PRIVILEGED AND CONFIDENTIAL ATTORNEY WORK PRODUCT — FOR INTERNAL USE ONLY
CALDERWOOD & HARKNESS LLP INTELLECTUAL PROPERTY LITIGATION
INTERNAL MEMORANDUM
May 15, 2023
TO: Calderwood & Harkness LLP, intellectual Property Litigation Group — Partners and Senior Associates
FROM: Alan Ngo (Responsible Partner), Omar Siddiqui (Lead Associate), Christy Bauer (Working Associate)
RE: Post-Mortem Analysis and Lessons Learned — Optiwave Semiconductor Corp. v. Lumos Photonics, Inc., Case No. 5:22-cv-03417-EJD (N.D. Cal.)
C&H Matter No.: 1014-00003
Distribution: Internal circulation only — IP Litigation Group; copies to Marcus Delgado (Billing Partner), Practice Group Leadership
Right-click to update Table of Contents
I. Executive Summary
The purpose of this memorandum is to provide a candid internal assessment of the Optiwave Semiconductor Corp. v. Lumos Photonics, Inc. patent infringement litigation for the benefit of the IP Litigation Group and to identify specific, actionable lessons for improving firm practices in future plaintiff-side patent enforcement matters. This memorandum is protected by the attorney work product doctrine and is intended solely for internal use within Calderwood & Harkness LLP. It reflects the frank and critical self-assessment that the work product doctrine is designed to protect and should not be shared with any person or entity outside the firm without the express authorization of the undersigned.
Calderwood & Harkness LLP ("C&H") opened Matter No. 1014-00003 on June 3, 2022, and filed the complaint on the same date in the United States District Court for the Northern District of California, San Jose Division, before the Honorable Edward J. Torvelis. The matter involved the assertion of four U.S. utility patents by our client, Optiwave Semiconductor Corp. ("OWS"), a fabless silicon photonics company headquartered in San Jose, California, against Lumos Photonics, Inc. ("Lumos"), an integrated device manufacturer publicly traded on the Nasdaq Stock Market (ticker: LMPH) and headquartered in Austin, Texas. OWS is majority-owned (approximately 62%) by Vantage Point Capital Partners Fund IV, LP, a growth-equity fund with a significant portfolio in the semiconductor and photonics sectors.
The four United States patents asserted in the litigation were:
• U.S. Patent No. 10,234,567 (the "'567 Patent"), directed to adiabatic coupling structures for silicon photonic waveguides;
• U.S. Patent No. 10,891,234 (the "'234 Patent"), directed to germanium-on-silicon photodetector architectures with graded buffer layers;
• U.S. Patent No. 11,045,678 (the "'678 Patent"), directed to monolithically integrated silicon photonic transceivers; and
• U.S. Patent No. 11,312,890 (the "'890 Patent"), directed to co-integrated modulator driver circuits for optical engines.
The accused products were the LumosBridge™ 800G Optical Engine and the LumosEdge™ 400ZR+ Pluggable Transceiver, both of which are high-volume commercial products in the data center interconnect market. OWS's initial damages claim exceeded $75 million, based on a 7.5% reasonable royalty rate applied to $148.6 million in accused product revenue, with additional claims for willfulness enhancement and pre-filing damages.
The case reached its decisive inflection point at the Markman hearing, held on January 30, 2023. The Court issued its Markman Order on February 17, 2023. In a 34-page opinion, the Court adopted Defendant's proposed construction — or a construction substantially aligned with Defendant's position — on eight of eleven disputed claim terms. The Court adopted Plaintiff's proposed construction on two terms and adopted a modified construction on one term. The effect of the Markman Order was devastating: the Court's narrow constructions effectively excluded Lumos's accused products from the scope of the asserted claims on three of the four patents and significantly weakened the infringement theory on the fourth.
Our internal post-Markman assessment, completed on February 21, 2023, concluded that infringement theories were "untenable as to three of four patents and significantly weakened as to the fourth." Following a Board meeting on February 24, 2023, OWS authorized settlement discussions. After approximately three weeks of negotiations, the case was resolved via a walk-away settlement: zero payment in either direction, mutual general releases, no license grant, and dismissal with prejudice. The Order of Dismissal was entered on April 18, 2023.
The total litigation spend was $2,549,568, comprising C&H fees of $1,847,312, costs and disbursements of $289,456, and expert fees of $412,800 (Whitfield Technical Consulting Group, Inc. and Apex Claim Analytics LLC combined). This spend was incurred against a Board-approved litigation budget of $4.2 million through trial. The client received zero recovery. The outcome represents both a financial loss and a reputational and strategic setback for OWS in the competitive silicon photonics market.
This memorandum identifies systemic lessons in five principal areas:
1. Pre-suit claim construction risk assessment;
2. Patent prosecution counseling and its intersection with litigation enforcement;
3. Litigation investment decision framework, including probability-weighted analysis;
4. Expert engagement timing and utilization; and
5. Client communications and expectation management.
Each of these areas is discussed in detail below, followed by specific recommendations for incorporation into the IP Litigation Group's standard practices going forward.
II. Case Summary and Procedural History
A. Pre-Litigation Phase
OWS maintains an extensive patent portfolio concentrated in the silicon photonics domain. At the time of filing, the portfolio comprised 87 issued U.S. utility patents and 14 pending applications. The portfolio is concentrated in three core technology areas: waveguide coupling architectures, germanium-on-silicon photodetectors, and modulator driver circuits. The portfolio was developed primarily through internal research and development conducted at OWS's San Jose facility, with named inventors including Dr. Kevin Fong (Chief Technology Officer), Dr. Amara Osei (Director of Photonics Design), and Hideo Tanaka (formerly Senior Staff Engineer, employed at OWS from 2016 to 2019).
