R2 Solutions LLC v. Databricks, Inc., No. 4:23-cv-01147 (E.D. Tex. Feb. 2, 2026)

Pending before the Court are four related motions: (1) Plaintiff R2 Solutions LLC’s Motion to Strike Databricks’ Fourth Supplemental Invalidity Contentions (Dkt. #205); (2) Plaintiff R2 Solutions LLC’s Motion to Strike Certain Expert Opinions (Dkt. #207); (3) Defendant Databricks, Inc.’s Motion to Strike R2 Solutions LLC’s Supplemental Infringement Contentions and Supplemental Opening Expert Report of Mr. William Davis (Dkt. #223); and (4) Defendant Databricks, Inc.’s Motion for Supplemental Claim Construction (Dkt. #224) (collectively, the “Motions”). Having considered the Motions, the relevant pleadings, and the applicable law, the Court finds as follows:

1. Plaintiff R2 Solutions LLC’s Motion to Strike Databricks’ Fourth

Supplemental Invalidity Contentions (Dkt. #205) should be GRANTED; 2. Plaintiff R2 Solutions LLC’s Motion to Strike Certain Expert Opinions

(Dkt. #207) should be GRANTED;

3. Defendant Databricks Inc.’s Motion to Strike R2 Solutions LLC’s

Supplemental Infringement Contentions and Supplemental Opening Expert Report of Mr. William Davis (Dkt. #223) should be DENIED; and 4. Defendant Databricks Inc.’s Motion for Supplemental Claim Construction

(Dkt. #224) should be DENIED.

2BACKGROUND

I. Factual Background

This is a patent infringement case. On December 23, 2023, Plaintiff R2 Solutions LLC (“Plaintiff” or “R2”) filed this action alleging that Defendant Databricks, Inc. (“Defendant” or “Databricks”), infringed U.S. Patent No. 8,190,610 (“the ʻ610 Patent”) (Dkt. #1 at p. 1). The ʻ610 Patent, entitled “MapReduce for Distributed Database Processing,” relates to novel and non-obvious inventions in the fields of data analytics and database structures (Dkt. #1 at ¶ 22).¹ Plaintiff’s Complaint alleges that Defendant knew of and directly or indirectly infringed, and continues to infringe, the ʻ610 Patent by, among other things, making, offering to sell, and/or selling the Accused Instrumentalities² in a manner that infringes the claims of the ʻ610 Patent (Dkt. #1 at ¶¶ 31, 43).

II. Procedural History

For context, the Court will provide a brief summary of the procedural history relevant to the four pending motions. On December 10, 2024, the Court entered its Claim Construction Memorandum Opinion and Order which, among other things, largely adopted Defendant’s3 proposed construction for the “reducing” limitation (Dkt. #71 at pp. 18-21).³ On January 9, 2025, Plaintiff served its Second Supplemental Infringement Contentions accusing, among other things, “hash join” functionality of meeting the “reducing” limitation (See Dkt. #205- 7 at pp. 66–67, 71, 77, 129). On January 29, 2025, Defendant served its Third Supplemental Invalidity Contentions alleging, inter alia, that prior art “hash joins” disclosed the “reducing” limitation (See Dkt. #205-9 at pp. 67–69; 71–72). Before the Court authorized the parties to serve supplemental contentions to address Defendant’s late document production, see infra pp. 4–5, Plaintiff’s Second Supplemental Infringement Contentions and Defendant’s Third Supplemental Invalidity Contentions were set to become the parties’ final contentions in the case.⁴

On January 24, 2025—less than three weeks before the fact discovery deadline and deadline to file motions to compel under the then operative scheduling order (see Dkt. #75 at p. 1)—during a deposition with one of Defendant’s co-founders Mr. Reynold Xin, Plaintiff learned, for the first time, that Defendant implements a proprietary source code base (Photon and Aether code) that was “designed so it can be [a] completely different implementation of the same set of functionalities that Spark would offer” (See Dkt. #82 at p. 5; Dkt. #82-2 at p. 72). At this time, Plaintiff had only accused infringement of the ʻ610 Patent via Defendant’s Data Intelligence Platform/Defendant’s4 Lakehouse Platforms that utilize “Apache Spark” or other similar functionality (See Dkt. #1 at ¶ 7; Dkt. #82 at p. 5). Following the discovery of this new and pertinent information, Plaintiff sent a letter to Defendant requesting that it produce all relevant documents relating to Photon and Aether on January 27, 2025 (Dkt. #82-13 at pp. 2–4). On January 29, 2025, during a meet and confer between the parties, Defendant refused to produce the documents relating to Photon and Aether. Defendant’s continued refusal to produce the requested documents led Plaintiff to seek this Court’s intervention to resolve the discovery dispute.

On February 5, 2025, Plaintiff filed its Motion to Compel seeking an order from the Court compelling Defendant to produce Photon and Aether source code as well as accompanying technical documentation (Dkt. #82 at p. 4).⁵ If the Court granted its Motion to Compel, Plaintiff’s motion further requested that the Court amend the current scheduling order to allow it time to address the new evidence in its infringement case (See Dkt. #82 at pp. 13–14 (explaining that under the current scheduling order fact discovery closes on February 13 and opening expert reports are due February 20)). On June 6, 2025, the Court granted Plaintiff’s Motion to Compel and ordered Defendant to produce source code and technical documentation for Photon and Aether, finding that Plaintiff’s requested discovery was relevant and proportional (See Dkt. #200 at pp. 4–10). The Court further ordered that all deadlines in the current scheduling order be moved back approximately sixty days and requested that the parties propose an amended scheduling order to this effect (See Dkt. #200 at pp. 10–13 (finding that good cause exists to amend the current scheduling order in this case)). On June 18, 2025, the parties filed their Joint Motion to Amend5 Scheduling Order (Dkt. #203). On June 23, 2025, the Court granted the Joint Motion to Amend Scheduling Order extending limited deadlines “to permit Plaintiff time to address Defendant’s production” and authorizing the parties to serve narrow supplemental contentions (See Dkt. #204).⁶ This order is now at the center of the dispute between the parties.

Pursuant to the Court’s June 23 Order, both parties served supplemental infringement and invalidity contentions. On August 1, 2025, Plaintiff served its Third Supplemental Infringement Contentions, identifying one additional infringing operation—Photon’s Shuffled Hash Join (“PSHJ”)—and explaining that Aether is involved with distribution of this operation across nodes (Dkt. #205-3 at pp. 12, 26). On August 22, 2025, Defendant served its Fourth Supplemental Invalidity Contentions, containing over 1,000 pages of supplemental charting and more than 7,000 pages of prior art production (Dkt. #205-2). Upon receipt, Plaintiff immediately objected to Defendant’s Fourth Supplemental Invalidity Contentions, arguing that its supplemental contentions violate the Court’s recent order because they are not “tied to R2’s supplements related to Photon and Aether” (Dkt. #205-4 at p. 4). On August 25, 2025, Defendant responded, claiming that its supplemental contentions “fully comply with the Court’s order” because they “identify prior art that discloses the same shuffle hash join functionality in a distributed system that R2 now alleges meets the requirements of the asserted claims” (Dkt. #205 at p. 3). On August 26, 2025, the parties met and conferred but were unable to informally resolve their dispute regarding the6 supplemental contentions, prompting the parties to file the instant Motions currently pending before the Court.

On September 16, 2025, the parties filed their third Joint Motion to Amend Scheduling Order seeking to extend the remaining case deadlines until after the Court resolves Plaintiff’s Motion to Strike Defendant’s Fourth Supplemental Invalidity Contentions and Plaintiff’s Motion to Strike Certain Expert Opinions in Defendant’s Supplemental Opening Report of Dr. Jon Weissman (See Dkt. #209). On September 24, 2025, the Court granted the Joint Motion to Amend Scheduling Order extending all remaining case deadlines to allow it to resolve the parties’ dispute regarding whether Defendant’s supplements violate the Court’s June 23 Order (See Dkt. #213).

III. The Parties Pending Motions

On August 29, 2025, Plaintiff filed its Motion to Strike Defendant’s Fourth Supplemental Invalidity Contentions, arguing that Defendant’s most recent contentions violate both the Court’s June 23 Amended Scheduling Order and June 6 Discovery Order because they are not “tied to R2’s supplements related to Photon and Aether” (Dkt. #205 at p. 5). On September 12, 2025, Defendant filed its Response, arguing that its supplemental contentions are “tied to R2’s supplements related to Photon and Aether” by identifying prior art and invalidity grounds that discloses the “Shuffle Hash Join” functionality that Plaintiff now alleges, for the first time, infringes the ʻ610 Patent (Dkt. #208 at pp. 7–8). On September 18, 2025, Plaintiff filed its Reply in Support of Motion to Strike Defendant’s Fourth Supplemental Invalidity Contentions (Dkt. #210).

On September 12, 2025, Plaintiff filed its Motion to Strike Certain Expert Opinions arguing that the Court should strike most of Defendant’s Supplemental Opening Expert Report of Dr. Weissman for, among other reasons, failing to comply with the Court’s June 23 Order (Dkt. #2077 at p. 5).⁷ On September 26, 2025, Defendant filed its Response, arguing that Dr. Weissman’s supplemental report fully complies with the Court’s order because his opinions are directly tied to Plaintiff’s supplemental contentions based on “Shuffle Hash Join” functionality (Dkt. #216 at pp. 6–7). On October 3, 2025, Plaintiff filed its Reply in Support of Motion to Strike Certain Expert Opinions (Dkt. #218). On October 10, 2025, Defendant filed its Sur-Reply to Plaintiff’s Motion to Strike Certain Expert Opinions (Dkt. #222).

