TexasLDPC Inc. v. Broadcom Inc.

IN THE UNITED STATES DISTRICT COURT FOR THE DISTRICT OF DELAWARE TEXASLDPC INC., Plaintiff, v. No. 18-cv-1966-SB BROADCOM INC., LSI CORP., and AVAGO TECHNOLOGIES U.S. INC., Defendants. Gregory R. Booker, Warren K. Mabey, Jr., FISH & RICHARDSON, P.C., Wilmington, Delaware; Lawrence K. Kolodney, FISH & RICHARDSON, P.C., Boston, Massachusetts; David M. Hoffman, FISH & RICHARDSON, P.C., Austin, Texas; Rodeen Talebi, FISH & RICHARDSON, P.C., Dallas, Texas. Counsel for Plaintiff. Adam W. Poff, Robert M. Vrana, YOUNG CONAWAY STARGATT & TAYLOR, LLP, Wilmington, Delaware; Steven J. Rizzi, Ramy E. Hanna, MCKOOL SMITH, PC, New York, New York. Counsel for Defendants. MEMORANDUM OPINION August 4, 2021 BIBAS, Circuit Judge, sitting by designation. When a plaintiff has brought plausible claims of infringement, dismissing its whole case before discovery would be an extreme remedy, and far from proper here. After TexasLDPC brought patent and copyright claims, the defendants lobbed back allegations of serious misconduct by the plaintiff and its lawyers. Because I was concerned, I let the defendants take some threshold discovery. Yet discovery has passed, and the defendants have found no proof. They still have not asked me to address any litigation misconduct. Instead, they raise two other arguments. First, they complain that the case is missing a required plaintiff. But they have failed to show enough prejudice to justify throwing out the case. Next, they argue that TexasLDPC granted them an implied license to use the inventions. But their evidence shows the opposite. I will thus deny their motions. Finally, because the defendants never asked for any relief on litigation misconduct, TexasLDPC has now moved for a ruling that there was none. I will deny that motion too: if the defendants find new evidence, they may try their misconduct argument later. I. BACKGROUND Kiran Gunnam was a Ph.D. student at Texas A&M University. D.I. 122-1, at 293. He studied decoders, devices for receiving data over noisy channels. Id. at 395. His research led to some patents and copyrights. See, e.g., id. at 325, 368; D.I. 42-7. Under the terms of his enrollment, he had to assign these rights to Texas A&M. D.I. 122-1, at 97–98. But if the school made money off his research, he would get about a 30% cut. Id. at 44, 61. After graduating, Gunnam took a job at LSI Corp. There, he kept working on decoders. He claims that some of his research at LSI involved improving his Ph.D. work. Id. at 299. From the start, he tried to get LSI to license that work. Id. at 149–150, 245. To help convince it, he showed it his inventions (with the school’s permission, he says). Id. at 148–49. But in the end, the negotiations fell through, LSI never bought a license, and Gunnam quit. Id. at 299; D.I. 79 ¶ 49. Now, Gunnam fears that LSI is infringing his Ph.D. work. He thinks that to use his improvements on that work, LSI also needs a license for his underlying graduate work—one it does not have. Id. at 116. But since Gunnam signed away his work to Texas A&M, he cannot sue for infringement himself. Yet his wife has found a way. She formed a company, TexasLDPC. In 2015, TexasLDPC bought a license from Texas A&M to Gunnam’s Ph.D. work. The license gave the company the exclusive right to sue for patent and copyright infringement. D.I. 119-1, at 2–4, 12. The right ends once the last patent expires (but before the copyrights do). Id. at 8. After TexasLDPC got the license, it sued LSI and some corporate affiliates for patent and copyright infringement. Although I denied the defendants’ motion to dismiss, I was concerned by Gunnam’s dual role in this case: he may have helped the defendants infringe but is married to the plaintiff’s owner. So I ordered threshold discovery on three issues: whether TexasLDPC is a proper plaintiff absent Texas A&M; whether it is barred from suing based on Gunnam’s time at LSI; and whether in suing, it wrongly used LSI’s confidential information. After discovery, the defendants asked me to dismiss the case on the first two grounds. They did not move on the third ground, and now TexasLDPC asks me to rule that there was no litigation misconduct. I will deny all the motions. II. THE CASE MAY GO ON WITHOUT TEXAS A&M Although Texas A&M owns the patents and copyrights at issue, it is not a plaintiff. The defendants object to that. They argue that the school is an indispensable party. But because it is a state school, it has sovereign immunity and cannot be forced to join this case. Gensetix, Inc. v. Bd. of Regents of Univ. of Texas Sys., 966 F.3d 1316, 1323 (Fed. Cir. 2020). Thus, the defendants say, I should dismiss the whole case. I will not. Even if having Texas A&M here would be better, the unfairness of end- ing this case before full discovery greatly outweighs any from pressing on. A. Although Texas A&M might be a “required party,” the case does not automatically end without it Under Rule 19, a court must try to add certain “required” parties