Lau on FRAND in Singapore and China

An article I only recently came across is Joseph Lau's FRAND Defences Against the Grant of Injunctive Relief:  Applying Huawei v. ZTE in Singapore and China, 16 Asian J. Comp. L. 33 (2021).  Here is  a link to the article, and here is the abstract:

From the size of A4 paper to 5G in the telecommunications sector, standards are ubiquitous. Standard essential patents (SEPs), which protect technology essential to standards, enable their proprietors to gain significant market power. Antitrust authorities therefore scrutinize the exercise of SEPs for breaches of competition law. In this regard, the ability of SEP proprietors to obtain injunctions against implementers as a remedy for infringement of SEPs where licensing negotiations have broken down or are ‘ongoing’ has proven controversial. Some fear that this enables SEP proprietors to threaten injunctions unless implementers agree to unfair, unreasonable, or discriminatory terms. In Huawei Technologies Co Ltd v ZTE Corp [2015] ECLI:EU:C:2015:477, the Court of Justice of the European Union identified circumstances where a SEP proprietor’s application for injunctive relief as a remedy for infringement of its SEP constitutes an abuse of a dominant position, with the classification of the SEP proprietor’s application as being abusive forming a ‘FRAND Defence’ which implementers may invoke against the grant of the injunction requested. This article analyzes whether this approach can be replicated by the Singapore Courts and whether the Chinese Courts, which have already dealt with SEP licensing disputes, adopt a similar approach.

I thought the article's discussion of how courts in Singapore might address FRAND cases was quite interesting.  The author argues that the most likely doctrinal tool would not be competition law, but rather the equitable nature of injunctive relief, Singapore being a common-law jurisdiction.  The article also includes a good discussion of Chinese case law and Chinese court guidelines as of 2020.

 

Two Observations about the Delhi High Court Decision in Lava v. Ericsson

I now have had the opportunity to read the relevant portions of the Delhi High Court’s recent Lava v. Ericsson decision, which (as previously noted by Florian Mueller) ordered Lava to pay Ericsson a global FRAND royalty, covering a portfolio of FRAND-committed SEPs for the period 2011-20, plus costs, totaling approximately USD $30 million (specifically, "1.05% of the net selling price of devices sold by Lava" from November 1, 2011 through May 8, 2020, amounting to Rs. 244,07,63,990, with postjudgment interest at 5% and costs).  In summary, the court considered Lava to be an unwilling licensee, because of its holdout behavior; relied on Ericsson’s proposed comparable licenses to come up with a FRAND rate; rejected Lava’s proposed comparable, and also its proposed top-down approach for lack of evidence on (among other things) the aggregate royalty burden; found no evidence that Ericsson had engaged in holdup or royalty stacking; rejected Lava’s argument that the appropriate royalty base would be the SSPPU, rather than the end product (smartphones); and also concluded that a FRAND license would be global in scope, even though Lava primarily sells phones in India.  All of this may be well-grounded in the law and the evidence (there are a lot of redactions concerning the comparables), but there are two things that, at least initially, I’m finding a bit confusing.