Prior to litigation, OWS engaged Clearpath IP Valuation Services, Inc. to conduct a portfolio-level valuation. Jennifer Armitage, CFA, CLP, of Clearpath prepared a valuation report assessing the total portfolio at $215 million and the four patents ultimately asserted in the litigation at $42.8 million collectively. This valuation was used by OWS management and the Board of Directors in evaluating the potential return on a litigation investment. As discussed in Section V below, the Clearpath valuation did not incorporate a probability-weighted assessment of litigation risk, including claim construction risk, and thus may have contributed to an overly optimistic assessment of the litigation's expected value.
On March 15, 2022, OWS sent a cease-and-desist letter to Lumos regarding the '567 Patent and the '234 Patent. The letter was drafted by Ted Marchetti, Vice President of intellectual Property at OWS, with assistance from C&H. The letter identified specific features of the LumosBridge™ 800G Optical Engine and the LumosEdge™ 400ZR+ Pluggable Transceiver that OWS believed infringed the identified patents and demanded that Lumos either cease the infringing activities or enter into a license agreement.
On April 4, 2022, Lumos responded through outside counsel, denying infringement and asserting that OWS's patents were invalid over the prior art. Lumos's response was detailed and technically specific, identifying several prior art references and challenging the scope of OWS's patent claims. In retrospect, the substantive quality of Lumos's pre-litigation response should have been an early signal of the strength of the defense team's preparation and technical sophistication.
During the period from April through May 2022, C&H prepared pre-suit infringement analyses for all four patents and a consolidated validity risk assessment. These analyses, discussed in greater detail in Section III, identified claim construction vulnerabilities for each of the four patents but assessed the risk of adverse narrow constructions as moderate rather than high. The analyses were prepared by the litigation team, with input from C&H's patent prosecution group on prosecution history issues.
On May 25, 2022, C&H delivered a comprehensive litigation strategy memorandum to the OWS Board of Directors and to representatives of Vantage Point Capital Partners, including Douglas Fairbanks III, Managing Partner, and Rebecca Torres, Operating Partner for the Semiconductor Sector. The memorandum recommended filing suit in the Northern District of California and presented projected damages exceeding $75 million, balanced against a projected litigation budget of approximately $4.2 million through trial. The memorandum characterized the overall claim construction risk as "manageable" — an assessment that, as discussed below, proved to be insufficiently calibrated to the actual risks presented by the prosecution histories.
During this pre-litigation period, OWS and Vantage Point also evaluated the potential for third-party litigation funding. Resolute Litigation Finance, LP conducted due diligence and provided a term sheet offering to fund a portion of the litigation costs in exchange for a contingent return on any recovery. Ultimately, OWS and Vantage Point declined the Resolute term sheet and elected to self-fund the litigation. The implications of this decision are discussed further in Section V.C.
On April 28, 2022, the OWS Board of Directors formally authorized the litigation, with the concurrence of Vantage Point Capital Partners. On May 12, 2022, C&H executed the engagement letter with OWS, and the matter was assigned internal matter number 1014-00003.
B. Litigation Phase
C&H filed the complaint on June 3, 2022, asserting infringement of all four patents by Lumos's LumosBridge™ 800G Optical Engine and LumosEdge™ 400ZR+ Pluggable Transceiver products. The complaint was filed in the United States District Court for the Northern District of California, San Jose Division, and was assigned to the Honorable Edward J. Torvelis. The complaint sought damages for patent infringement under 35 U.S.C. § 284, enhanced damages for willful infringement under 35 U.S.C. § 284, attorneys' fees under 35 U.S.C. § 285, and a permanent injunction under 35 U.S.C. § 283.
Lumos retained Brannick Howell & Stein LLP ("BHS") as litigation counsel. BHS assembled a formidable team led by Victoria "Tori" Ashworth, a partner with 18 years of intellectual property litigation experience and a reputation for aggressive and technically sophisticated Markman briefing. The associate team was led by Daniel Kowalski, who holds a PhD in Optics from the University of Rochester, providing BHS with deep in-house technical expertise on the precise photonics technologies at issue. BHS also retained Dr. Raymond Ng, formerly a professor at the University of Colorado Boulder and currently of counsel to BHS, as a technical consultant and tutorial witness. As discussed in Section VIII, the technical depth and coordination of BHS's team proved to be a significant factor in the outcome of the Markman hearing.
On July 29, 2022, Lumos filed its Answer and Counterclaims. The counterclaims sought declaratory judgments of non-infringement and invalidity under 35 U.S.C. §§ 102 (anticipation), 103 (obviousness), and 112 (written description and enablement). The invalidity counterclaims were supported by a detailed identification of prior art references and prosecution history arguments, further signaling the depth of BHS's technical preparation.
The Initial Case Management Conference was held on August 15, 2022. Judge Torvelis set a Markman hearing date of January 30, 2023, and established a claim construction briefing schedule. The Court's scheduling order reflected a front-loaded approach to claim construction that is characteristic of the Northern District of California's Patent Local Rules.
The claim construction phase proceeded on the following timeline:
• September 12, 2022: Plaintiff's Preliminary Claim Construction Contentions served.
• October 3, 2022: Defendant's Responsive Claim Construction Contentions served.
• October 14, 2022: Joint Claim Construction and Prehearing Statement filed, identifying eleven disputed claim terms across the four asserted patents.
• October 31, 2022: Claim construction discovery closed. During this period, depositions were taken of named inventors Dr. Kevin Fong, Dr. Amara Osei, and Hideo Tanaka. Documentary discovery focused on prosecution histories, inventor notebooks, technical publications, and industry standards relevant to the disputed claim terms.
• December 5, 2022: Plaintiff's Opening Claim Construction Brief filed (45 pages). The brief presented OWS's proposed constructions for all eleven disputed terms, supported by intrinsic evidence from the patent specifications and prosecution histories, as well as extrinsic evidence from industry standards, technical publications, and expert testimony.
• December 19, 2022: Defendant's Responsive Claim Construction Brief filed. BHS's brief was highly effective in leveraging the prosecution history record against OWS's proposed constructions, as discussed in Sections III and VIII below.
• January 9, 2023: Plaintiff's Reply Claim Construction Brief filed (15 pages).
• January 19, 2023: Defendant's Sur-Reply filed by leave of Court.