On October 10, 2025, Defendant filed its Motion to Strike Plaintiff’s Supplemental Infringement Contentions and Supplemental Opening Expert Report of Mr. William Davis, arguing that the Court should strike the entirety of Plaintiff’s supplemental contentions and supplemental expert report of Mr. Davis because Plaintiff waived its ability to assert its new infringement theory based on a join strategy called “Shuffle Hash Join” (Dkt. #223 at p. 5). On October 20, 2025, Plaintiff filed its Response, arguing that there is no merit to Defendant’s motion because Defendant knows well that Plaintiff’s supplemental contentions and expert reporting are timely and expressly authorized by the Court’s orders (Dkt. #227 at p. 4). On October 27, 2025, Defendant filed its Reply in Support of its motion to strike (Dkt. #231). On November 3, 2025, Plaintiff filed its Sur-Reply in Opposition to Defendant’s motion (Dkt. #234).

On October 15, 2025, Defendant filed its Motion for Supplemental Claim Construction requesting that the Court supplement its construction of the claim term “reducing” based on Plaintiff’s supplemental infringement theory that now accuses that “hash join” functionality—a functionality it clearly and unmistakably disclaimed in IPR—meets the “reducing” requirements of the asserted claims (Dkt. #224 at p. 5). On October 30, 2025, Plaintiff filed its Response, arguing8 that its identification of “hash join” functionality for the “reducing” limitation does not raise new claim construction issues considering that it has mapped “hash join” functionality to the “reducing” limitation since well before claim construction began (Dkt. #233 at p. 5). On November 6, 2025, Defendant filed its Reply in Support of its motion (Dkt. #236). On November 13, 2025, Plaintiff filed its Sur-Reply in opposition to Defendant’s motion (Dkt. #239).

The Motions are now ripe for adjudication.

LEGAL STANDARD

I. The Local Patent Rules

This dispute is governed in part by the Local Patent Rules. The Court, once again, faces the conflict between the necessity for orderly and timely disclosure of necessary information on the one hand, and the natural desire for litigants to attain important tactical advantages by delaying disclosure of key elements of their case. See Anascape, Ltd. v. Microsoft Corp., No. CIV.A. 9:06-CV-158, 2008 WL 7180756, at *2 (E.D. Tex. May 1, 2008) [hereinafter Anascape]. Over time, the Federal Rules of Civil Procedure have been amended in favor of full and open discovery. See FED. R. CIV. P. 26 Advisory Committee Note (1993 Amendment). The Rules are to be “construed and administered to secure the just, speedy, and inexpensive determination of every action.” FED. R. CIV. P. 1. Some district courts, including this one, “have recognized that the normal discovery process can be very cumbersome and expensive in a complicated patent case.” Anascape, 2008 WL 7180756, at *2. To balance the competing interest of the parties, the Eastern District of Texas, like many other courts, have adopted Local Patent Rules and frequently enter special scheduling orders. Id. This tracks suggestions made in the MANUAL FOR COMPLEX LITIGATION, FOURTH, § 11.212.

Local Patent Rules are a valid exercise of the Court’s authority to manage its docket in accordance with the Federal Rules of Civil Procedure. See FED. R. CIV. P. 83; O2 Micro Int’l Ltd. v.9 Monolithic Power Sys., Inc., 467 F.3d 1355, 1365–66 (Fed. Cir. 2006) [hereinafter O2 Micro]. The Eastern District of Texas unanimously adopted the Local Patent Rules. Anascape, 2008 WL 7180756, at *2. They are posted on the Court’s website, and each party was informed in the Order Governing Proceedings that they would apply (See Dkt. #18 at p. 3). The Scheduling Order in this case set deadlines for the parties to complete actions required by the various provisions of the Local Patent Rules (See e.g., Dkt. #28; Dkt. #75).

The Local Patent Rules are strictly enforced and “exists to further the goal of full, timely discovery and provide all parties with adequate notice and information with which to litigate their cases, not to create supposed loopholes through which parties may practice litigation by ambush.” Anascape, 2008 WL 7180756, at *2 (citation omitted). A court has the inherent power to enforce its scheduling orders and to impose sanctions. See FED. R. CIV. P. 16(f ); O2 Micro, 467 F.3d at 1367; Flaska v. Little River Marina Const. Co., 389 F.2d 885, 887 n. 3 (5th Cir. 1968). However, interpretation and enforcement of discovery provisions of the Local Patent Rules should not conflict with, and should harmonize with, the discovery provisions of the Federal Rules of Civil Procedure. O2 Micro, 467 F.3d at 1365. “Deciding whether invalidity contentions should be struck, or whether to allow amendments to contentions, is similar to deciding whether evidence should be excluded for discovery violations.” Anascape, 2008 WL 7180756, at *2. It is also akin to deciding whether the pleading deadlines of a scheduling order should be extended. Id. Therefore, the Court will consider the kinds of factors identified as important in making both types of decisions. See O2 Micro, 467 F.3d at 1366; Finisar Corp. v. DirectTV Grp., Inc., 424 F. Supp. 2d 896, 899 (E.D. Tex. 2006). A non-exclusive list of factors considered by courts includes: (1) the reason for the delay, including whether it was within the reasonable control of the party responsible for the delay; (2)10 whether the offending party was diligent in seeking an extension of time, or in supplementing discovery, after an alleged need to disclose the matter became apparent; (3) the importance of the particular matter; (4) the danger of unfair prejudice to the non-movant; and (5) the length of the delay and its potential impact on the judicial proceedings. See Anascape, 2008 WL 7180756, at *3 (citation omitted).

II. Patent Rule 3-3

Plaintiff contends that Defendant’s Fourth Supplemental Invalidity Contentions violate the express requirements of Patent Rules 3-3(c) and 3-3(b) (Dkt. #205 at pp. 14–15). Generally, Local Patent Rule 3-3 (“Rule 3-3”) governs invalidity contentions. See P.R. 3-3; Allergan, Inc. v. Teva Pharms. USA, Inc., No. 2:15-CV-1455-WCB, 2017 WL 11807449, at *1 (E.D. Tex. Aug. 3, 2017). The invalidity contentions must include “[t]he identity of each item of prior art that alleged anticipates each asserted claim or renders it obvious.” P.R. 3-3(a). The contentions must include “[w]hether each item of prior art anticipates each asserted claim or makes it obvious.” P.R. 3-3(b). The contentions must also include invalidity charts “identifying where specifically in each alleged item of prior art each element of each asserted claim is found.” P.R. 3-3(c). The purpose of Rule 3-3 is to place plaintiff on notice of potentially invalidity art that defendants will assert in their case and at trial. See PerdiemCo LLC v. IndusTrack LLC, No. 2:15-CV-00727-JRG-RSP, 2016 WL 8189021, at *1 (E.D. Tex. Oct. 25, 2016). As such, if a defendant fails to comply with Rule 3-3, then the Court will strike the deficient invalidity contentions from the record. See Saffran v. Johnson & Johnson, No. 2:07-CV-0451-TJW, 2009 WL 8491495, at *1 (E.D. Tex. Feb. 24, 2009) (striking defendant’s invalidity for failing to comply with the plain reading and spirit of Local Patent Rule 3-3). In short, Rule 3-3“requires, at a minimum, that defendants (1) identify each prior art reference; (2) identify whether each reference anticipates or renders obvious each asserted claim; (3) identify11 the combination of references that makes a claim obvious and the source of the motivation to combine such terms; and (4) provide a chart identifying where specifically in each alleged item of prior art each element of each asserted claim is found.” CyEee Grp. Ltd. v. Samsung Elecs. Co., No. 2:17-CV-140-WCB, 2018 WL 4100760, at *2 (E.D. Tex. July 2, 2018).

III. Patent Rule 3-6(b)

Defendant contends that its Fourth Supplemental Invalidity Contentions comply with Patent Rule 3-6(b) (Dkt. #208 at p. 15). Rule 3-6(b) allows a party to supplement its infringement contentions “by order of the Court, which shall be entered only upon a showing of good cause.” P.R. 3-6(b). “The Court has broad discretion to determine what constitutes good cause to amend invalidity contentions.” MacroSolve v. Antenna Software, Inc., No. 6:11-CV-287-MHS-JLD, 2013 WL 3833079, at *2 (E.D. Tex. July 23, 2013) (citing S&W Enterprises, L.L.C. v. SoutTrust Bank of Alabama, NA, 315 F.3d 533, 535 (5th Cir. 2003)). Courts in this district consider four factors when determining good cause: “(1) the explanation for the failure to meet the deadline; (2) the importance of the thing that would be excluded if the proposed amendment is not allowed; (3) potential prejudice in allowing the thing that would be excluded; and (4) the availability of a continuance to cure such prejudice.” Greenthread, LLC v. OmniVision Tehcs., Inc., No. 2:23-CV-00157-JRG, 2024 WL 1744069, at *2 (E.D. Tex. Apr. 23, 2024) (citing Intellectual Ventures II LLC v. FedEx Corp., No. 2:16-cv-00980, 2017 WL 4812436, at *2 (E.D. Tex. Oct. 25, 2017)).

