The Patent Trial and Appeal Board (PTAB) has knocked out all asserted claims from one of the two VLSI Technology LLC patents that Intel was found to infringe just over two years ago, accounting for $657M of a $2.2B damages award. That final written decision came in an inter partes review (IPR) that has recently been at the center of a heated debate over the proper punishment for gamesmanship and misconduct in IPR proceedings: Starting late last year, USPTO Director Kathi Vidal issued an evolving set of sanctions rulings against the original petitioner, OpenSky Industries, over a series of unusual financial offers designed to extract payments from both VLSI and joined copetitioner Intel. After first demoting OpenSky to a silent understudy role and then dismissing it altogether, Vidal reinstated it to the proceeding in February, taking similar steps against another petitioner facing similar accusations (Patent Quality Assurance, or PQA). The invalidity decision follows OpenSky’s unsuccessful attempt last month to get Intel dismissed from the IPR on estoppel grounds.
VLSI’s litigation against Intel began in the Northern District of California in October 2017, with the dispute subsequently expanding to include cases in the District of Delaware and the Western District of Texas. Three of the plaintiff’s West Texas cases have gone to trial: the first ending in the aforementioned $2.2B verdict in March 2021 as to two patents (7,523,373 and 7,725,759); the second ending in Intel’s favor, turning away a $3B ask from VLSI for two more patents (6,366,522 and 6,633,187) the following month; and the third going for VLSI, with a jury returning a $949M verdict as to a fifth patent (7,606,983) in November 2022. OpenSky and PQA have both stated that they were formed as a response to the first verdict, arguing that by ensuring that a patent involved in such a large verdict would undergo a validity review (prior Intel IPRs against the ‘373 and ‘759 patents having been denied institution), their IPR petitions would protect the “integrity of the patent system”.
Yet in spring 2022, after the PTAB had already instituted trial in both actions—an OpenSky IPR against the ‘759 patent (IPR2021-01064) and a PQA IPR against the ‘373 patent (IPR2021-01229)—allegations of unscrupulous behavior came to light: VLSI disclosed that OpenSky had offered to sabotage its own IPRs in exchange for payment, prompting widespread outcry and leading Vidal to initiate director review in those proceedings. It was soon revealed that OpenSky had also sought payments, and even an acquisition offer, from Intel. Vidal issued a sweeping sanctions order against OpenSky in October as a result, finding that it abused the IPR process through this behavior and by failing to comply with discovery ordered during the director review. However, Vidal stopped short of dismissing OpenSky altogether, barring it from actively participating and making Intel the lead petitioner for remand proceedings. December saw Vidal double down, dismissing OpenSky entirely and doing the same to PQA for similar behavior and discovery misconduct, and for making additional misrepresentations. However, on January 27 and February 3, Vidal restored both PQA and OpenSky as petitioners and ordered OpenSky to pay attorney fees and costs to VLSI.
On April 4, the PTAB then denied VLSI’s motion to terminate Intel, which as noted earlier had been designated lead petitioner. The patent owner argued that because the jury in the parallel district court case found that Intel had not shown the tried claims of the ‘739 patent (claims 14, 17, 18, and 24) to be invalid, the resulting final judgment of no invalidity barred Intel from challenging the claims due to claim preclusion—contending that the IPR was a collateral attack on the infringement verdict. Yet the PTAB agreed with Intel that because the statute establishing IPR’s predecessor, inter partes reexamination, provided for two-way estoppel (where district court judgments may estop parties from raising certain arguments in the reexam and vice versa), whereas the America Invents Act replaced this with a provision explicitly providing one-way estoppel (barring petitioners from asserting, in district court, arguments they “raised or reasonably could have raised” before the PTAB, but with no estoppel occurring in the other direction), Congress had shown its “intent that claim preclusion not apply in the circumstances here”.