Dr. Harold Whitfield of the Whitfield Technical Consulting Group, Inc. served as OWS's technical tutorial witness and submitted a declaration in support of OWS's proposed constructions with the Opening Claim Construction Brief. Dr. Whitfield, who holds a PhD in Electrical Engineering (photonics concentration) from the University of California, Berkeley, and is Professor Emeritus of Electrical and Computer Engineering at the University of California, Santa Barbara, brought impeccable academic credentials and extensive experience in the integrated photonics field, with over 200 peer-reviewed publications. His declaration addressed the understanding of a person of ordinary skill in the art ("POSITA") regarding each of the disputed claim terms and supported broader constructions that would have encompassed Lumos's accused product architectures.
The Markman hearing was held on January 30, 2023, and lasted approximately six and a half hours. Both sides presented technical tutorials: Dr. Whitfield for OWS and Dr. Ng for Lumos. The tutorials addressed the underlying photonics technology, the specific designs of the accused products, and the technical distinctions relevant to claim construction. The hearing included oral argument on each of the eleven disputed terms.
C. Markman Order and Its Impact
On February 17, 2023, Judge Torvelis issued a 34-page Markman Order construing all eleven disputed claim terms. The Order adopted Defendant's proposed construction or a construction substantially aligned with Defendant's position on eight of the eleven terms. The Court adopted Plaintiff's proposed construction on two terms and adopted a modified construction (closer to Defendant's position) on one term.
The constructions that proved fatal to OWS's infringement theories are summarized below. Each of these constructions excluded the specific technical features of Lumos's accused products from the scope of the asserted claims:
'567 Patent — "adiabatically coupled": The Court adopted the construction: "coupled through a gradual, continuous taper designed and configured to prevent excitation of higher-order modes over the entire transition length." This construction required a continuous taper geometry and expressly excluded Lumos's segmented (non-continuous) taper design, which employs discrete coupling sections in the LumosBridge™ 800G Optical Engine. The Court's analysis relied heavily on the prosecution history, noting that the "adiabatically coupled" limitation was added by amendment to overcome the Tanaka prior art (U.S. Patent No. 9,456,012), and that the applicant's remarks during prosecution described the claimed invention as employing a "gradual, continuous taper" in contrast to the prior art's segmented approach.
'234 Patent — "continuously varying germanium concentration": The Court adopted verbatim Defendant's proposed construction: "a buffer layer in which the germanium atomic fraction increases monotonically and without step changes from the silicon substrate interface to the absorber layer." This construction excluded Lumos's step-graded SiGe buffer layer architecture, which employs four discrete composition steps with germanium fractions of 0.25, 0.50, 0.75, and 1.0. The Court reasoned that the claim language "continuously varying" was unambiguous on its face and that resort to the specification to broaden its scope was not warranted. The Court noted that while the specification described both "continuously graded" and "step-graded" embodiments, the claims used only the "continuously varying" language, and the Court would not read limitations out of the claims or expand their plain meaning to encompass embodiments described only in the specification.
'678 Patent — "monolithically formed on a single silicon die": The Court adopted the construction: "fabricated entirely within the same front-end-of-line and back-end-of-line semiconductor processing sequence on a single silicon substrate without heterogeneous bonding or attachment of separately fabricated optical components." This construction excluded Lumos's architecture, which employs micro-transfer-printed III-V gain elements that are separately fabricated and then bonded to the silicon photonic integrated circuit (PIC). The Court's analysis relied on the prosecution history, in which the applicant added the "monolithically formed" limitation by amendment to overcome prior art references based on hybrid integration approaches. The Court quoted the applicant's prosecution remarks — "the present invention is distinguished from heterogeneous integration approaches" — and concluded that the applicant had disclaimed hybrid and heterogeneous integration methods.
'890 Patent — "co-integrated": The Court adopted Defendant's proposed construction requiring "fabrication in the same semiconductor process, sharing doping profiles and process layers, without bond wires, interposers, or package-level routing." This construction excluded Lumos's design, which employs a separate CMOS driver die that is flip-chip bonded to the photonic die via copper pillar micro-bumps. The Court again relied on the prosecution history, noting that the "co-integrated" limitation was added during prosecution specifically to distinguish prior art designs in which driver circuits are separately fabricated and bonded to the photonic die.
The cumulative effect of these constructions was that Lumos's accused products fell outside the scope of the asserted claims under any reasonable infringement analysis. For three of the four patents ('567, '678, and '890), the constructions were outcome-determinative — no viable infringement theory survived. For the '234 Patent, the construction significantly weakened the infringement theory, though a narrow argument regarding Lumos's continuous-composition prototype (never commercialized) might have survived; however, this theory would have involved minimal damages and was not a viable basis for continuing the litigation.
D. Settlement and Closure
Following the issuance of the Markman Order, the litigation team moved quickly to assess the impact and advise the client. On February 21, 2023 — four days after the Order — C&H completed and delivered an internal post-Markman assessment memorandum concluding that the infringement theories were untenable as to three of four patents and significantly weakened as to the fourth. The memorandum recommended that OWS pursue settlement discussions and outlined the range of likely settlement outcomes, from a nominal cross-license to a walk-away dismissal.
On February 24, 2023, the OWS Board of Directors convened a special meeting to review the post-Markman assessment. The C&H litigation team presented the Markman Order's impact on each patent, the remaining litigation options (including the possibility of an interlocutory appeal under 28 U.S.C. § 1292(b), which was assessed as unlikely to be granted and unlikely to change the outcome), and a recommendation to pursue settlement. The Board authorized settlement discussions.
Settlement negotiations took place between March 6 and March 28, 2023. OWS initially proposed a cross-license arrangement under which Lumos would pay a nominal royalty of $500,000 in exchange for a license to the four asserted patents, with OWS receiving a reciprocal license to a subset of Lumos's optical interconnect patents. Lumos rejected this proposal, taking the position that the Markman Order had effectively established non-infringement and that no payment or license was warranted. Lumos counteroffered a walk-away settlement with mutual general releases and dismissal with prejudice.