IV. Motion to Strike Expert Reports

The Local Patent Rules “exist to further the goal of full, timely discovery and provide all parties with adequate notice and information with which to litigate their cases.” Computer Acceleration Corp. v. Microsoft Corp., 503 F. Supp. 2d 819, 822 (E.D. Tex. 2007). A plaintiff must “set forth specific theories of infringement at the outset of the case.” Orion IP, LLC v. Staples, Inc.,12 407 F. Supp. 2d 815, 817 (E.D. Tex. 2006). These “contentions must be reasonably precise and detailed to provide a defendant with adequate notice of the plaintiff’s theories of infringement, [but] they need not meet the level of detail required, for example, on a motion for summary judgment on the issue of infringement.” Realtime Data, LLC v. Packeteer, Inc., No. CIV.A. 6:08CV144, 2009 WL 2590101, at *5 (E.D. Tex. Aug. 18, 2009) (citation modified). “Nevertheless, a party may not rely on vague, conclusory language in its infringement contentions.” Longhorn HD LLC v. NetScout Sys., Inc., No. 2:20-CV-00349-JRG-RSP, 2022 WL 903931, at *2 (E.D. Tex. Mar. 27, 2022).

Expert infringement reports may not introduce theories not previously set forth in infringement contentions. See Anascape, 2008 WL 7180756, at *4 (striking portions of expert report that exceeded scope of invalidity contentions); Visto Corp. v. Seven Networks, Inc., No. CIV.A. 2:03CV333TJW, 2006 WL 5153146, at *1 (E.D. Tex. Mar. 27, 2006) (striking portions of expert testimony relying on prior art not disclosed in invalidity contentions). However, “the scope of infringement contentions and expert reports are not coextensive.” Longhorn HD LLC, 2022 WL 903931, at *2.

Infringement contentions need not disclose “specific evidence nor do they require a plaintiff to prove its infringement case,” EON Corp. IP Holdings, LLC v. Sensus USA Inc., No. 6:09-CV-116, 2010 WL 346218, at *2 (E.D. Tex. Jan. 21, 2010), “whereas expert reports must include a complete statement of the expert’s opinions, the basis and reason for them, and any data or other information considered when forming them.” Longhorn HD LLC, 2022 WL 903931, at *2 (citing FED. R. CIV. P. 26(a)(2)(B)).

13“Whether assessed as a late amendment to a pleading or some other discovery violation, the factors relevant to permitting the amendment or excluding the evidence are substantively identical.” Biscotti Inc. v. Microsoft Corp., No. 2:13-CV-01015-JRG-RSP, 2017 WL 2267283, at *2 (E.D. Tex. May 24, 2017) (citation omitted). The Fifth Circuit applies a four-factor test to determine whether exclusion of expert testimony based on undisclosed information is an appropriate remedy: “(1) the importance of the evidence; (2) the prejudice to the opposing party of including the evidence; (3) the possibility of curing such prejudice by granting a continuance; and (4) the explanation for the party’s failure to disclose.” Longhorn HD LLC, 2022 WL 903931, at *2 (citing Primrose Operating Co. v. Nat’l Am. Ins. Co., 383 F.3d 546, 563–64 (5th Cir. 2004)). The movant carries the burden on this issue. Godo Kaisha IP Bridge 1 v. Broadcom Ltd., No. 2:16-CV-00134-JRG, 2017 WL 2844646, at *1 (E.D. Tex. May 19, 2017).

ANALYSIS

I. Plaintiff R2 Solutions LLC’s Motion to Strike Databricks’ Fourth Supplemental

Invalidity Contentions (Dkt. #205) Plaintiff argues that Defendant’s Fourth Supplemental Invalidity Contentions should be stricken from the record in their entirety because (1) Defendant’s supplemental invalidity contentions are not “tied to” Plaintiff’s supplements related to Photon and Aether; (2) Defendant has no justification for the late addition of new theories and references; (3) Defendant should not be allowed a windfall born of its own discovery; and (4) Defendant’s supplement violates Local Patent Rules 3-3(c) and 3-3(b) (See Dkt. #205 at pp. 8–16). The Court addresses each argument in turn.

14A. Defendant’s Supplemental Invalidity Contentions Are Not “Tied to”

Plaintiff’s Supplements Related to Photon and Aether.

Plaintiff moves to strike the entirety of Defendant’s supplemental invalidity contentions on the grounds that they are not “tied to R2’s supplements related to Photon and Aether,” as required by the Court’s June 23 Order (Dkt. #205 at p. 5). Plaintiff argues that Defendant’s supplemental contentions added entirely new invalidity theories—including prior art references, anticipation positions, obviousness combinations, and “state-of-the-art” documents—that have nothing to do with Photon or Aether (See Dkt. #205 at pp. 7, 9). For example, Defendant’s supplemental contentions do not analyze Photon or Aether, do not reference PSHJ, and do not demonstrate how any of the newly added references rebuts its Photon/Aether-specific allegations (See Dkt. #205 at pp. 7, 9). Instead, Defendant just broadly cites these new references as discussing hashing algorithms and parallel joins—material that Plaintiff contends bears no unique relevance to the compelled code bases and was just as relevant to its Original Infringement Contentions as it is to its Third Supplemental Infringement Contentions (Dkt. #205 at p. 9). According to Plaintiff, its mapping of Photon/Aether is not the type of scenario that justifies the sort of unrestricted overhaul that Defendant has made to its most recent invalidity defenses (Dkt. #205 at pp. 9, 13).

In response, Defendant argues that its supplemental invalidity contentions fully comply with Court’s order because it diligently served contentions that are “tied to R2’s supplements related to Photon and Aether” (Dkt. #208 at pp. 6–7). Defendant contends that each of its supplemental contentions identify prior art that discloses the same “Shuffle Hash Join” functionality that Plaintiff now alleges, for the first time, infringes the ʻ610 Patent (Dkt. #208 at p. 6). In support of this argument, Defendant asserts that the term “Shuffled Hash Join” or “Shuffle Hash Join” is an abbreviation for a specific join strategy that includes three general steps15 (Dkt. #208 at p. 7). First, a “scan” step reads portions of various input files as partitions (Dkt. #208 at p. 7); Second, a “shuffle” step (also known as an “exchange”) uses “hash partitioning” to form new partitions and redistribute the scanned data to these new partitions so that similar data is located on the same partitions (Dkt. #208 at pp. 7–8). And third, a “hash join” step executes on each partition in parallel to join the data on the partition by building a hash table from one set of inputs and then probing that hash table with the other set of inputs to identify matching values (Dkt. #208 at p. 8). Based on this new infringement theory, Defendant contends that it supplemented its invalidity contentions, which included seven new claim charts that each demonstrate how the charted reference discloses a Shuffle Hash Join that meets the asserted claims “under R2’s interpretation of the asserted claim in R2’s infringement contentions” (Dkt. #208 at p. 8).⁸ From Defendant’s perspective, there can be no dispute that its supplemental invalidity contentions are “tied to” Plaintiff’s supplements related to Photon and Aether and thus comply with the Court’s order (Dkt. #208 at p. 10).

The Court agrees with Plaintiff that Defendant’s supplemental invalidity contentions violate the Court’s June 23 Order because they are not “tied to” Plaintiff’s supplements related to Photon and Aether. Defendant’s main argument that its supplemental contentions comply with the Court’s order is based on its erroneous belief that Plaintiff’s narrow supplemental contentions introduced a new theory of infringement (See e.g., Dkt. #208 at p. 6 (alleging that “[f ]or the first16 time, in its supplemental infringement contentions, R2 alleged infringement by a ʻShuffle Hash Join’ join strategy’”) (emphasis added))). They do not. Contrary to Defendant’s belief, Plaintiff’s supplemental infringement contentions merely substantiated its infringement case using entirely new information that was not otherwise available to Plaintiff until recently. This has been done by identifying just a single additional infringing operation, PSHJ, identified and made known to Plaintiff only after Defendant was compelled to produce it pursuant to the Court’s June 6 Order. As such, the Court finds that this is not a scenario that allows for the type of carte blanche amendments that Defendant has made to its most recent supplemental invalidity contentions.⁹ Indeed, Plaintiff’s supplemental contentions did not modify the scope of its original infringement contentions. Specifically, Plaintiff’s supplemental contentions did not seek to add any new asserted claims or infringement theories. For example, Plaintiff continues to only assert that Defendant directly and/or indirectly infringes Claims 1, 5, 17, and 21 of the ʻ610 Patent (Dkt. #205- 3 at p. 5). Similarly, Plaintiff neither altered its claim construction nor proffered any new claim constructions. Instead, Plaintiff has simply mapped the existing claim limitations in this case to its sole additional supplemental infringement contention, which, again, is based entirely on new and previously unknown information—Photon and Aether source code—that was produced only pursuant to this17 Court’s order.¹⁰ Put simply, Plaintiff’s supplemental contentions do not change its overall infringement case: the same patents, patent claims, and infringement theories apply, and nothing new was added other than one additional infringing operation—Photon’s Shuffled Hash Join. Thus, Plaintiff’s supplemental contentions did not open the door for Defendant to assert, without limitation, entirely new invalidity contentions and prior art references that were not tethered to Plaintiff’s one additional infringing operation.¹¹ Accordingly, the Court finds that Defendant’s supplemental invalidity contentions violate the Court’s June 23 Order because they are not “tied to R2’s supplements related to Photon and Aether.”