On April 20, the Board adjusted the final written decision deadline for the -1064 IPR to May 15, as provided under 37 C.F.R. § 42.100(c) (allowing such an adjustment when a case involves joinder). The panel in that IPR then issued its final written decision three days ahead of schedule, on May 12. In that opinion, the panel found that OpenSky and Intel (collectively treated as the petitioner) had shown all challenged claims—i.e., claims 1, 14, 17, 18, 21, 22, and 24, or all claims tried in West Texas plus claims 1, 21, and 22—to be unpatentable. The panel determined that the petitioner had shown those claims to be invalid on all asserted bases: claims 1, 14, and 17 were invalid as obvious in light of the “Shaffer” (the 6,298,448 patent) and “Lint” (the 7,360,103 patent) references and in light of the “Chen” (the 2003/0159080 application) and “Terrell” (5,838,995) references. Claims 18, 21, 22, and 24 were held to be obvious in light of those same combinations of prior art references, both with the addition of one further reference, “Kiriake” (the 2004/0098631 application).
The PTAB has not yet issued a final decision in PQA’s IPR against the ‘373 patent, in which trial has been instituted against all 16 claims.
VLSI’s Disclosure Debacle
Another issue throughout much of the VLSI litigation has been Intel’s license defense, which the defendant has raised, or sought to raise, in California, Delaware, and Texas federal courts as well as in Delaware’s Court of Chancery. That defense stems from a 2012 settlement agreement between Intel, Finjan, Inc., and Finjan Software, Inc., which according to Intel, gave the company a “license to the patents of Finjan and its affiliates, current or future, through a capture period of November 20, 2022”. Intel contends that upon the completion of Fortress Investment Group LLC’s acquisition of Finjan in 2020, those Finjan entities became affiliates of VLSI—which, per court filings, is a subsidiary of CF VLSI Holdings LLC, a Delaware limited liability company that is “owned by ten separate entities, which in turn are owned by pension funds and third-party investors”, those pension funds “managed” by Fortress Investment Group.
While Judge Albright dispensed with that defense as “futile” in the West Texas litigation, Delaware’s Chief Judge Colm F. Connolly came to the opposite conclusion as to that same issue. As 2022 neared its end, it appeared likely that VLSI would have to reveal information relevant to Intel’s defense—in particular, details on the individual funds that comprise VLSI’s funding for its litigation—as a result of the heightened disclosure requirements that Judge Connolly imposed in his court through a pair of April 2022 standing orders, one requiring litigants to disclose details related to any nonrecourse funding arrangements with third parties and a second requiring the disclosure of a wide range of information on corporate control. In July, Judge Connolly directed VLSI and Intel to certify that they had complied with the orders, leading to a back-and-forth in which he stated that he was “troubled” by VLSI’s “clearly inadequate” disclosures—leading the court to unseal a declaration by the plaintiff’s CEO and stay the case until VLSI complied fully.
As that dispute dragged on, Judge Connolly set a hearing for December to address those concerns, similar to those in which he has grilled the individuals behind a variety of other familiar parties—including plaintiffs associated with IP Edge LLC or Dynamic Deals IP LLC (d/b/a DynaIP)—over their failure to comply with his standing orders. However, that hearing was subsequently continued into January, and by the end of December the parties informed the court that they had stipulated to end the Delaware litigation entirely.
This meant that Intel’s license defense remained viable only in the Northern District of California. In that court, Intel filed a motion to compel additional discovery (through both document requests and interrogatories) concerning VLSI’s investors, contending that it needed the discovery to mount its license defense, as well as to assist the court in evaluating possible conflicts and to challenge the credibility of VLSI’s witnesses.
In March, Magistrate Judge Nathaniel Cousins sounded an ominous note for Intel’s license defense in an order denying Intel’s motion. As noted in RPX’s prior coverage, the court in part dispensed with the license defense justification: “Intel asserts that a patent license it has with another entity covers ‘affiliates’ and that it wishes to explore if VLSI is an affiliate. The Court agrees with VLSI that this issue was already decided between the parties in VLSI v. Intel Corp., 2022 WL 1261322 (W.D. Tex. Apr. 21, 2022)”. The court referred to Judge Albright’s decision, covered in more detail here, in which that court ruled, among other things, that the “proposed amendment would be futile because VLSI is not a party to the agreement”.
On May 8, District Judge Beth Labson Freeman denied Intel’s motion for relief from that order. While her rationale is not yet clear, as the ruling remains conditionally sealed while the parties propose redactions due to the presence of confidential information, the decision could spell the end for the defendant’s license defense barring any subsequent appeal.