On March 28, 2023, the OWS Board of Directors and Vantage Point Capital Partners approved the walk-away settlement terms, concluding that the costs and risks of continuing the litigation — including the near-certainty of summary judgment of non-infringement and the possibility of an exceptional-case finding under 35 U.S.C. § 285 — outweighed any potential benefit.
The Settlement Agreement and Mutual Release was executed on April 5, 2023, signed by Sandra Ellison (General Counsel) on behalf of OWS and by Richard Talbot (Senior Vice President and General Counsel) on behalf of Lumos. The agreement provided for zero payment in either direction, mutual general releases of all claims and counterclaims, no license grant, no admission of validity or invalidity, and dismissal of the action with prejudice. A Joint Stipulation of Dismissal with Prejudice was filed on April 10, 2023, and the Court entered the Order of Dismissal on April 18, 2023.
The matter was closed on C&H's docket following the entry of the Order of Dismissal. Final financial reconciliation was completed in May 2023. Total retainer deposits received from OWS were $400,000, paid in four quarterly installments of $100,000 each. A retainer balance of $62,488 was returned to OWS. The net retainer applied to fees and costs was $337,512. The remaining fees of $1,509,800 were collected via monthly invoices throughout the engagement.
III. Lessons Learned: Pre-Suit Claim Construction Risk Assessment
A. Inadequate Weighting of Narrow Construction Risk
The single most significant lesson from this matter is that our pre-suit infringement analyses — prepared during April and May 2022 and incorporated into the May 25, 2022 litigation strategy memorandum — failed to adequately weight the risk of adverse narrow claim constructions. This is the core finding of this post-mortem assessment, and the one with the most direct implications for future practice.
To be clear, the pre-suit analyses were not silent on claim construction risk. Each of the four patent-specific infringement memoranda identified specific claim terms that were potential construction vulnerabilities. The '567 Patent memorandum flagged "adiabatically coupled" as susceptible to narrow construction based on the prosecution history. The '234 Patent memorandum noted that the claim language "continuously varying germanium concentration" might be construed to exclude step-graded architectures. The '678 Patent memorandum identified the "monolithically formed" limitation as a prosecution-history vulnerability. And the '890 Patent memorandum discussed the risk that "co-integrated" would be construed to require fabrication in a single semiconductor process, excluding Lumos's flip-chip bonded driver die.
The problem was not the identification of risks but the weighting of those risks in the overall litigation assessment. Each pre-suit memorandum characterized the relevant claim term as a "potential vulnerability" but assessed the probability of an adverse construction as moderate — typically in the range of 30–40%. The consolidated litigation strategy memorandum to the Board, drawing on these individual assessments, characterized the claim construction risk as "manageable" and expressed the view that the strength of the infringement case on a favorable construction, combined with the damages potential, justified proceeding.
In hindsight, the risk of adverse narrow constructions was substantially higher than we assessed, and our analysis suffered from several identifiable biases and gaps. First, we gave insufficient weight to the prosecution history record, which contained multiple narrowing amendments and applicant remarks that created clear prosecution history estoppel arguments. Second, we overweighted our own assessment of the technically "correct" construction based on POSITA understanding and underweighted the likelihood that the Court would rely primarily on the intrinsic record — specifically the prosecution history — consistent with the Phillips v. AWH Corp., 415 F.3d 1303 (Fed. Cir. 2005) (en banc), hierarchy of claim construction evidence. Third, we did not sufficiently consider the cumulative effect of having construction vulnerabilities across all four patents: even if the probability of an adverse construction on any single term was only 40%, the probability that at least some of the constructions would be adverse was substantially higher, and the interdependence of the infringement theories meant that adverse constructions on even a subset of the patents could undermine the economic viability of the entire case.
B. Specific Analytical Gaps
A closer examination of each patent reveals specific analytical gaps that contributed to the underassessment of claim construction risk:
'567 Patent ("adiabatically coupled"): Our pre-suit analysis correctly identified that the "adiabatically coupled" limitation was added during prosecution to overcome the Tanaka prior art — U.S. Patent No. 9,456,012, which was, ironically, co-invented by Hideo Tanaka before he joined OWS and was cited against the '567 Patent under 35 U.S.C. § 102(a)(1). We flagged prosecution history estoppel risk and noted that the amendment narrowed the scope of the claims. However, we did not sufficiently consider how the Court would construe the scope of the estoppel and, in particular, how the applicant's narrowing amendments would be read to define the metes and bounds of the claimed invention. Our analysis took the position that "adiabatically coupled" was a term of art in the photonics field that encompassed both continuous and segmented taper designs, and that a POSITA would understand the term to include any coupling structure that achieves adiabatic mode evolution. This position, while arguably correct as a matter of technical understanding, failed to account for the fact that the applicant's prosecution remarks had specifically contrasted the claimed continuous taper with the segmented approach of the Tanaka prior art. The Court, consistent with the Ridgeway & Co. v. Conceptronic, Inc., 90 F.3d 1576 (Fed. Cir. 1996), approach, gave controlling weight to the intrinsic record and concluded that the applicant had disclaimed segmented taper designs.
'234 Patent ("continuously varying germanium concentration"): The '234 Patent specification contained embodiments describing both "continuously graded" and "step-graded" buffer layers, and our analysis relied on the specification's disclosure of both embodiments to support a broader construction that would encompass Lumos's step-graded SiGe buffer architecture. This was a reasonable position, and the Federal Circuit has at times looked to the specification's full scope of disclosed embodiments to inform claim construction. However, we should have assigned a higher probability to the plain-meaning narrow construction. The claim language — "continuously varying" — is on its face a description of a continuous function, not a step function. The Court correctly observed that the claim language was not ambiguous and did not require resort to the specification to expand its scope beyond its plain meaning. We should have recognized that this was a case where the claim drafter's choice of language would control, and that arguing for a broader construction would require us to overcome the heavy burden of showing that the plain meaning was inconsistent with the specification — a burden we could not carry because the specification did in fact describe continuous-grading embodiments that aligned with the plain meaning.