B. Defendant Has No Justification for the Late Addition of New Theories and

References.

Plaintiff also argues that Defendant has no justification for the late addition of its new theories and references included in its supplemental contentions (Dkt. #205 at p. 10). Specifically, Plaintiff contends that Defendant’s sole argument in support of its expansive supplementation— that its “supplemental invalidity contentions identify prior art that discloses the same shuffle hash join functionality in a distributed system that R2 now alleges meets the requirements for the asserted claims”—is misleading and incorrect for two reasons (Dkt. #205 at p. 10). First, Defendant is playing word games by simply re-describing the prior art with the name of Photon/Aether’s18 infringing functionality in an attempt to manufacture a connection between PSHJ and its new invalidity theories (Dkt. #205 at p. 10). Second, Defendant’s argument falsely implies that Plaintiff’s mapping of PSHJ somehow raised new unforeseen issues justifying the addition of entirely new art and invalidity theories (Dkt. #205 at p. 11). In support, Plaintiff asserts that its theory as to PSHJ is identical to its theory as to Apache Spark—Plaintiff’s supplement just applied that theory to the newly-produced materials (Dkt. #205 at p. 11). Additionally, Plaintiff contends that the parties pre-Photon/Aether contentions confirm that Defendant is being disingenuous in arguing that its latest supplements are “tied to” Plaintiff’s Photon/Aether supplements (Dkt. #205 at p. 12). For example, all the functionality that Defendant cites its new references for—including “shuffling,” “hashing,” “joining,” and “hash joins”—were discussed at length in Plaintiff’s original infringement read on Apache Spark (Dkt. #205 at p. 12). And it is also indisputable that these topics were explicitly discussed by Defendant in its Third Supplemental Invalidity Contention served seven months ago in response to Plaintiff’s Second Supplemental Infringement contentions (Dkt. #205 at p. 12). From Plaintiff’s perspective, this confirms that Defendant understood in January 2025 that the functionality it now cites its new references for was relevant to validity; however, Defendant failed to identify these references and theories until its Fourth Supplemental Invalidity Contentions (Dkt. #205 at p. 12). As a result, Defendant cannot now claim that its new references and theories are “tied to” Plaintiff’s Photon/Aether supplements, because, according to its own previous contentions, these materials are “tied to” Plaintiff’s original infringement contentions (Dkt. #205 at p. 12).

The Court agrees that Defendant has no justification for the late addition of the new theories and references contained in its supplemental invalidity contentions. Notwithstanding19 Defendant’s best efforts to complicate the issue, it is clear to the Court that Defendant is attempting to use Plaintiff’s supplemental contentions as a pretext to introduce a host of new references that all could (and should) have been included in its previous invalidity contentions. As a preliminary matter, the Court finds that Defendant’s primary argument in support of its expansive supplements—that is, “[f ]or the first time in its supplemental infringement contention R2 alleged infringement by a ʻShuffle Hash Join’ join strategy’” (Dkt. #208 at p. 6)—is refuted by the record in this case. Indeed, a review of each iteration of Plaintiff’s infringement contentions clearly shows that it has always accused infringement by “shuffle hash join.”¹² What’s more, Plaintiff’s infringement theory as to PSHJ is identical to its theory as to Apache Spark (See Dkt. #205-2 at pp. 15, 36–37, 57–59 (applying its infringement theory for Apache Spark to its infringement theory for PSHJ)). The Court finds the Federal Circuit’s decision in Shire LLC v. Amneal Pharms., LLC, 802 F.3d 1301 (Fed. Cir. 2015), to be instructive here. In Shire, the Federal Circuit affirmed the district court’s decision denying defendant’s request to amend its contentions to assert an “on-sale bar” defense as untimely. Id. at 1305, 1309. The Federal Circuit explained that defendant should have asserted its “on-sale bar” defense earlier because “the summary of documents produced by [plaintiff ] on May 21, 2012, indicate that [d]efendants had access to the information regarding the on-sale bar, or documentation that should have led them to it earlier than they now claim.” Id. (citation modified). Here, just like the defendant in Shire, Defendant has had20 access to the information regarding “shuffle hash join” as early as June 2024 when Plaintiff served its preliminary infringement contentions. Nevertheless, Defendant failed to include any of these newly alleged prior art references and theories found in its invalidity contentions; instead, waiting until more than a year later to raise them in its Fourth Supplemental Invalidity Contentions. Thus, Defendant has no justification, other than its own lack of diligence, for failing to assert these invalidity defenses earlier. Accordingly, the Court finds that Defendant’s supplemental contentions are untimely and should be stricken in their entirety.

Moreover, the Court will not reward Defendant for its gamesmanship by allowing its expansive and untethered supplemental contentions—all of which are based on information it has known about since June 2024—to remain at this late of a stage in the proceedings. The Local Patent Rules “exist to further the goal of full, timely discovery and provide all partis with adequate notice and information with which to litigate their cases, not to create supposed loopholes through which parties may practice litigation by ambush.” Computer Acceleration Corp., 503 F. Supp. 2d at 822 (citation omitted). The Court agrees with Plaintiff that allowing Defendant to serve it untimely contentions “would pave the way for future defendants to . . . stonewall discovery until compelled, force patentees to finalize their case without full information, and then use late-produced materials as a springboard for new invalidity theories” (Dkt. #205 at pp. 13–14). See MASS Engineered Design, Inc. v. Ergotron, Inc., 250 F.R.D. 284, 287 (E.D. Tex. 2008) (denying defendant’s request to add its untimely invalidity contentions explaining that doing so “would open the floodgates for other accused infringers to circumvent the Local Patent Rules, thereby completely nullifying Patent Rule 3-3”). Furthermore, the Court is not persuaded by Defendant’s argument that it would “suffer significant unfair prejudice” if the Court allows Plaintiff’s supplements to remain while precluding21 Defendant from presenting its supplemental invalidity contentions made in response (See Dkt. #223 at p. 7). Any prejudice that Defendant may suffer is a direct consequence of its own misconduct in this case.¹³ See Coopervision, Inc. v. Ciba Vision Corp., 480 F. Supp. 2d 884, 889 (E.D. Tex. 2007) (explaining that “[a] party that fails to disclose information in a timely manner has little room to complain” about the new claims contained in plaintiff’s supplemental invalidity contentions); Alexsam, Inc. v. IDT Corp., No. 2:07-CV-420-CE, 2011 WL 108725, at *2 (E.D. Tex. Jan. 12, 2011) (allowing plaintiff’s supplemental infringement contentions explaining that it will not reward a party “for [its] late production of relevant discovery documents by limiting the claims made against them” (citation modified)).

C. The Court Should Not Allow Defendant a Windfall Born of its Own Discovery

Misconduct.

Plaintiff also argues that the Court should strike Defendant’s supplemental invalidity contention in their entirety, because failing to do so would allow Defendant to benefit from its own discovery misconduct (Dkt. #205 a p. 13). Plaintiff contends that this would be fundamentally unfair for two reasons: First, allowing the supplemental invalidity contentions to stand will reward Defendant’s misconduct (Dkt. #205 at p. 13). Second, allowing the latest supplements is severely prejudicial to Plaintiff (Dkt. #205 at p. 14). Here, having already determined that Defendant’s supplemental contentions should be stricken in their entirety for violating the Courts June 23 Order, the Court need not delve into whether Defendant’s discovery misconduct constitutes an22 independent basis for granting Plaintiff’s Motion to Strike Defendant’s Fourth Supplemental Invalidity Contentions. Accordingly, the Court will not address the merits of this argument.

D. Defendant’s Supplement Violates P.R. 3-3(c) and 3-3(b).

Alternatively, Plaintiff argues that Defendant’s supplemental invalidity contentions should be stricken in their entirety for the independent reason that they violate P.R. 3-3(c) and P.R. 3-3(b) (See Dkt. #205 at pp. 14–16). Because Plaintiff offers this as an alternative basis in support of its motion; and having already determined that Defendant’s supplemental invalidity contentions should be stricken in their entirety for violating the Court’s June 23 Order, the Court need not decide whether Defendant’s supplemental contentions should be stricken for the independent reason that they violate P.R. 3-3(c) and 3-3(b). Accordingly, the Court will not address the merits of either parties’ arguments regarding whether Defendant’s supplemental contentions violate P.R. 3-3(c) or P.R. 3-3(b).

E. Defendant’s Invalidity Contentions Compliance with P.R. 3-6

In its Response, Defendant argues, sue sponte, that its supplemental contentions should not be stricken on the basis that pursuant to P.R. 3-6(b) “good cause” exists to allow its supplemental invalidity contentions to remain (Dkt. #208 at p. 15).¹⁴ First, Defendant strictly complied with the Court’s August 22, 2025 deadline to serve supplemental invalidity contentions (Dkt. #208 at p. 16). Second, Defendant’s supplemental contentions are “very important” because they grant a reasonable means to defend against Plaintiff’s new infringement contentions accusing23 “Shuffle Hash Join” (Dkt. #208 at p. 16). Third, Plaintiff will not be prejudiced by its supplemental contentions because they are expressly authorized and in strict accordance with the Court’s order (Dkt. #208 at p. 16). Fourth, no continuance is necessary as the Court’s Amended Scheduling Order not only permits Defendant to supplement its invalidity contentions but also permits supplemental expert reports and depositions (Dkt. #208 at p. 16).