to a case. Fed. R. Civ. P. 19(a)(2). In an infringement case, that includes the patent or copyright owner. Lone Star Silicon Innovations LLC v. Nanya Tech. Corp., 925 F.3d 1225, 1236 (Fed. Cir. 2019); Charles Alan Wright, Arthur R. Miller, Mary Kay Kane, 7 Federal Practice and Procedure § 1614 (3d. ed. 2021). The parties debate whether Texas A&M is a required party. The defendants argue that the school must be joined. TexasLDPC responds that it need not be joined because its license effectively makes TexasLDPC the patent owner. I need not resolve this debate. If a court cannot add a required party (because it has sovereign immunity, for instance), the case does not automatically end. Instead, the court must decide “whether, in equity and good conscience, the action should proceed among the existing parties or should be dismissed.” Fed. R. Civ. P. 19(b). In deciding that, the court must consider “whether the plaintiff would have an adequate remedy if the action were dismissed.” Id. It must then weigh that concern against any prejudice from pressing on. In doing so, the court must consider whether “a judgment rendered in the [required party’s] absence would be adequate.” Id. Here, these factors cut strongly against dismissing. B. The case should go on without Texas A&M If I dismissed the case, TexasLDPC would suffer enormous prejudice: it would have no remedy for infringement and could lose a valuable contract right. This “coun- sels strongly against dismissal.” Gardiner v. V.I. Water & Power Auth., 145 F.3d 635 , 642 (3d Cir. 1998). The defendants respond that continuing without Texas A&M would be unfair to both the school and them. They are probably wrong about the school. And any prejudice that they might suffer would not be enough to warrant my ending the case. 1. Texas A&M will be fine. The defendants’ main argument for dismissal is a sud- den anxiety for their would-be opponent, Texas A&M. They have drummed up a few ways in which the school’s interests are not perfectly aligned with TexasLDPC’s. For example, the school has licensed Gunnam’s Ph.D. work to another company, Marvell. D.I. 119-1, at 3. Thus, the defendants say, TexasLDPC might argue for a reading of the patents that excludes Marvell’s applications, and so threaten the school’s Marvell royalties. This convenient altruism for Texas A&M is not convincing. For one thing, these risks are theoretical. The defendants have pointed to no specific reason why TexasLDPC’s patent theories would undercut Marvell’s applications. If anything, TexasLDPC has an interest in the patents’ covering Marvell’s applications: it has a license for all applications, and thus Marvell’s too. Cf. A123 Sys., Inc. v. Hydro-Quebec, 626 F.3d 1213 , 1221–22 (Fed. Cir. 2010) (upholding a dismissal under Rule 19(b) when the plaintiff had a license for only some applications of a patent and when it still could sue elsewhere). The defendants’ other arguments are even weaker. They claim that they or TexasLDPC will try to shift the blame to Texas A&M. But they have not explained why this would have any legal effect on the school. They also claim that the school could get more royalty money if TexasLDPC waited to sue. But the contract lets the company sue now, so any harm would not be prejudicial. Plus, the school itself does not seem worried. It gave TexasLDPC the exclusive right to sue. D.I. 119-1, at 3–4. It knows about this case. It has even taken part in discovery. And yet it has neither joined the case nor asked me to dismiss it—not even in an amicus brief. This “strongly suggest[s] that its interest will not be impeded if the suit goes forward without it.” Gardiner, 145 F.3d at 641. In fact, the school would likely be much worse off if I dismissed the suit: it gets a cut of any winnings. D.I. 1991, at 4. Still, there is a slight risk that this case could backfire against Texas A&M. The defendants argue that this is enough; I must dismiss the case if there is even a “potential for injury to the interests of the absent sovereign.” D.I. 118, at 14 (quoting Republic of the Philippines v. Pimentel, 553 U.S. 851, 867 (2008)). But the defendants misread Pimentel. That “case dealt with foreign sovereign im- munity, not state sovereign immunity, and the Court placed some weight on the comity and dignity interests at play.” Gensetix, 966 F.3d at 1326 (internal quotation marks omitted). Plus, in Pimentel, the sovereign was effectively a would-be defendant, would have been clearly worse off if the case had gone on, and said so vocally. Pimentel, 553 U.S. at 858–59; see id. at 867 (“[D]ismissal of the action must be ordered” when “sovereign immunity is asserted.” (emphasis added)). Here, by contrast, Texas A&M has not objected to the case, likely wants it to go on, and has even submitted to discovery. I will not dismiss the case because of the off chance that the school is secretly opposed to it and has just been shy. 2. Any prejudice to the defendants will be mild. More likely, the defendants are