'678 Patent ("monolithically formed"): The "monolithically formed on a single silicon die" limitation was added by amendment during prosecution to overcome prior art based on hybrid integration references. Most critically, the applicant's prosecution remarks explicitly stated: "the present invention is distinguished from heterogeneous integration approaches." This statement created a textbook prosecution history estoppel under Festo Corp. v. Fenmore & Co., 535 U.S. 722 (2002), and its progeny. Our pre-suit analysis acknowledged this remark but assessed the risk as moderate, reasoning that the applicant's remark could be read as distinguishing only the specific prior art references at issue rather than broadly disclaiming all heterogeneous integration. In retrospect, this was an overly optimistic reading of the prosecution history. The applicant's remark was categorical — "distinguished from heterogeneous integration approaches" — and was reasonably read as a broad disclaimer. We should have weighted this risk as high, particularly given that Lumos's micro-transfer-printed III-V gain elements are unambiguously a form of heterogeneous integration.
'890 Patent ("co-integrated"): The "co-integrated" limitation was added during prosecution to distinguish prior art designs in which driver circuits are separately fabricated and bonded to the photonic die. Lumos's architecture — a separate CMOS driver die flip-chip bonded to the photonic die via copper pillar micro-bumps — was precisely the type of design the applicant had distinguished during prosecution. Our analysis recognized this risk but assessed it as moderate, in part because we believed that "co-integrated" could be construed more broadly to encompass any design in which the driver and photonic components are integrated into a single functional package, regardless of the fabrication method. This position was difficult to sustain in the face of the prosecution history, and we should have recognized it as such.
C. Recommendation: Enhanced Pre-Suit Markman Simulation
Based on the experience of this matter, the firm should implement a formal pre-suit "Markman simulation" exercise for any plaintiff-side patent case in which claim construction will be outcome-determinative — which, as a practical matter, includes nearly all patent infringement cases. The recommended process is as follows:
First, a team member who was not involved in the prosecution analysis or the pre-suit infringement assessment should be assigned to play the role of "devil's advocate." This team member should prepare the strongest possible defendant's claim construction brief, drawing on the prosecution history, prior art, specification, and any other intrinsic or extrinsic evidence that could support narrow constructions. The devil's advocate should be given explicit instructions to argue for the narrowest reasonable construction of each key claim term, without regard to the firm's desired outcome.
Second, the firm should conduct a red-team/blue-team exercise in which the devil's advocate presents the defendant's claim construction arguments to the litigation team, and the team responds in the posture of a Markman hearing. This exercise should be structured to simulate the adversarial dynamic of the actual hearing and to force the team to confront the strongest versions of the defendant's arguments.
Third, the firm should develop a standardized claim construction risk scoring rubric with weighted factors for use in pre-suit assessments. The rubric should include, at minimum, the following factors: (a) prosecution history estoppel risk (weighted heavily), including the number and specificity of narrowing amendments and applicant remarks; (b) specification embodiment scope, including whether the claims are broader or narrower than the disclosed embodiments; (c) prior art landscape, including the proximity of the prior art to the claimed invention and the likelihood that the defendant will argue for constructions that align with the prior art boundary; (d) claim amendment history, including the number of office action rejections and the extent of claim narrowing during prosecution; and (e) the specific judge's track record on claim construction, including any discernible tendencies toward narrow or broad constructions.
With respect to the last factor, our pre-suit assessment did not include a systematic study of Judge Torvelis's Markman order history. A review of his publicly available claim construction orders would have provided useful data on his approach to prosecution history estoppel and his general inclination toward intrinsic versus extrinsic evidence. In future matters, this analysis should be conducted as a standard part of the pre-suit assessment, regardless of the forum or the assigned judge.
D. Integration of Technical Expert at the Pre-Suit Stage
Dr. Harold Whitfield was not retained until August 2022, approximately two months after the complaint was filed. His engagement was timed to support the claim construction briefing and Markman hearing, which was the standard approach for technical expert engagement in our practice at that time.
However, earlier involvement of a technical expert — ideally during the pre-suit investigation phase — might have provided a more critical and technically grounded assessment of the claim construction landscape. A technical expert with deep knowledge of the silicon photonics field would have been well-positioned to assess whether a POSITA would understand the disputed claim terms in the broader sense that OWS advocated, or whether the technical community's understanding aligned more closely with the narrow constructions that the defendant ultimately proposed. More importantly, a technical expert engaged at the pre-suit stage could have provided a candid assessment of whether the intrinsic record — particularly the prosecution history — was consistent with or undermined the broader constructions.
We recommend that, in future cases turning on technical claim terms, the firm engage the technical expert during the pre-suit investigation phase to stress-test claim construction positions before committing the client to litigation. This engagement should be structured as a critical evaluation, with the expert asked to identify weaknesses in the proposed constructions rather than to advocate for them.
IV. Lessons Learned: Patent Prosecution Counseling
A. Prosecution Decisions That Narrowed Claim Scope
A critical underlying cause of the adverse Markman outcome was the narrowing of claim scope during patent prosecution. Three of the four patents-in-suit — the '567, '678, and '890 Patents — were narrowed during prosecution through claim amendments adding limitations to overcome prior art rejections. These amendments created the prosecution history estoppel and narrow-construction arguments that BHS exploited and that the Court adopted:
• '567 Patent: The "adiabatically coupled" limitation was added during prosecution to overcome the Tanaka prior art (U.S. Patent No. 9,456,012). The '567 Patent underwent two office action rejections before the claims were amended to include this limitation and the patent was allowed. The amendment and the applicant's accompanying remarks specifically contrasted the claimed continuous taper design with the segmented approach of the Tanaka prior art.
• '678 Patent: The "monolithically formed on a single silicon die" limitation was added during prosecution to overcome prior art references based on hybrid integration approaches. The '678 Patent underwent three office action rejections, making it the most heavily prosecuted of the four asserted patents. The applicant's prosecution remarks included the categorical statement that "the present invention is distinguished from heterogeneous integration approaches," creating a broad disclaimer that the Court enforced.