Before addressing the merits of this argument, the Court notes that Defendant’s reliance on Local Patent Rule 3-6(b) to argue that the Court should allow its supplemental contentions to remain is nonsensical. Unlike the other cases dealing with P.R. 3-6(b),¹⁵ the Court has already found that good cause exists and entered an order expressly authorizing narrow supplementation of the parties’ contentions (See Dkt. #200 at pp. 10–13; Dkt. #204 at pp. 1–2 (applying the four-factor test used by courts in this district and finding that good cause exists to amend the scheduling order to allow both parties to serve supplemental contentions)). Indeed, both parties, pursuant to the Court’s order, have already served their supplemental contentions, which are now the subject of the instant Motions before the Court. Thus, this is not a scenario where Defendant needs to request the Court’s permission to file its supplement contentions—Defendant already has such permission. Instead, the issue before the Court is whether the supplemental contentions served by Defendant strictly comply with the Court’s order authorizing narrow supplementation (See Dkt. #204 at p. 1 (authorizing Defendant “to serve its supplemental Invalidity Contentions tied to R2’s supplements related to Photon and Aether)). Nevertheless, even if the Court assumes that24 Defendant raised this as an alternative argument in the event that the Court had found that its supplemental contentions are improper—like it did above—the result is the same. Specifically, after applying the four factors to this case, the Court finds that Defendant has not shown good cause exists to allow its improper supplements to remain.

For the first factor, Defendant’s explanation for its failure to meet the deadline, the Court finds that Defendant has not shown sufficient reasons for why it did not include the current references its earlier invalidity contentions. As discussed above, Plaintiff’s PSHJ infringement theory is identical to its Apache Spark theory—which Plaintiff included in its earlier infringement contentions served more than seven months prior. See supra Section I.B. Thus, Defendant’s failure to meet the deadline for including prior art and invalidity grounds that discloses “shuffle hash joins” stems only from its own negligence. See Finisar Corp., 424 F. Supp. 2d at 902 (finding that defendant “could have done a more complete job of analysis and research earlier in the case in order to comply with its disclosure obligation”). The second factor, the importance of the information, the Court finds that Defendant’s supplemental contentions at issue are of high importance as exclusion would prohibit Defendant from asserting an invalidity defense. Thus, this factor weighs in favor of finding that good cause exists; however, it also underscores Defendant’s inadequate explanation for failing to plead its “highly important” invalidity defenses on time. See MASS Engineered Designs, Inc., 250 F.R.D. at 286 (finding that the fact the motion to strike would exclude highly important information to the case further undermines any explanation given for a party’s failure to meet its disclosure deadline). As to the third factor, considering the potential prejudice, the Court finds that Defendant’s will suffer little to no prejudice. Defendant arguing that it would be prejudiced by Plaintiff’s supplements overlooks that these supplements were a result of the25 Court ordering Defendant to produce relevant documents that it was wrongfully withholding. See Firtiva Corp. v. Funimation Glob. Grp., LLC, No. 2:21-CV-00111-JRG-RSP, 2022 WL 1792818, at *2 (E.D. Tex. June 1, 2022) (rejecting defendant’s prejudice argument because it glosses over the fact that the new opinions were a result of the court ordering defendant to produce documents after the close of fact discovery and after plaintiff served its opening expert report). Moreover, Defendant’s prejudice argument is further undermined by the fact Plaintiff’s supplements did not change or alter its infringement theory; instead, Plaintiff merely substantiated infringement using entirely new information that was not otherwise available to it until recently. See Empire Tech. Dev. LLC v. Samsung Elecs. Co., No. 2:23-CV-00427-JRG-RSP, 2025 WL 1550220, at *1 (E.D. Tex. May 30, 2025) (finding any prejudice on defendant to be minimal where the supplemental source code additions did not change plaintiff’s infringement theory). Finally, as to the fourth factor—the possibility of curing any such prejudice by a continuance—the Court finds that a continuance of the Final Pretrial Conference, the trial, or any other unexpired deadlines is not desirable in this case. Indeed, the Court has already extended important case deadlines on three earlier occasions.¹⁶ See Estech Sys. IP, LLC v. Carvana, LLC, No. 2:21-CV-0482-JRG-RSP, 2023 WL 210837, at *3 (E.D. Tex. Jan. 17, 2023) (quoting S&W Enters., L.L.C. v. SouthTrust Bank of Alabama, NA, 315 F.3d 533, 537 (5th Cir. 2003)) (“In view of district judges’ power to control their dockets by refusing to give ineffective litigants a second change to develop their case . . . we conclude that it was within26 the judge’s sound discretion not to grant a continuance.”). Thus, the Court finds that Defendant has not shown good cause for allowing its supplemental invalidity contentions to remain.

F. Defendant’s Disclaimer Argument

Additionally, Defendant argues that Plaintiff ignores the impact of its prior disclaimer when it argues that Defendant should have raised its supplemental “Shuffle Hash Join” prior art in response to its original infringement contentions (Dkt. #208 at p. 17). Specifically, Defendant contends that in Plaintiff’s Patent Owner Preliminary Response (“POPR”) in IPRs it repeatedly distinguished Shuffle Hash Join from the “entirely different” MapReduce requirements of the ʻ610 Patent asserted claims (Dkt. #208 at p. 17). Indeed, Plaintiff went as far as contending that a hash join is incompatible with MapReduce because the MapReduce architecture “would have to [be] gutted and replaced with a collection of hash join functions, rendering impotent the entire parallelized architecture” (Dkt. #208 at p. 18). Moreover, Plaintiff previously avoided and never mapped a Shuffle Hash Join to each asserted claim requirement in its infringement contentions or in its opening expert report (Dkt. #208 at p. 19). In short, Plaintiff disclaimed Shuffled Hash Join before the parties served their original contentions in June and July 2024, and thus Defendant reasonably understood that “Shuffle Hash Join” was outside of the scope of the asserted claims which caused it not to assert “Shuffle Hash Join” prior art until now (Dkt. #208 at pp. 18– 19).

In response, Plaintiff asserts that Defendant’s argument that Plaintiff disclaimed “Shuffle Hash Join” during IPR fails for two reasons (Dkt. #210 at pp. 6–7). First, Defendant waived the disclaimer arguments by failing to raise them during claim construction (Dkt. #210 at p. 7). Specifically, Plaintiff contends that Defendant has always known of the supposed disclaimer, but it failed to raise it during claim construction despite Plaintiff identifying “Shuffled Hash Join” functionality in all contentions, including before claim construction (Dkt. #210 at p. 7). Second,27 there was no clear and unmistakable disclaimer in IPR as to “Shuffle Hash Join” (Dkt. #210 at p. 7). Plaintiff contends that the term “Shuffle Hash Join” was never used in IPR, making it impossible for it to unmistakenly disclaim it (Dkt. #210 at p. 8). Moreover, Defendant ignores the fact that the Chowdhuri reference Plaintiff distinguished in IPR is not equivalent to PSHJ and thus distinctions over Chowdhuri cannot extend to the now accused PSHJ functionality (Dkt. #210 at p. 8).

The Court agrees with Plaintiff that Defendant’s disclaimer argument is without merit. With respect to disclaimer, the law is clear: “[f ]or prosecution declaimer to attach, our precedent requires that the alleged disavowing actions or statement made during prosecution be both clear and unmistakable.” Droplets, Inc. v. eBay, Inc., No. 2:11-CV-401-JRG-RSP, 2014 WL 4217376, at *35 (E.D. Tex. Aug. 22, 2014) (citing Omega Eng’g, Inc. v. Raytek Corp., 334 F.3d 1314, 1325–26 (Fed. Cir. 2003)). Here, based on the evidence presented, it cannot be said that there was a clear and unmistakable declaimer of “Shuffle Hash Join” by Plaintiff in IPR. Defendant argues that during IPRs Plaintiff repeatedly distinguished “Shuffle Hash Join” from the “entirely different” MapReduce requirements of the ʻ610 Patent asserted claims (See Dkt. #208 at pp. 17– 18). Nonetheless, Plaintiff has persuasively argued that the Chowdhuri hash join reference it distinguished during IPR is not equivalent to PSHJ (See Dkt. #210 at p. 8). Thus, Plaintiff’s distinguishing the “Shuffle Hash Join” disclosed in Chowdhuri has no limiting effect here. See Cordis Corp. v. Medtronic AVE, Inc., 339 F.3d 1352, 1359 (Fed. Cir. 2003) (finding that there is no clear and unmistakable disclaimer where the statement made during prosecution “is amenable to multiple reasonable interpretations”).

• * * *

28In sum, the Court finds that Defendant’s supplemental infringement contentions violate the Court’s June 23 Order because they are not “tied to” Plaintiff’s supplement related to Photon and Aether. However, even if it could be said that Defendant’s supplements complied with the Court’s order (they do not), the Court finds that they should still be stricken based on Defendant’s failure to raise them in their earlier contentions. Accordingly, Plaintiff’s Motion to Strike Defendant Databricks’ Fourth Supplemental Invalidity Contentions (Dkt. #205) in their entirety is hereby GRANTED.

