). The common disclosure in the '562 and '847 Patents uses the term "micro-channel" consistently to refer to channels with sub-millimeter hydraulic diameters, without any limitation to the 10 µm to 200 µm range that Dr. Sundaram proposes. The continuation relationship and shared disclosure further support OWS's proposed construction of "micro-channel heat sink."
96. Dr. Sundaram's declaration does not address the continuation relationship between the '562 and '847 Patents or its implications for consistent construction. This is a significant omission, because the shared specification language across the two patents provides additional intrinsic evidence supporting the broader construction.
D. Common Provisional Application for the '193 and '015 Patents
97. The '193 Patent (filed April 7, 2017) and the '015 Patent (filed August 30, 2019) both claim priority to Provisional Application No. 62/499,117, filed April 7, 2017. The provisional application discloses, among other things, the silicon-substrate integration approaches and the polarization-diverse grating coupler architectures that are at issue in the claim-construction disputes involving those two patents.
98. The provisional application provides additional context for understanding the scope of terms common to both the '193 and '015 Patents. In particular, the provisional application uses the terms "integrated on a silicon substrate" and "polarization-diverse grating coupler" in the same broad manner as the issued patents, without any of the structural or process-specific limitations that Dr. Sundaram seeks to import through his proposed constructions.
99. Dr. Sundaram does not meaningfully engage with the provisional application in his declaration. Although he cites the provisional application in his list of materials reviewed (Sundaram Decl. ¶ 10(f)), he does not analyze its disclosure or explain how it supports (or contradicts) his proposed constructions. The provisional application is part of the intrinsic record and should be considered in construing the claims of the '193 and '015 Patents.
IX. IMPACT OF DR. SUNDARAM'S PROPOSED CONSTRUCTIONS ON THE INFRINGEMENT ANALYSIS
100. I recognize that claim construction is a legal determination for the Court, and I offer the observations in this section solely to assist the Court in understanding the practical implications of the competing proposed constructions. The ultimate resolution of the claim-construction disputes is, of course, a matter for the Court to decide.
101. That said, the practical impact of Dr. Sundaram's proposed constructions is to exclude Lumicore's accused products from the scope of the asserted claims in a manner that is not supported by the intrinsic record of the asserted patents. In my professional opinion, Dr. Sundaram's constructions are results-oriented — they are crafted to achieve a particular outcome in the infringement analysis rather than to faithfully reflect the ordinary meaning of the claim terms as established by the intrinsic evidence.
102. "Waveguide Multiplexer." If Dr. Sundaram's construction of "waveguide multiplexer" is adopted — limiting the term to "an arrayed waveguide grating that combines optical signals using a slab coupler and phase-shifted waveguide array" — it would exclude the multiplexer architectures used in Lumicore's LumiLink 400G-ZR+ and LumiLink 800G-DR8 products. The Benchmark Semiconductor Analysis, LLC teardown reports for these products identify ring-resonator-based wavelength multiplexing architectures that fall squarely within the '847 Patent's disclosure but would be excluded by Dr. Sundaram's narrow construction. The LumiLink 400G-ZR+ generated FY 2023 revenue of approximately $94.8 million, and the LumiLink 800G-DR8 generated FY 2023 revenue of approximately $128.4 million.
103. "Integrated on a Silicon Substrate." If "integrated on a silicon substrate" is limited to monolithic fabrication — excluding hybrid bonding and heterogeneous integration — it would effectively exclude all three accused product lines. The Benchmark Semiconductor Analysis teardown reports for the LumiLink 400G-ZR+, LumiLink 800G-DR8, and LumiEdge 1600G-LR consistently identify hybrid-integration techniques in which III-V compound semiconductor gain elements are bonded onto a silicon photonic substrate. This is precisely the approach described in the '193 Patent specification at Column 4, lines 32–45. The combined FY 2023 accused-product revenue for all three product lines is approximately $264.9 million.
104. "Micro-Channel Heat Sink." If "micro-channel heat sink" is narrowed to a closed-loop liquid cooling system with hydraulic diameters of 10 µm to 200 µm, it would exclude the passive micro-channel thermal management structures observed in the Benchmark Semiconductor Analysis teardowns of the LumiLink 400G-ZR+ (teardown report dated December 12, 2022, lead analyst Kevin Ng) and LumiLink 800G-DR8. Those teardown reports identify passive micro-channel structures with hydraulic diameters in the range of approximately 300 µm to 600 µm — well within the '562 Patent specification's "less than 1 mm" threshold but outside the 10 µm to 200 µm range that Dr. Sundaram proposes.
105. "Polarization-Diverse Grating Coupler." If "polarization-diverse grating coupler" requires simultaneous two-dimensional TE/TM coupling, it would exclude the sequential polarization-handling architectures observed in the LumiLink 800G-DR8 and LumiEdge 1600G-LR products. The teardown reports for those products identify grating coupler architectures that employ a one-dimensional grating coupler with a separate on-chip polarization rotator — a configuration that achieves polarization diversity sequentially rather than simultaneously. The LumiEdge 1600G-LR generated FY 2023 revenue of approximately $41.7 million.
106. In each of these instances, Dr. Sundaram's proposed construction would exclude accused products that fall within the scope of the asserted patents' express disclosures. This pattern — in which every proposed construction conveniently excludes the accused products from infringement — reinforces my opinion that Dr. Sundaram's constructions are driven by the desired litigation outcome rather than by a faithful reading of the intrinsic record.
X. CONCLUSIONS
107. For the reasons set forth in this Supplemental Declaration, I hold the following opinions to a reasonable degree of professional certainty:
108. POSITA Definition. Dr. Sundaram's proposed POSITA definition — requiring a Ph.D. and at least five years of direct industry experience in the design and fabrication of silicon photonic integrated circuits at a commercial foundry or semiconductor company — is improperly elevated and does not accurately reflect the level of ordinary skill in the art durin