• '890 Patent: The "co-integrated" limitation was added during prosecution to distinguish prior art designs in which driver circuits are separately fabricated and bonded to the photonic die. The '890 Patent underwent two office action rejections before the limitation was added.
These prosecution amendments directly limited the scope of the doctrine of equivalents under Festo and created narrow literal-construction arguments that the Court adopted. The amendments were made by OWS's prosecution counsel — not C&H, as C&H was not engaged in prosecution of these patents — but the enforcement consequences of the amendments fell squarely on the litigation team.
The '234 Patent presented a different but related issue. The '234 Patent was allowed on the first office action after an examiner interview, and there were no narrowing amendments during prosecution. However, the claim language — "continuously varying germanium concentration" — was narrower than the specification's full disclosure, which described both continuously graded and step-graded buffer layer embodiments. The claim drafter chose language that corresponded to only one of the disclosed embodiments, leaving the broader embodiment unclaimed. While this is not a prosecution history estoppel issue in the traditional sense, it reflects a prosecution drafting choice that narrowed the claim scope relative to the disclosed invention and limited the patent's enforcement value.
B. Implications for Client Counseling
The experience of this matter highlights the critical importance of integrating litigation enforcement considerations into the patent prosecution process. When counseling clients on patent prosecution strategy, the litigation team should be consulted to assess the enforceability implications of proposed claim amendments, particularly for high-value patents that the client may later seek to enforce against competitors.
We offer the following specific recommendations:
First, C&H should implement a cross-functional review process in which the IP litigation group reviews prosecution amendments for key client patents before they are entered. This review should be targeted — it is neither practical nor necessary to review every amendment on every application — but should be applied to patents that the client has identified as strategically important or that cover technologies likely to be the subject of future enforcement actions. The review should assess whether proposed amendments create prosecution history estoppel risks that could limit the patent's enforcement value and, where possible, should suggest alternative amendment strategies that achieve allowance without unnecessary narrowing.
Second, for clients like OWS with significant patent portfolios (87 issued patents, 14 pending applications at the time of filing), C&H should develop a "litigation-readiness" prosecution strategy for the highest-value patents. This strategy should ensure that claims are drafted and prosecuted with an eye toward maintaining breadth sufficient to cover foreseeable design-arounds by competitors. The strategy should include, at minimum, the following elements: (a) claims drafted to cover both current embodiments and anticipated alternative designs; (b) specifications that describe a broad range of embodiments and explicitly define key claim terms in a manner that supports broader constructions; and (c) prosecution conducted with an awareness that every amendment and every remark in the prosecution history will be scrutinized by opposing counsel in future litigation.
Third, when applicants must narrow claims to overcome prior art, the prosecution team should consider filing continuation applications with broader claims that may capture competitor designs not covered by the narrowed parent claims. In this matter, none of the four asserted patents had pending continuation applications at the time of filing. If continuation applications with broader claims had been on file, OWS might have had additional enforcement options even after the adverse Markman Order on the parent claims. While continuation applications cannot avoid the prosecution history of the parent, they can present new claim language that is not burdened by the same narrowing amendments.
Fourth, prosecution counsel should be trained on the litigation consequences of prosecution remarks that distinguish prior art. The applicant's remarks on the '678 Patent — "the present invention is distinguished from heterogeneous integration approaches" — were devastating at the Markman hearing. These remarks went beyond what was necessary to distinguish the specific prior art references at issue and created a broad disclaimer that the Court enforced against the patentee. Prosecution counsel should be trained to make narrowly tailored arguments that distinguish specific prior art references without making broad categorical statements that can be used to narrow the claims in future litigation.
C. Tanaka Prior Art Problem
A unique and especially frustrating vulnerability in this matter arose from the prior patent activity of Hideo Tanaka, a named inventor on the '567 Patent. Tanaka was employed as a Senior Staff Engineer at OWS from 2016 to 2019 and was a co-inventor of the '567 Patent during that period. However, before joining OWS, Tanaka was the co-inventor of U.S. Patent No. 9,456,012, which was issued on September 6, 2016. This prior patent was cited by the USPTO as prior art against the '567 Patent under 35 U.S.C. § 102(a)(1), and the '567 Patent claims were narrowed specifically to overcome Tanaka's own prior work.
The irony of this situation is not lost on the litigation team. The very inventor whose contributions made the '567 Patent possible also created the prior art that forced the claim narrowing that ultimately destroyed the patent's enforcement value. The prosecution history estoppel that BHS exploited at Markman was the direct result of amendments made to overcome Tanaka's pre-OWS prior art.
The lesson from this experience is straightforward but important: pre-filing due diligence for any patent enforcement action should include a thorough review of all named inventors' prior patent and publication histories, including work performed at prior employers or academic institutions, to anticipate prior art challenges and the prosecution narrowing that those challenges may have caused. In this case, the connection between Tanaka's prior patent and the '567 Patent's prosecution history was known to the litigation team but was not given sufficient weight in the overall risk assessment. A more rigorous pre-filing review — one that specifically examined the prosecution history amendments prompted by each prior art reference, including those authored by the patent's own inventors — would have revealed the full scope of the estoppel risk.
V. Lessons Learned: Litigation Investment Decision Framework
A. The Decision to File Suit
The OWS Board of Directors approved a litigation budget of $4.2 million through trial based on projected damages exceeding $75 million. The damages projection was developed by Dr. Lawrence Kessler of Apex Claim Analytics LLC and was based on a 7.5% reasonable royalty rate applied to $148.6 million in accused product revenue (LumosBridge™ and LumosEdge™ combined), with additional amounts for pre-suit damages accruing from the date of the cease-and-desist letter and a potential willfulness enhancement of up to treble damages. On its face, the projected return on litigation investment — a potential recovery exceeding $75 million on a $4.2 million investment — appeared highly attractive, representing a potential return of approximately 18:1.