II. Plaintiff R2 Solutions LLC’s Motion to Strike Certain Expert Opinions (Dkt. #207)

Plaintiff’s Motion to Strike Certain Expert Opinions argues that the Court should strike most of Defendant’s Supplemental Opening Expert Report of Dr. Jon Weissman¹⁷ for two main reasons: (1) Dr. Weissman’s Supplemental Opinions are not “tied to” Plaintiff’s supplements; and (2) Defendant has no justification for its late opinions (See Dkt. #207 at pp. 7– 14). The Court need not address the underlying merits of these arguments to reach a decision on the relief requested in Plaintiff’s motion. Having already excluded Defendant’s Fourth Supplemental Invalidity Contentions, see supra Section I, the Court also excludes the September 5th Supplemental Opening Report of Dr. Weissman to the extent that it opines on the stricken invalidity contentions set forth in Defendant’s Fourth Supplemental Invalidity Contentions. In other words, the Court excludes the September 5th Supplemental Opening Report of Dr. Weissman to the extent it opines on invalidity contentions not set forth in Defendant’s January 29th Third Supplemental Invalidity29 Contentions, as these opinions are now irrelevant. Accordingly, Plaintiff’s Motion to Strike Certain Expert Opinions (Dkt. #207) is hereby GRANTED.

III. Defendant Databricks, Inc.’s Motion to Strike R2 Solutions LLC’s Supplemental

Infringement Contentions and Supplemental Opening Expert Report of Mr. William Davis (Dkt. #223) Defendant argues that Plaintiff’s supplemental infringement contentions and its supplemental expert report of Mr. Davis should be stricken in their entirety because Plaintiff has waived its ability to assert its new infringement theory based on “Shuffle Hash Join” (Dkt. #223 at p. 5). Specifically, Defendant contends that Plaintiff has known about “Shuffle Hash Join” since the original round of fact and expert discovery but chose not to accuse it—instead, accusing only a different join strategy called “sort merge join” (Dkt. #223 at p. 5). Thus, Plaintiff cannot resurrect its abandoned theory by arguing that shuffle hash join in Spark and Photon allegedly infringe for “identical” reasons (Dkt. #223 at p. 7). Moreover, Defendant argues that the four factors set forth in Hayes v. Locke Supply Co., weigh strongly in favor of striking Plaintiff’s supplemental contentions and supplemental expert report (Dkt. #223 at p. 7).¹⁸ First, Plaintiff has no justifiable excuse for failing to include its shuffle hash join theory in its original report other than for “simplicity’s sake” which is not a valid justification (Dkt. #223 at p. 7). Second, choosing to abandon the theory undermines any alleged importance (Dkt. #223 at p. 7). Third, Defendant would be unfairly prejudiced if Plaintiff’s supplements were not stricken (Dkt. #223 at p. 7). Specifically, Plaintiff seeks to completely change course to now allege a new infringement theory based on “Shuffle Hash Join,” yet also seeks to preclude Defendant from presenting its timely disclosed invalidity theories30 tied to its untimely infringement theories (Dkt. #223 at p. 7). Fourth, no amount of continuance will cure such prejudice (Dkt. #223 at p. 7). As such, Mr. Davis’s untimely opinions regarding “Shuffle Hash Join” should be stricken (Dkt. #223 at p. 7).¹⁹

In response, Plaintiff argues that there is no merit to Defendant’s motion as Defendant knows well that its supplemental infringement contentions and expert reports are timely and expressly authorized by the Court’s orders (Dkt. #227 at p. 4). Specifically, Plaintiff contends that Defendant’s only argument in support of its motion—that is, Plaintiff waived its theory against PSHJ because it could have, but did not, offer it earlier—fails for numerous reasons (Dkt. #227 at p. 4). First, Plaintiff’s supplements cannot be improper because they were already explicitly authorized by the Court (Dkt. #227 at pp. 4–5). Second, Plaintiff relied entirely on new information, produced only pursuant to the Court’s order, to supplement its infringement case; indeed, Plaintiff could not have provided its supplemental contentions or expert opinions on PSHJ without the compelled information (Dkt. #227 at p. 5). The fact that it previously served contentions regarding shuffled hash join found in spark is a red herring—Plaintiff could not possibly have known that Photon/Aether infringed via shuffle hash join functionality until it saw it (Dkt. #227 at p. 5). Third, Defendant’s insistence that PSHJ and Spark shuffled hash join (“SSHJ”) is also immaterial (Dkt. #227 at p. 5). Identical or not, Plaintiff had no way of knowing that PSHJ and SSHJ worked the same until it looked at the PSHJ code, which was the lynchpin to its infringement theory based on31 Photon/Aether (Dkt. #227 at pp. 5–6). Lastly, none of the Hayes factors Defendant discusses in its motion come close to substantiating the striking of Plaintiff’s supplements (See Dkt. #227 at p. 6). In sum, Plaintiff’s PSHJ supplements are timely and authorized by the Court’s order and thus Defendant lacks any basis to have them stricken in their entirety.

The Court finds that Plaintiff did not waive its ability to assert its infringement theory based on “Shuffle Hash Join.” Defendant’s reliance on Astellas as a basis for arguing that the Court should strike the entirety of Plaintiff’s supplemental infringement contentions and the entirety of the supplemental expert report of Mr. Davis is misplaced (See Dkt. #233 at pp. 8–9). In Astellas, the plaintiff’s new infringement theory made during the supplemental discovery period was based on documents produced over a year before the theory was disclosed and thus could have been raised during the original discovery period. See Astellas, 2022 WL 17998229, at *3 (explaining that because plaintiff’s theory “was based on documents produced to [plaintiff ] in August 2020, [plaintiff ] could have raised this theory during the original period of discovery (which ended in October 2020 for fact discovery and April 2021 for expert discover” (citation modified)). As a result, the Federal Circuit concluded that “[t]here was no reason, other than [plaintiff’s] own litigation choices, that the [new] infringement theory could not have been asserted earlier.” Id. (citation modified). In making its ruling, the court, however, expressly recognized that this was “not a case where [plaintiff ] relied on new information . . . to craft a new theory of infringement.” Id. (citation modified). Here, unlike the plaintiff in Astellas, Plaintiff’s PSHJ infringement theory is based exclusively on new information—Photon and Aether source code as well as accompanying technical documentation—that Defendant produced only pursuant to the Court’s order compelling it to do so (See Dkt. #200). It cannot be said, then, that there was “no reason, other than [plaintiff’s] own32 litigation choices, that the [new] infringement theory could not have been asserted earlier.” Astellas, 2022 WL 17998229, at *3 (citation modified). Indeed, if anything, Plaintiff’s failure to raise its PSHJ theory earlier was attributed solely to Defendant’s own litigation misconduct—that is, voluntarily withholding critical materials throughout fact discovery, expert reporting, and dispositive motions practice. Thus, this case presents the alternative scenario that the court in Astellas indicates would have led the court to reach a different result—that is, where Plaintiff “relies on new information” to craft the new infringement theory contained in its supplemental contentions. See Astellas, 2022 WL 17998229, at *3. Accordingly, the Court finds that Plaintiff did not waive its ability to assert its new infringement theory based on “Shuffle Hash Join,” and thus Defendant’s request for them to be stricken in their entirety on this basis is hereby DENIED.