The first is the doctrinal basis for granting a global portfolio license.  Lava initiated the litigation, asking the court to declare inter alia that Ericsson “is bound to grant an irrevocable license under its standard essential patents, including patents which are essential and/or claimed to be essential by the Defendant to 2G and/or 3G standards, on fair, reasonable and non-discriminatory (FRAND) terms, to the Plaintiff herein,” and to “[d]eclare the fair, reasonable and non-discriminatory (FRAND) terms, including royalty rates, on which the Defendant should grant a license under its Indian patents and patent applications which are enforceable and essential to 2G and/or 3G standards, to the Plaintiff herein” (p.15).  In turn, Ericsson sued Lava for the infringement of eight specific (assertedly standard-essential) patents.  If I am understanding this correctly, though, by the time the case was litigated on the merits, Lava no longer wanted the above declaration, and instead proposed that the court limit its attention to the eight patents in suit.  The court nevertheless set the terms of a global portfolio license, stating in paragraph 635 that “[t]he issue the Court is adjudicating . . .  is whether Ericsson can obtain a declaration that the royalty rates offered by them to Lava in respect of their portfolio of SEPs, are indeed FRAND.”  So maybe the court understood Ericsson to be seeking a declaratory judgment of some sort permitted under Indian law.  (At p.16, the court states that Ericsson sought inter alia a declaration "that the rates offered by [Ericsson] qua its portfolio of Standard Essential Patents are FRAND in nature.")  But the resolution of the case seems to me somewhat different from the way these types of cases have been litigated elsewhere.  In the U.K., for example, a decision that the FRAND-committed SEPs in suit are valid and infringed may result in the court offering the defendant a choice between being enjoined from practicing those patents in the U.K. or accepting the terms of a court-determined global license (the so-called FRAND injunction).  But here, by 2020 there was no threat of injunction, because all of the patents in suit had expired.  Another possibility in some countries might be for the defendant to allege that it is a third-party beneficiary of the patentee’s FRAND commitment, and to sue for breach of contract and/or a declaratory judgment; the end result may be for the court to declare what the terms of a FRAND license would be, as in Microsoft v. Motorola.  Or maybe the patentee can seek a declaration that its offer is FRAND (and therefore that the patentee is not in breach of its obligations), as indeed Ericsson was doing here.  What I’m not sure I’ve seen before, however, is a court adjudicating a patent infringement suit involving a discrete number of SEPs and awarding, as past damages for the infringement, a FRAND royalty covering the entire portfolio, including patents not in suit.  Of course, in this case, if the defendant doesn’t sell any products outside of India, it isn’t paying a royalty on any non-Indian patents, even if the non-Indian patents affect the global royalty rate Ericsson charges; but it's not clear to me that the patents in suit are the only Indian SEPs Ericsson owns (see para. 725, in which Lava asserts that Ericsson has "about 30 Indian patents," though not specifying whether any of the other 22 or so are SEPs).  This seems unusual to me; again, if I’m understanding correctly, this case was litigated as a patent infringement suit, and the court is awarding damages for patent infringement, not merely declaring what the terms of a global license would be.  If, however, Ericsson has only eight Indian SEPs (one of which the court found to be invalid, thus reducing the number to seven), then maybe the global rate (based on Ericsson's global portfolio) multiplied by the number of products Lava sold in India is the correct measure of damages for the infringement of the seven valid Indian SEPs.  But this is not entirely clear to me from the decision, though it's possible I'm overlooking something.  (I need to check to see if any of the Chinese FRAND cases have followed this fact pattern; off the top of my head, I'm not sure.)  Or maybe it's simpler than I am making it out to be; perhaps the court is saying that to determine the damages for the infringement of the seven patents in suit, you need to take into account that the patentee would have conditioned the use of those seven patents on the defendant's agreement to a global license.  In other words, but for the infringement, the defendant would have agreed to a global license, so the damages needed to compensate the patentee for the infringement of the seven patents equal the defendant's turnover multiplied by the global rate.  I'm not sure that argument would fly under U.S. patent damages law--it sounds almost like a convoyed goods argument, and the U.S. approach to damages for convoyed goods is somewhat more restrictive than in some countries (see my recent post here)--but maybe it works in India, and that might be the correct interpretation of paras. 680 et seq. of the decision.  

The other thing that seems odd to me is the statute of limitations issue.  Apparently the general statute of limitations under Indian law is three years, but the court concluded that certain provisions of the Patent Act trumped this general provision: §§ 11A(7) and 45(3), which allow the patentee to recover damages covering the period of time beginning from the date of publication of the patent application, as long as the patent subsequently issues (a common provision in many countries’ patents laws), and § 111, which states that the patentee can sue for damages once the defendant knows or has reason to know of the existence of the patent (a common provision in some Commonwealth countries).  The court reads these provisions as allowing Ericsson to sue for damages beginning in 2011, which is when it put Lava on notice, even though Ericsson did not file suit for infringement until 2015.  If I am understanding correctly, the logic of the decision would seem to eliminate the statute of limitations altogether for patent infringement suits, by allowing the patentee to sue for damages at any time once it puts the defendant on notice (though perhaps subject to the equitable doctrine of laches?).  Maybe that is correct as a matter of Indian law—it’s not a matter on which I claim any expertise—but it strikes me as a somewhat unusual practice. 