However, this return projection was contingent on several assumptions, the most critical of which was that OWS would survive the Markman hearing with viable infringement theories intact. Markman was a binary risk event: a favorable outcome would leave the case on track for substantial damages, while an adverse outcome would effectively end the case. Our pre-suit assessment acknowledged this risk but did not model it with sufficient rigor. The projected damages and the litigation budget were presented to the Board as the expected scenario, without adequate emphasis on the probability-weighted expected value of the case after accounting for the Markman risk.
The actual spend of $2,549,568, representing 61% of the Board-approved budget, was consumed before the case reached any merits determination beyond claim construction. The entire amount was lost with zero recovery. This outcome — a total loss of $2,549,568 — was within the range of foreseeable outcomes had the Markman risk been properly modeled, but it was not presented to the Board as a prominent scenario.
B. Decision-Tree / Probability-Weighted Modeling
The most significant gap in our litigation investment analysis was the absence of a rigorous decision-tree model with probability-weighted outcomes at key decision nodes. We presented the Board with a budget (the cost side) and projected damages (the upside scenario) but did not present a formal expected-value analysis that incorporated the probability of adverse outcomes at Markman, summary judgment, and trial.
In retrospect, had we constructed a decision-tree model and assigned a realistic probability to the Markman risk, the analysis would have looked substantially different. Our own pre-suit analyses identified claim construction vulnerabilities across all four patents and assessed the risk of adverse constructions as moderate — which, in the framework of our internal discussions, corresponded to a probability range of approximately 30–40% for each patent. Even taking the lower end of that range, the cumulative probability of surviving Markman on all four patents with viable infringement theories was substantially less than 100%.
For illustrative purposes, if we assign a 60% probability of surviving Markman with viable infringement theories (the complement of a 40% adverse-construction risk), a 70% probability of surviving summary judgment conditional on surviving Markman, and a 55% probability of prevailing at trial conditional on surviving summary judgment, the probability of reaching a favorable verdict would be approximately 23% (0.60 × 0.70 × 0.55 = 0.231). Applied to a $75 million damages projection, the probability-weighted expected recovery would be approximately $17.3 million, which, after deducting the full $4.2 million litigation budget, would yield a probability-weighted net expected value of approximately $13.1 million. This would still have justified proceeding, but it would have presented a materially different picture than the upside-only scenario that was emphasized in the Board presentation.
More importantly, if we had assigned a more realistic 50–60% probability of losing Markman — which, in hindsight, the prosecution history vulnerabilities warranted — the expected value would have been substantially lower. At a 50% Markman survival probability, the probability-weighted expected recovery drops to approximately $14.4 million; at a 40% survival probability, it drops to approximately $11.6 million. And these figures assume that a favorable Markman outcome would not affect the damages figure, which is itself a simplifying assumption. Factoring in the possibility that even a partially favorable Markman outcome might require damages adjustments, the expected value could fall further.
The critical point is not whether any particular probability estimate would have changed the ultimate decision to file suit — it might not have — but that the Board was entitled to a transparent, probability-weighted analysis when making a $4.2 million investment decision. The recommendation for future practice is clear: for any matter with significant claim construction risk, the firm should prepare a formal decision-tree model with probability nodes for each major phase of the litigation and present the probability-weighted expected value alongside the upside scenario when advising clients and investors on litigation investment decisions.
C. Third-Party Litigation Funding Consideration
During the pre-litigation phase, OWS and Vantage Point Capital Partners evaluated third-party litigation funding as a potential mechanism for reducing downside risk. Resolute Litigation Finance, LP conducted due diligence on the case and provided a term sheet offering to fund a portion of the litigation costs in exchange for a contingent return on any recovery. The specific terms of the Resolute proposal included funding of up to $3.0 million in litigation costs in exchange for a priority return of 2.5x the funded amount from any recovery, with OWS and Vantage Point retaining the remainder.
OWS and Vantage Point declined the Resolute term sheet and elected to self-fund the litigation. The reasons for declining were not fully documented in the matter file, but discussions with the client indicated that the primary factors were Vantage Point's confidence in the case, the perceived cost of the funder's contingent return (which would have consumed a significant portion of any recovery below $20 million), and a general preference for maintaining full control of the litigation without third-party involvement.
In hindsight, accepting third-party funding would have shifted a significant portion of the downside risk away from OWS. Under the Resolute term sheet, if the case had resulted in a zero recovery (as it did), Resolute would have borne the funded costs and OWS's exposure would have been limited to costs not covered by the funding arrangement. While the funder's contingent return would have reduced OWS's upside in the event of a recovery, the risk-transfer benefit of funding would have been valuable given the claim construction vulnerabilities that were identified — but underweighted — in the pre-suit analysis.
More broadly, the experience of this matter suggests that third-party litigation funding can serve as a useful market check on case quality when claim construction risk is identified as significant. Litigation finance firms like Resolute employ sophisticated analysts who conduct independent assessments of case merits, including claim construction risk, as part of their due diligence. The willingness (or unwillingness) of a funder to invest in a case provides an external data point on the case's expected value that can supplement the firm's own assessment. In this case, Resolute did provide a term sheet, which suggests that its analysts assessed the case favorably — but the terms of the funding (particularly the 2.5x priority return) may have reflected a discount for the claim construction risk that we underweighted in our own analysis.
The lesson for future practice is that the decision to accept or decline third-party litigation funding should be a deliberate, risk-adjusted decision rather than a default to self-funding. When claim construction risk or other significant pre-trial risks are identified, the firm should advise the client to carefully consider the risk-transfer benefits of third-party funding and to document the rationale for the ultimate decision, whether to accept or decline.
D. Relationship Between Pre-Litigation Valuation and Litigation Risk
Clearpath IP Valuation Services, Inc. provided a pre-litigation valuation of $42.8 million for the four asserted patents and $215 million for OWS's entire portfolio. The Clearpath valuation was prepared using standard methodologies — including comparable licensing transactions, income-based approaches, and market-based approaches — and was relied upon by the OWS Board and Vantage Point in evaluating the potential return on litigation investment.