Moreover, the Court also finds that the four factors set forth in Hayes do not “weigh strongly in favor of striking R2’s supplemental contention and supplemental expert report” (Dkt. #223 at p. 7). First, Plaintiff does indeed have a justifiable explanation for failing to include its “Shuffle Hash Join” theory in its earlier contentions: Defendant’s decision to withhold critical materials throughout fact discovery, expert reporting, and dispositive motion practice until being compelled by this Court to produce them. See STMicroelectronics, Inc. v. Motorola, Inc., 307 F. Supp. 2d 845, 852–53 (E.D. Tex. 2004) (explaining that where a patent plaintiff’s supplemental contentions are based on new information obtained during discovery which, despite all diligence was not available when its original disclosure were due, constitutes good cause warranting the untimely supplementation). Second, it would be impossible for Plaintiff to abandon an infringement theory that remained completely unknown to Plaintiff until just recently when Defendant was forced to produce the relevant information that it had been withholding—i.e., Photon/Aether source code33 as well as accompanying technical documentation. See Ciena Corp. v. Nortel Networks Inc., 233 F.R.D. 493, 496 (E.D. Tex. 2006) (allowing party to serve supplemental contentions containing new claims where such claims were based on additional information it acquired after the deadline to amend had passed); see also O2 Micro, 467 F.3d at 1366 (explaining that “refusing to allow any amendment to contentions based on new information developed in discovery could be contrary to the spirit of the Federal Rules.” (citation modified)).²⁰ Third, if Plaintiff’s supplements are permitted, Defendant would suffer little to no prejudice whereas Plaintiff would suffer significant unfair prejudice. As already discussed ad nauseam by this Court above, see supra Sections I.A., I.B., Plaintiff’s supplements do not, as Defendant repeatedly contends, “completely change course” or “allege for the first time” infringement via “Shuffle Hash Join” functionality (See Dkt. #208 at p. 6; Dkt. #223 at p. 7). Instead, Plaintiff has just merely substantiated its infringement case using entirely new information that was not otherwise made available to it until just recently. See supra34 Section I.A.²¹ In contrast, unlike Plaintiff, Defendant already had ample time to include these new references and opinions in its earlier contentions, as Plaintiff’s original infringement contentions on Apache Spark (served on January 9, 2025) discussed its “Shuffle Hash Join” theory at length (See Dkt. #205 at pp. 11–12). Indeed, Plaintiff’s January 9 infringement contentions were the operative contentions underlying Defendant’s original expert report (Dkt. #205 at p. 12), and thus Defendant’s failure to include these new references and opinions at that time is a result of only its own negligence in defending against this patent infringement suit. See Finisar Corp Inc., 424 F. Supp. 2d at 902 (striking most of the new prior art references in defendant’s supplemental contentions finding in part that it “could have done a more complete job of analysis and research earlier in the case in order to comply with its disclosure obligations.”); see also Estech Sys. IP, LLC, 2023 WL 210837, at *3 (denying supplemental contentions derived from other sources independent of those contained in the supplemental discovery responses explaining that it refuses to give35 “ineffective litigants a second change to develop their case”).²² Fourth, having found that allowing Plaintiff’s supplements will cause Defendant to suffer little to no prejudice, considering whether a continuance would cure any such prejudice is unnecessary. However, even if Plaintiff’s supplements were prejudicial to Defendant (they are not), the Court finds that under these circumstances a continuance is not desirable.²³ Thus, contrary to Defendant’s arguments otherwise, each of the four factors set forth in Hayes weigh strongly in favor of not striking Plaintiff’s supplemental contentions and supplemental expert report. Accordingly, the Court finds that Defendant’s requests for Plaintiff’s supplements to be stricken in their entirety on this basis is hereby DENIED.

36IV. Defendant Databricks, Inc.’s Motion for Supplemental Claim Construction (Dkt.

#224) Defendant’s Motion for Supplemental Claim Construction argues that the Court should supplement its construction of the “reducing”²⁴ claim limitation because (1) the parties raise an actual dispute regarding the scope and meaning of the “reducing” element of the asserted claims (Dkt. #224 at pp. 13–14); and (2) Plaintiff clearly and unmistakenly disclaimed hash joins from the “reducing” requirement of the asserted claims (Dkt. #224 at pp. 14–19). The Court addresses each argument in turn.

A. The Court Should Resolve the Parties’ Dispute Regarding the Scope of the

“Reducing” Element of the Asserted Claims.

At the outset, Defendant argues that O2 Mirco requires the Court to intervene to resolve the parties dispute regarding the scope and meaning of the “reducing” claim limitation (See Dkt. #224 at p. 13 (arguing that when the parties raise an actual dispute regarding the proper scope of these claims, the Court, not the jury must resolve the dispute, and the failure to do so constitutes reversible error)). Specifically, Defendant contends that the dispute arises from Plaintiff’s new infringement theory that a generic hash join implemented by Defendant’s accused products meets the “reducing” requirement of the asserted claim (Dkt. #224 at p. 14). According to Defendant, its current construction of “reducing” defines this limitation in light of the intrinsic record— specifically, Plaintiff’s repeated statements disclaiming hash join from the scope of the “reducing” limitation of the asserted claims (Dkt. #224 at p. 14). Plaintiff new infringement theory, however, requires disregarding this disclaimer and adopting a broader interpretation of “reducing” as37 including hash join functionality (Dkt. #224 at p. 14). Thus, there remains a clear dispute between the parties regarding the meaning and scope of the “reducing” claim limitation making supplemental construction of this term by the Court necessary (Dkt. #224 at p. 14).

In response, Plaintiff argues that there is no actual dispute between the parties regarding the scope and meaning of the “reducing” limitation (Dkt. #233 at p. 12). Plaintiff contends that, contrary to Defendant’s belief, the salient and undisputable fact is that it identified hash join functionality and explained how it meets the “reducing” limitation numerous times before claim construction was complete (Dkt. #233 at p. 12). Indeed, Defendant even responded in kind citing prior art disclosures of hash joins in its invalidity contentions (Dkt. #233 at p. 12). According to Plaintiff, there can be no doubt, then, that Defendant had every opportunity to presents its current construction to the Court during the Markman proceedings: it was on notice of Plaintiff’s theory, it timely argued for IPR disclaimer of other terms, and it even argued for (and prevailed on) a different construction for the “reducing” limitation (Dkt. #233 at p. 13). Thus, O2 Micro does not require the Court to entertain Defendant’s request to reopen claim construction under these circumstances (Dkt. #233 at p. 13).²⁵

This case presents a scenario in which more than one year after the Court issued its Claim Construction Memorandum Opinion and Order (Dkt. #71), Defendant now claims, for the first time, that the parties have a fundamental dispute regarding the scope of the “reducing” claim term. Defendant contends that, as a result, under O2 Mirco the Court has a duty to resolve the claim38 construction dispute prior to trial (See Dkt. #224 at pp. 13–14). The Court disagrees. It is well- established that an O2 Mirco issue arises only when the parties raise an actual dispute regarding the scope of the asserted claims. See O2 Micro, 521 F.3d at 1360 (“When the parties raise an actual dispute regarding the proper scope of these claims, the court, not the jury, must resolve that dispute.”). “There is not necessarily an O2 Micro issue, however, whenever further claim construction could resolve the parties dispute—rather, a party must sufficiently request further construction of the relevant limitation to raise an actual dispute.” Kaufman v. Microsoft Corp., 34 F.4th 1360, 1369 (Fed. Cir. 2020) (citation modified); see also Eon Corp. IP Holdings v. Silver Spring Networks, 815 F.3d 1314, 1319 (Fed. Cir. 2016) (“[A] district court’s duty at the claim construction stage is . . . to resolve a dispute about claim scope that has been raised by the parties” (citation modified)). Thus, “if there is no indication that the court was aware of the supposed [claim construction] dispute, a party is considered to have forfeited the O2 Micro issue and cannot [seek to] resurrect its argument [later on]” Kaufman, 34 F.4th at 1369–70 (citation modified). Here, in light of Defendant’s failure to timely raise this construction dispute regarding the “reducing” limitation during the Markman hearing—opting instead to wait and present its construction argument for the first time in its motion filed more than a year later—the Court finds that it failed to preserve its O2 Miro challenge.

As a preliminary matter, the Court finds that there is ample evidence in the record to support its finding that Defendant failed to preserve its O2 Micro challenge. Despite Defendant’s arguments to the contrary, each iteration of infringement contentions that Plaintiff has served reveals that it has always alleged that the “hash join” functionality meets the “reducing” limitation39 of the asserted claims.²⁶ Likewise, each iteration of invalidity contentions that Defendant has served has also always alleged that the prior art hash join functionality disclosed the “reducing” limitation.²⁷ In sum, the record reveals that Defendant has always been (or at least should have been) aware that Plaintiff has accused “hash join” functionality of meeting the “reducing” limitation both before and after claim construction took place. At no time, however, did Defendant ever argue that Plaintiff had disclaimed that the “reducing” limitation could be met by “hash join” functionality until well into the late stage of these proceedings. Furthermore, Defendant not only failed to timely object to the Court’s construction of “reducing,” but it defended the construction of “reducing” that this Court largely adopted in its claim construction order.²⁸ As such, the Court agrees with Plaintiff that Defendant “had every opportunity to present its current construction to the Court during Markman proceedings” yet failed to raise it for more than a year after the Markman (Dkt. #233 at pp. 13–14). In fact, this is precisely the type of eleventh-hour attempt to litigate a newly minted claim construction controversy that courts have routinely rejected. See Video Sols. Pte. Ltd. v. Cisco Sys., Inc., No. 2:23-CV-00222-JRG, 2025 WL 1397200, at *2 (E.D. Tex. May40 14, 2025) (denying defendant’s motion for further construction of “endpoint” finding that the motion is precisely the type of “eleventh-hour attempt” to propose new construction that courts have discretion to not consider); Broadcom Corp. v. Qualcomm Inc., 543 F.3d 683, 694 (Fed. Cir. 2008) (refusing to consider defendant’s claim construction argument finding that it is an improper “eleventh-hour attempt” to litigate a new claim construction controversy that was not previously raised with the district court). Thus, the Court finds that even if the parties’ disagreement concerned the scope of the claim limitation, Defendant waived any claim construction argument it may have had. See Bettcher Indus., Inc. v. Bunzl USA, Inc., 661 F.3d 629, 640–41 (Fed. Cir. 2011) (citation modified) (affirming district court’s decision to prohibit the defendant from adding new claim construction theories where the defendant “had ample opportunity to seek construction of the [claim] limitation” but failed to do so until “a year after the Markman hearing”).