worried about themselves. And they have identified some real potential prejudice. But not enough. Two of their arguments are baseless. First, they say that they will have a harder time getting discovery from Texas A&M if it is not a party. But the defendants have not explained why that would change things. So far, the school has cooperated, turning over thousands of documents and offering a deposition witness. Until that changes, the defendants have not shown prejudice. See Gardiner, 145 F.3d at 641. Second, the defendants are worried that when TexasLDPC’s license to sue expires, the case will still be pending and it will have to start over with Texas A&M as the plaintiff. I doubt that. The license does not expire until at least 2028. D.I. 118, at 16. This case will surely be over by then. Still, the defendants do raise a fair point about serial litigation: After the patents expire and the license ends, the copyrights will still be valid and will revert to Texas A&M. So even if the defendants beat the copyright claims now, they might face another suit from the school in a decade. Thus, a judgment for the defendants might not give them complete relief. See Fed. R. Civ. P. 19(b)(3). (Likewise, the patent licenses could end early, so there is an outside risk of a second patent suit.) This is a reasonable point, but it is not enough. Parties must face multiple suits on the same subject all the time. It is inconvenient but not the end of the world. This mild potential prejudice pales next to TexasLDPC’s losing its whole case. The balance is not close. “In equity and good conscience,” this case should go on. I will thus deny the motion to dismiss. III. TEXASLDPC DID NOT GIVE THE DEFENDANTS AN IMPLIED LICENSE Next, the defendants ask me to rule that they have an implied license to use Gunnam’s Ph.D. work. They do not. For their implied license, the defendants point to “legal estoppel.” Under this doctrine, when you sell a patent for an invention, you also implicitly sell any other patents you own (or will own) that are needed to make the invention. TransCore, LP v. Elec. Transaction Consultants Corp., 563 F.3d 1271 , 1278–79 (Fed. Cir. 2009). This rule is “an interpretative presumption” that “parties are free to contract around.” Gen. Protecht Grp., Inc. v. Leviton Mfg. Co., 651 F.3d 1355 , 1361–62 (Fed. Cir. 2011). The defendants think that the doctrine fits here. They note that when Gunnam joined LSI, he agreed to assign LSI everything he invented while working for them. D.I. 122-1, at 255. When Gunnam worked for LSI, they claim, he improved his Ph.D. work. Per his employment contract, he assigned those improvements to the defendants. But to practice the improvements, the defendants would also need to practice his underlying Ph.D. work. So under legal estoppel, the defendants argue, Gunnam’s assignment of his improvements implicitly included a license to his Ph.D. work. Thus, he may not assert the Texas A&M patents against them. And since TexasLDPC (they claim) is his agent, the company may not assert the patents either. D.I. 121, at 12– 19. Even if TexasLDPC is Gunnam’s agent, this argument flops. Gunnam’s contract with LSI expressly excluded his Ph.D. work: it distinguished his “prior inventions” (which were excluded) from future “improvements” to those inventions (which were not). D.I. 122-1, at 256; see also D.I. 121, at 13. The defendants respond that this exclusion is only a limit on what Gunnam assigned, not a limit on what he licensed. Not so. The exclusion applies to the whole “Agreement”—the only document that the defendants claim created an implied license. D.I. 122-1, at 256. That document could not have been clearer: if the defendants wanted to use Gunnam’s Ph.D. work, they needed to pay separately. Cf. AMP Inc. v. United States, 389 F.2d 448 , 454–55 (Ct. Cl. 1968) (finding legal estoppel when an exclusion to a license did not cover a prior patent). In sum, Gunnam never gave the defendants an implied license to his Ph.D. work; indeed, he expressly refused them one. So there was no legal estoppel. I will deny the motion for summary judgment. IV. THE DEFENDANTS MAY RAISE LITIGATION MISCONDUCT LATER Although the defendants asked for threshold discovery on litigation misconduct, they never filed a motion. According to TexasLDPC, the defendants have thus waived their misconduct argument. I disagree. The defendants say that they do not yet have the evidence to make their best case. D.I. 140, at 2. Given that, I commend them for not moving yet and will not punish them for waiting. So I will deny TexasLDPC’s motion. Still, this case has been delayed long enough by the defendants’ request for threshold discovery. I will not let them delay it any longer. If they decide to move for relief based on litigation misconduct, they now must wait until after discovery. And if they do move then, their motion will not be well taken unless it relies at least partly on evidence that is unavailable now. I will thus deny this motion too.