Van Dongen on Proportionality and Injunctions in the UPC

Lisa Van Dongen has posted a paper on ssrn titled Proportionality and Flexibilities in Final Injunctive Relief, forthcoming in The Unified Patent Court: Problems, Possible Improvements and Alternatives (Alain Strowel et al. eds., Ledizioni, 2023).  Here is a link to the paper, and here is the abstract:

          In 2006, the patent world was shaken to the core by eBay v MercExchange, a case that questioned several basic principles in patent enforcement that were considered well established. The US Supreme Court sent a clear signal that patent rights were not to be considered absolute, and courts should thus not enforce them in automated fashion with injunctive relief. This case has received considerable attention globally, with many patent scholars analysing it in meticulous detail and questioning the European approach. Even though there is no agreement in the field on the optimal balance in patent enforcement (and likely never will be), even the most adamant proponents of strong patent enforcement agree that there may be other interests that merit the denial or tailoring of final injunctive relief. Yet, the automated tendencies in patent enforcement in Europe - the finding of an infringement automatically leading to the (blanket) grant of a permanent injunction - remain not only as prevalent as they have been for several decades, but also without any clear departures by courts (apart from English judges) from such tendencies indicative of course changes. What is more, is that the possibility for Europe to break with automated tendencies in enforcement will soon be further complicated by the addition of another layer to Europe’s existing patent systems, namely by the creation of the Unified Patent Court (UPC) and the unitary patent. If this system takes off, decisions of this new court will carry significant weight in European patent enforcement due to several organisational and territorial aspects. The UPC has even been described as a potential judicial counterbalance to pro-patent tendencies in patent offices, particularly the European Patent Office (EPO). However, considering the strange construct of its creation, it is questionable that the UPC will be that judicial counterbalance and lead the way for other courts in Europe. Some of these aspects might also create some tension with other systems it will have to co-exist and interact with. A closer look is thus imperative. This paper aims to do just that, testing the hypothesis: The UPC will not bring about a change in the current automated tendencies in granting final injunctions, but rather cement them. This paper explains why there will be no push from the EU to try and do so based on the current status of EU harmonisation in patent enforcement, questions the UPC’s capability and willingness to break with the existing automated tendencies based on the UPCA’s formulations and organisational features of the UPC, and explores some legislative solutions at the European level to move Europe away from automated tendencies in patent enforcement.

This is an insightful paper, and I suspect that the author is correct in predicting that the UPC will not depart from the status quo in favor of near-automatic granting of injunctive relief to the prevailing patent owner.  Her recommendation that the EU consider legislation providing more detail on when and how proportionality might result in limitations on (tailoring) injunctive relief, or in some instances denying such relief altogether, seems to have merit.

New Indian SEP Case

Multiple sources are reporting on a March 28, 2024 judgment of the Delhi High Court, awarding Ericsson the equivalent of approximately USD$30 million in a SEP dispute against a firm called Lava.  I have not yet succeeded in obtaining a copy of the judgment, which is said to be quite long, but recommend that readers take a look at Florian Mueller's discussion of it on his new site, ipfray.  I probably will have some comments on the decision after I obtain a copy, read it, and digest it, which might however take a while.

Alicante IP Colloquia

For anyone finding themselves in southeastern Spain on April 19, Alicante IP Colloquia is putting on an in-person event at noon titled «Hacia la cuantificación del daño» ("Toward the quantification of damages").  Panelists will be Judges Florencio Molina López and Gustavo Andrés Martin Martín.  The event is free, but registration is mandatory (link available here).  Judge Molina López is the author of a recent book on protective letters (see here).  I'm guessing the panelists will address the recent decision of the Madrid Court of Appeals in Eli Lilly v. Teva, previously noted on this blog here and here.

Federal Circuit Affirms Denial of Preliminary Injunction

The case is Biomedical Device Consultants & Laboratories of Colorado, LLC v. ViVitro Labs, Inc., nonprecedential opinion published last Thursday and authored by Judge Lourie, joined by Judges Dyk and Stark.  Plaintiff BDC, which manufactures and sells heart valve durability testing devices, sued its competitor ViVitro for infringement of U.S. Patent 9,237,935, which is “directed toward accelerated rate fatigue testing devices for prosthetic valves” (p.2). The district court denied BDC’s motion for a preliminary injunction, “finding that it failed to establish a likelihood of success on the merits for two independent reasons,” namely substantial questions as to both infringement and validity.  On appeal, the Federal Circuit affirms on the basis that there is a “substantial question of validity,” and does not address the infringement issue (though toward the end of the opinion it "caution[s] that claim terms are generally not limited to the preferred embodiments"(p.14), which might be interpreted as expressing some possible discomfort with the district court's claim construction and hence infringement analysis).  Without rehashing the evidence before the court, I would note only that the opinion does not appear to break any new ground on the legal questions.  It cites Titan Tire Corp. v. Case New Holland, Inc., 566 F.3d 1372 (Fed. Cir. 2009), which is probably the leading opinion addressing the parties’ burdens when a defendant responding to a motion for preliminary injunction challenges validity, only in passing, but the analysis seems consistent with that case.  In particular, the court states that