Moreover, the Court finds that O2 Micro is distinguishable. In O2 Micro, the district court was clearly aware of the parties’ disagreement about the claim term “only if,” and the court refused to construe it beyond its ordinary meaning. See O2 Micro, 521 F.3d at 1357 (citation modified) (“The parties agreed, for the most part, that a previously issued claim construction order . . . controlled in this case . . . . However, the parties presented a handful of additional terms for the court to construe of which ʻonly if’ was one.’”); id. at 1361 (“The parties presented a dispute to the district court regarding the scope of the asserted claims.”); see also id. at 1361 (“”[T]he district court failed to resolve the parties’ dispute because the parties disputed not the meaning of the words themselves, but the scope that should be encompassed by this claim language.” (emphasis in original)). In other words, in O2 Micro the claim construction “issues were fully litigated and decided at the Markman stage of the litigation.” Id. at 1359. Here, unlike O2 Micro, Defendant41 expressly agreed to both the meaning and scope of the term during claim construction: “reducing” means “representing all mapped (intermediate) key-value pairs sharing the same key as a single key-value pair or a list of values” (Dkt. #71 at p. 21). Indeed, Defendant agreed to this meaning with no further limitations, including the negative limitation that it now requests in this case. The lack of further limitations was itself a characteristic of the construction to which both parties agreed. Defendant cannot argue at this late stage in the proceedings—after the majority of the parties’ trial preparation has been framed and directed by the Markman construction—that it is dissatisfied with its own proposed construction. Having obtained its desired construction of “reducing,” Defendant should not be allowed to reverse course and demand a new construction of that very same term. To do otherwise would be to allow any party who later becomes dissatisfied with the operative claim construction order in the case to unduly delay trial by making a similar eleventh-hour request for further claim construction. See Nuance Commc’ns, Inc. v. ABBYY USA Software House, Inc., 813 F.3d 1368, 1373 (Fed. Cir. 2016) (emphasis added) (“At the Markman hearing, the district court found in Nuance’s favor by adopting the plain and ordinary meaning of the term “identifying.” The fact that shortly before trial Nuance became dissatisfied with its own proposed construction and sought a new one does not give rise to an O2 Micro violation.” (emphasis added)). Thus, contrary to Defendant’s belief, O2 Micro does not require the Court to reopen claim construction and supplement its prior construction of the “reducing” limitation in these circumstances.

B. Plaintiff Clearly and Unmistakenly Disclaimed Hash Join from the

“Reducing” Requirement of the Asserted Claims.

Next, Defendant argues that the Court should supplement its construction of the “reducing” claim limitation because Plaintiff clearly and unmistakenly disclaimed hash join from42 the “reducing” requirement of the asserted claims (Dkt. #224 at pp. 14–19). Defendant contends that Plaintiff’s arguments repeatedly, clearly, and unmistakably distinguishing Chowdhuri’s hash join during IPR amount to a disclaimer that require imposing a negative limitation—that is, construing the “reducing” limitation as excluding “hash joins” (Dkt. #224 at pp. 16–17).²⁹ For example, Plaintiff repeatedly stated that the Chowdhuri hash join “work[s] very differently” than the MapReduce architecture of the ʻ610 Patent—with “Chowdhuri [working] through hash join functions, and MapReduce [working] through phases of distributed map and reduce functions” (Dkt. #224 at p. 17). Indeed, Plaintiff even concluded that “Chowdhuri discloses an entirely different mechanism than MapReduce to process data, including through hash join functions” (Dkt. #224 at p. 17). From Defendant’s perspective, Plaintiff’s IPR statements shown that it clearly and unmistakably interpreted the “reducing” limitation as different from and excluding the “hash join” functionality it now accuses in its most recent supplements (Dkt. #224 at p. 17). In other words, Plaintiff’s supplements now accuse the same generic hash join functionality for its PSHJ theory that it previously disclaimed during IPR (See Dkt. #224 at pp. 12, 17). Thus, Plaintiff should not be permitted to obtain a broader construction of “reducing” than the position it advocated to the PTAB (Dkt. #224 at p. 17).

In response, Plaintiff argues that it did not clearly and unmistakably disclaim all “hash join” functionality from meeting the “reducing” limitation (Dkt. #233 at p. 15). Specifically, Plaintiff contends the standard for prosecution disclaimer is exacting and is not met here for at least four43 reasons (Dkt. #233 at p. 15). First, during IPR Plaintiff expressly recognized functional overlap between “hash joins” and “reducing” operations, noting that a hash join can serve similar purposes as, or even replace, a reduce function (Dkt. #233 at p. 16). Second, when taken in context, Plaintiff’s statements do not rise to the level of clear and unmistakable disclaimer that no hash join could ever accomplish the claimed “reducing” limitation (Dkt. #233 at p. 16). Indeed, telling is the fact that not once in its substantive arguments does Defendant offer a full and accurate quote, much less in the proper context (Dkt. #233 at p. 16).³⁰ Third, the broader litigation record reveals that there is no clear and unmistakable disclaimer (Dkt. #233 at p. 18). For example, during claim construction Defendant pursued and obtained a construction of “reducing” without asking the Court to exclude “hash join” functionality (Dkt. #233 at p. 19). Lastly, even if its IPR statements could be read to disclaim Chowdhuri’s hash joins from meeting the “reducing” limitation, such disclaimer cannot extend to categorically exclude all hash join functionality, much less the specific PSHJ hash join functionality at issue (Dkt. #233 at p. 19). In sum, Plaintiff did not announce a technology-wide rule that “hash joins” can never satisfy the “reducing” limitation and thus there is no basis to rewrite “reducing” to exclude “hash joins” (Dkt. #233 at p. 19).

The Court agrees with Plaintiff that it did not disclaim “hash join” functionality from meeting the “reducing” limitation of the asserted claim during IPR. As mentioned above, the law regarding prosecution disclaimer is clear: “[f ]or prosecution declaimer to attach, our precedent requires that the alleged disavowing actions or statement made during prosecution be both clear and unmistakable.” Droplets, Inc., 2014 WL 4217376, at *35 (citation omitted). Here, despite44 narrowing its disclaimer argument,³¹ the Court again finds that Defendant failed to prove that Plaintiff’s statements during IPR amount to a “clear and unmistakable” disclaimer that requires reading a negative limitation into the “reducing” limitation to exclude “hash join” functionality. See Azurity Pharms., Inc. v. Alkem Lab’ys Ltd., 133 F.4th 1359, 1366 (Fed. Cir. 2025) (“The party seeking to invoke prosecution history disclaimer bears the burden of proving the existence of a ʻclear and unmistakable’ disclaimer that would have been evidence to one skilled in the art.’” (citation omitted)). Defendant relies on Plaintiff’s statements in its Patent Owner’s Preliminary Response regarding how Chowdhuri’s “hash join” functionality works very differently and does not correspond to the MapReduce architecture of the ʻ610 patent (See Dkt. #224 at pp. 16– 17 (arguing that these IPR statement show that Plaintiff clearly and unmistakably interpreted “reducing” as different from and excluding “hash joins”)). Nonetheless, Plaintiff has persuasively demonstrated that its IPR statements distinguishing Chowdhuri’s “hash join” functionality were not made specifically with regard to the ʻ610 patent or the “reducing” limitation (See Dkt. #233 at pp. 16– 19 (arguing that its IPR statements were distinguishing Chowdhuri from Pike and not Chowdhuri from the ’610 Patent as well as attacking Defendant’s misguided combination of Pike and Chowdhuri)). Instead, its IPR statements distinguishing Chowdhuri were contextually limited to grounds wholly unrelated to its use of “hash join” functionality (See Dkt. #233 at pp. 18–19 (arguing that its statements distinguished a specific prior-art architecture that bypassed mapping and intermediate data and did not announce a technology-wide rule that “hash joins” can never45 satisfy the “reducing” limitation)). Thus, Plaintiff’s IPR statements distinguishing Chowdhuri’s “hash join” functionality has no limiting effect on the construction of the “reducing” limitation. See Avid Tech., Inc. v. Harmonic, Inc., 812 F.3d 1040, 1045 (Fed. Cir. 2016) (citation modified) (“Where the alleged disavowal is ambiguous, or even amendable to multiple reasonable interpretations, we have declined to find prosecution disclaimer.”).

• * * *

In sum, both of Defendant’s arguments offered in support of its motion requesting that the Court supplement its construction of the “reducing” claim limitation are without merit. First, Defendant failed to show that an O2 Micro issue exists in this case; however, even if it did, the Court finds that Defendant waived any O2 Micro challenge that it may have had by waiting until more than a year after the Markman hearing to raise it. Second, Defendant has failed to show that Plaintiff’s IPR statements amount to a clear and unmistakable disclaimer of “hash joins” from the “reducing” claim limitation. Accordingly, Defendant Databricks, Inc.’s Motion for Supplemental Claim Construction (Dkt. #224) is hereby DENIED.

CONCLUSION

It is therefore ORDERED that Plaintiff R2 Solutions LLC’s Motion to Strike Databrick’s Fourth Supplemental Invalidity Contentions (Dkt. #205) is hereby GRANTED.

It is further ORDERED that Plaintiff R2 Solutions LLC’s Motion to Strike Certain Expert Opinions (Dkt. #207) is hereby GRANTED.

It is further ORDERED that Defendant Databricks Inc.’s Motion to Strike R2 Solutions LLC’s Supplemental Infringement Contentions and Supplemental Opening Expert Report of Mr. William Davis (Dkt. #223) is hereby DENIED.

46It is further ORDERED that Databricks Inc.’s Motion for Supplemental Claim Construction (Dkt. #224) is hereby DENIED.

IT IS SO ORDERED.