However, the Clearpath valuation did not account for the probability-weighted litigation risk of narrow claim constructions. The valuation assessed the patents' intrinsic value based on their technical scope and market coverage, but did not discount that value for the probability that a court might construe the claims narrowly in a way that would exclude the accused products. This is not a criticism of Clearpath's methodology — patent valuation firms typically do not conduct detailed claim construction risk analyses — but it highlights a gap in the overall litigation investment analysis.
When patent valuations are used to support litigation investment decisions, the litigation team should provide the valuation firm with an explicit claim construction risk assessment so that the valuation can be discounted accordingly. Alternatively, the litigation team should present the Board with both the undiscounted valuation and a risk-adjusted valuation that reflects the probability-weighted litigation risk. In this case, a risk-adjusted valuation that incorporated a 40–50% probability of adverse claim constructions would have yielded a significantly lower value for the asserted patents, which might have influenced the Board's assessment of the litigation's expected return.
VI. Lessons Learned: Expert Engagement and Utilization
A. Technical Expert — Dr. Harold Whitfield
Dr. Harold Whitfield of the Whitfield Technical Consulting Group, Inc. (Palo Alto, California) served as OWS's primary technical expert throughout the claim construction phase. Dr. Whitfield holds a PhD in Electrical Engineering with a concentration in photonics from the University of California, Berkeley, and is Professor Emeritus of Electrical and Computer Engineering at the University of California, Santa Barbara. He has authored over 200 peer-reviewed publications in the field of integrated photonics and is widely recognized as a leading authority on silicon photonic waveguide design, germanium-on-silicon photodetector architectures, and monolithic photonic integration.
Dr. Whitfield was retained in August 2022 and provided claim construction support throughout the claim construction phase. His work included: (a) review of the asserted patent specifications and prosecution histories; (b) analysis of the accused Lumos products; (c) preparation of a declaration in support of OWS's Opening Claim Construction Brief; and (d) presentation of a technical tutorial to the Court at the January 30, 2023 Markman hearing. Dr. Whitfield's total fees, combined with those of Apex Claim Analytics, were $412,800.
Dr. Whitfield's testimony and declaration were strong on the technical merits. His POSITA opinions supported OWS's broader constructions and were grounded in a thorough understanding of the silicon photonics field. His technical tutorial at the Markman hearing was clear, well-organized, and effectively communicated the technical concepts to the Court. On the merits of the technical arguments, Dr. Whitfield performed well.
However, the Court's Markman Order gave more weight to the intrinsic evidence — specifically the prosecution history — than to extrinsic expert testimony regarding POSITA understanding. This is consistent with the well-established hierarchy of claim construction evidence set forth in Phillips v. AWH Corp., 415 F.3d 1303 (Fed. Cir. 2005) (en banc), which holds that intrinsic evidence — the claims, specification, and prosecution history — is the primary source for claim construction and that extrinsic evidence, including expert testimony, is secondary and cannot be used to contradict the intrinsic record.
In this case, Dr. Whitfield's opinions, while technically sound, were asking the Court to adopt constructions that were arguably inconsistent with the prosecution history record. The applicant's narrowing amendments and distinguishing remarks during prosecution had created an intrinsic record that pointed toward the narrow constructions that Defendant proposed. Dr. Whitfield's POSITA testimony could not overcome this intrinsic record, and the Court correctly applied the Phillips hierarchy in reaching its conclusions.
Lumos's technical consultant, Dr. Raymond Ng, was highly effective in his Markman tutorial. Dr. Ng, who holds a PhD in Optical Engineering and is of counsel to BHS, tailored his presentation to highlight the specific design differences between OWS's patented embodiments and Lumos's accused products: segmented versus continuous taper, step-graded versus continuously graded buffer layers, heterogeneous versus monolithic integration, and flip-chip bonded versus co-integrated driver circuits. His tutorial effectively reinforced BHS's prosecution history arguments by demonstrating that the accused products embodied precisely the designs that the applicant had distinguished during prosecution. The Court credited Dr. Ng's presentation, and his influence on the Markman Order should not be underestimated.
The primary lesson regarding Dr. Whitfield's engagement is one of timing rather than quality. As discussed in Section III.D above, earlier engagement of Dr. Whitfield — during the pre-suit investigation phase rather than after the complaint was filed — could have provided a more critical reality check on whether the intrinsic record would support OWS's proposed constructions. A technical expert of Dr. Whitfield's caliber, engaged in a pre-suit advisory capacity and asked to evaluate the prosecution history from the defendant's perspective, might have identified the claim construction vulnerabilities more clearly and informed a more accurate risk assessment.
B. Damages Expert — Dr. Lawrence Kessler
Dr. Lawrence Kessler of Apex Claim Analytics LLC (Chicago, Illinois) served as OWS's damages expert. Dr. Kessler holds a PhD in Economics from Northwestern University and specializes in patent damages analysis, with particular expertise in reasonable royalty analysis under the Georgia-Pacific framework.
Dr. Kessler was retained in October 2022 and began developing a preliminary damages framework based on a reasonable royalty theory. His work included: (a) analysis of accused product revenue data obtained through discovery; (b) identification and analysis of comparable license agreements in the silicon photonics and optical transceiver industries; (c) preliminary Georgia-Pacific factor analysis; and (d) development of a damages model applying a 7.5% reasonable royalty rate to $148.6 million in accused product revenue.
The Georgia-Pacific factor analysis was in progress when the Markman Order was issued on February 17, 2023, rendering the damages work moot. Dr. Kessler's engagement was suspended, and no damages report was filed or served.
The combined expert fees for Dr. Whitfield and Dr. Kessler totaled $412,800. While we cannot disaggregate this figure precisely due to the combined billing arrangement, Dr. Kessler's share of this total was substantial, reflecting significant work on the damages analysis during the period from October 2022 through February 2023.
The lesson regarding Dr. Kessler's engagement relates to the staging of expert work relative to litigation milestones. In this matter, Dr. Kessler was engaged and began substantive damages analysis before the Markman hearing, which meant that a significant portion of his work was rendered moot by the adverse Markman Order. For cases with front-loaded claim construction risk — where the viability of the infringement theories depends on sur