At the preliminary injunction stage, a defendant may raise a substantial question of validity “on evidence that would not suffice to support a judgment of invalidity at trial.” Amazon.com, 239 F.3d at 1358. The question here is one of “vulnerability,” which “requires less proof than the clear and convincing showing necessary to establish invalidity itself.” Id. Furthermore, the district court’s assessment of prior art references is an issue of fact reviewed for clear error. . . (p.8).

Proceeding from these premises, the court concludes that “the district court did not make a clear error in its assessment of the prior art,” and “therefore did not abuse its discretion in finding that BDC failed to demonstrate that ViVitro’s anticipation defense lacked substantial merit.  See Titan Tire Corp. v. Case New Holland, Inc., 566 F.3d 1372, 1377 (Fed. Cir. 2009) (“[I]t is the patentee, the movant, who must persuade the court that, despite the challenge presented to validity, the patentee nevertheless is likely to succeed at trial on the validity issue.”)” (pp. 10-11).  The court also rejects BDC’s arguments that the district court did not address four of the eight claims in suit, finding that the district court did indeed conclude that there was a substantial question of obviousness as to those claims, and did not abuse its discretion in so concluding.

Federal Circuit: Damages for Extraterritorial Injuries Caused by Domestic Infringement Are in Principle Recoverable

The decision is Brumfield v. IBG, LLC, opinion by Judge Taranto, joined by Judges Prost and Hughes.  The case involved four patents, relating to “improved graphical user interfaces for commodity trading and methods for placing trade orders using those interfaces” (p.3), owned by Trading Technologies International, Inc. (TT).  (They are now owned by a trust of which Mr. Brumfield, the appellant named in the caption, is the sole trustee.  The court refers to “TT” as the entity in interest throughout.)  TT alleged that IBG infringed various claims of these patents, and that “the instrument of the alleged infringement was the BookTrader module (trading tool) that is part of IBG’s Trade Workstation Platform (TWS), software that traders load onto their computers and use for buying and selling on exchanges, such as commodities exchanges” (p.11).  The district court found the asserted claim of two of the patents in suit invalid for lack of patentable subject matter, however, a conclusion that the Federal Circuit now affirms (and that I will omit for present purposes).  The matter went to trial on five method claims and one computer readable medium (CRM) claim of the ’304 Patent, and three method and two CRM claims of the ’132 Patent.  The jury found these claims infringed and not invalid, and awarded $6,610,985 in damages.  TT nevertheless challenges the judgment, on the ground that the district court improperly excluded TT’s expert’s damages theory as it related to foreign activities (in which case the damages awarded presumably would be undercompensatory.)  In particular, the district court, uncertain whether the Supreme Court’s decision in WesternGeco LLC v. ION Geophysical Corp., 485 U.S. 407 (2018), overrules the Federal Circuit’s decision in Power Integrations, Inc. v. Fairchild Semiconductor International, Inc., 711 F.3d 1148 (Fed. Cir. 2016), had excluded testimony that apparently would have at least partly based the reasonable royalty due for the domestic infringement of the claims in suit on the value the patentee would have expected to derive from foreign activity that was enabled by the domestic infringement.  WesternGeco specifically held that a U.S. patent owner can recover damages, where the defendant violates Patent Act § 271(f) by exporting components from the United States to be combined abroad, and the resulting extraterritorial combination deprives the patent owner of the profit it would have earned on extraterritorial sales it would have made but for the infringement; it did not expressly address the question of whether the analysis would the same or different when the act of domestic infringement is the (much more common) violation of § 271(a) (making, using, selling in the United States), although I and (some) others have argued that the reasoning of the opinion would apply in this context as well.

In what may prove to be a landmark opinion, the Federal Circuit concludes that WesternGeco overrules Power Integrations , and that a reasonable royalty  awarded for domestic infringement that enables further activity abroad may under some circumstances reflect the value to the infringer of that subsequent foreign activity.  So the district court was wrong on this issue; but unfortunately for TT, the court also concludes that because the expert’s model was premised on domestic manufacture, and the plaintiff hadn’t proven that any infringing domestic manufacture by the defendant actually caused the defendant to gain any foreign benefit, the expert’s testimony on this matter was correctly excluded.

In concluding, as it does, that “[f]or a determination whether patent damages are properly awarded in a particular case based partly on conduct abroad, the decision in WesternGeco established a framework of analysis that necessarily supersedes the analysis set forth in our earlier decision Power Integrations” (p.25), the court’s analysis is consistent with what I argued in my article Extraterritorial Damages in Patent Law, 39 Cardozo Arts & Enter. L.J. 1 (2021).  In particular:

            The first doctrinal issue before us is whether the WesternGeco framework applies when the direct infringement in question (either itself or as a component of indirect infringement) is one of the acts at issue here accused of infringing under § 271(a). We readily conclude that it does.

 

           Nothing about the WesternGeco analysis of § 284, the damages provision, or about § 281, the cause-of-action provision, is altered when “the infringement” at issue is infringement under § 271(a) rather than § 271(f). Under WesternGeco we must examine the particular acts alleged to constitute infringement under particular statutory provisions to determine if the allegations focus on domestic conduct. Section 271(a) provides that “whoever without authority makes, uses, offers to sell, or sells any patented invention, within the United States or imports into the United States any patented invention during the term of the patent therefor, infringes the patent.” 35 U.S.C. § 271(a) (emphases added). At least the making, using, offering to sell, and selling provisions are expressly limited to domestic acts. . . .

 

            If the exporting covered by § 271(f)(2) is a domestic act for purposes of the extraterritoriality analysis, as WesternGeco held, so too are the § 271(a)-covered acts at issue in this case. The WesternGeco extraterritoriality framework for damages under § 284 therefore applies to the infringement under § 271(a) here. . . .

 

           We also conclude that the WesternGeco framework applies to a reasonable-royalty award, not just a lost-profits award, under § 284, though its application must reflect the established differences in standards for the two types of awards. . . .

 

            This case involves a reasonable royalty, and repeatedly articulated standards frame how the particular issue presented here is properly formulated. . . .

 

            Those principles point to a minimum requirement for a patentee seeking reasonable-royalty damages based on foreign conduct that is not independently infringing. Under the foregoing principles, the hypothetical negotiation must turn on the amount the hypothetical infringer would agree to pay to be permitted to engage in the domestic acts constituting “the infringement.” 35 U.S.C. § 284. If the patentee seeks to increase that amount by pointing to foreign conduct that is not itself infringing, the patentee must, at the least, show why that foreign conduct increases the value of the domestic infringement itself—because, e.g., the domestic infringement enables and is needed to enable otherwise-unavailable profits from conduct abroad—while respecting the apportionment limit that excludes values beyond that of practicing the patent. This kind of causal connection, framed in terms of the agreement-to-pay aspect of a hypothetical negotiation, is a necessary beginning—we need not here say it is sufficient—for a foreign-conduct analysis in a reasonable-royalty case. Cf. Carnegie Mellon, 807 F.3d at 1307 (noting that defendant’s sales abroad were “strongly enough tied to its domestic infringement as a causation matter to have been part of the hypothetical-negotiation agreement,” before moving on to apply extraterritoriality standards based on Power Integrations, now superseded by WesternGeco) (pp. 33-37).

The court goes on to note, however, that there has to be proximate cause in addition to cause-in-fact, and that the proximate cause analysis as it relates to reasonable royalties may be tricky:

          The foregoing authorities [various Supreme Court decisions] raise questions about the proper approach to determining, based on “other doctrines, such as proximate cause,” WesternGeco, 585 U.S. at 417 n.3, when foreign conduct can properly play a role in calculating patent damages. One such question is whether the “reasonable, objective foreseeability” presumptive standard for lost profits, Rite-Hite, 56 F.3d at 1546, is applicable where the damages are for a (non-established) reasonable royalty, whose conceptual foundation is notably different from that of lost profits. Another question concerns the long-recognized general avoidance of extraterritorial reach that is an aspect of the statutory context. . . . What, if any, room is there to take that consideration into account in applying the proximate-cause requirement, itself not addressed in WesternGeco, without contradicting the Supreme Court’s ruling in WesternGeco? We need not and do not here suggest answers to, or further explore, those or other questions (pp. 39-40).

For my own ruminations on proximate cause and royalties in this context, see my article above at pp. 39-42, 51-52.

The reason the court doesn’t need to explore those issues for now is its factual conclusion that the expert had not shown “the needed causal relationship to the foreign conduct for which recovery is sought” (p.40).  The asserted infringement on which the expert relied was “Making the Accused Product,” but that (the court says) cannot refer to the method claims in suit, because you don’t “make” a method (p.41).  The expert’s analysis therefore would have had to refer to the CRM claims, but the expert did not focus on the defendant’s making of “an individual memory-device unit,” but rather its TWS BookTrader software—“software in the abstract”—which was not itself claimed in the patents in suit (pp. 41-42).  The expert’s analysis therefore didn’t start from an act of domestic infringement, i.e., “making a claimed CRM (or method)” (p.42).  Instead—if I am understanding this correctly—the expert sought to include the foreign use of copies of the TWS BookTrader software in the royalty base (see p.16), but without tying that use to a predicate act of infringing domestic manufacture.  At least, I think that’s the gist of it.  The court further observes:

          We may assume (without deciding) that IBG had to make early CRMs domestically (or practice the claimed method) as part of its process of developing its software and that the value of such development work to IBG might reflect prospective foreign-earned revenue for the resulting product. Cf. Carnegie Mellon, 807 F.3d at 1294, 1297, 1307 (referring to payment for domestic infringement that is part of development work that, when completed, would produce large foreign revenues). In this case, however, according to TT and Ms. Lawton, IBG’s development of its BookTrader product meeting all claim limitations occurred before TT’s patents issued: TT accused IBG of marketing its BookTrader product before July 20, 2004, which caused infringement to begin precisely when the ’304 patent issued. On that premise, IBG’s making of CRMs in the initial creation of a BookTrader product meeting all claim limitations was not infringing under § 271(a), and IBG therefore did not need to pay TT anything for that work, which could not properly be included in the calculation in the hypothetical negotiation held “just before” July 20, 2004. . . .   

 

         Later domestic making of BookTrader-containing CRMs (or practicing of the claimed methods) could be infringing, of course, and properly be subject to a royalty. But TT was permitted to introduce evidence that some foreign users of BookTrader obtained their copies from domestic acts of making a copy or selling. FRE 702 Opinion, 2021WL 5038754, at *2. The only disallowed proposal therefore had to involve making copies abroad for foreign users (and foreign sales).

 

           On TT’s and Ms. Lawton’s premise that pre-July 20, 2004 versions of TWS BookTrader met the limitations of the ’304 and ’132 patents’ claims, TT has not offered a concrete, coherent account of why, in the hypothetical negotiation, the royalty for new domestic acts of making claimed CRMs (or practicing claimed methods), starting July 20, 2004, would have properly been increased to reflect the prospective making and sale of CRMs abroad for use abroad. On the noted premise, IBG, even before the patents issued, already had CRMs containing TWS BookTrader that met the patents’ limitations. “[N]either export from the United States nor use in a foreign country of a product covered by a United States patent constitutes infringement.” Johns Hopkins University v. CellPro, Inc., 152 F.3d 1342, 1366 (Fed. Cir. 1998). And TT has not argued that the making of CRMs abroad would be infringing, even if the software installed abroad came from the United States, either under § 271(a), see Centillion Data Systems, LLC v. Qwest Communications International, Inc., 631 F.3d 1279, 1288 (Fed. Cir. 2011); Deepsouth, 406 U.S. at 527, or under § 271(f), see Microsoft, 550 U.S. at 449–50 (software itself is not a “component” under § 271(f)). . . .

 

          IBG might of course infringe by domestically making new CRMs containing upgraded versions of TWS BookTrader. But TT has not shown how value added by the upgrades would be properly added to the royalty in light of the apportionment requirement to avoid charging for value not attributable to the claimed invention. In particular, TT has not explained how such upgrade value would be anything but the value of features beyond what is required by the patent claims . . . . (pp. 43-45).