Reset A A Font size: Print

United States Court of Appeals, Federal Circuit.



Decided: September 01, 2017

Before PROST, Chief Judge, NEWMAN, LOURIE, DYK, MOORE, O'MALLEY, REYNA, WALLACH, TARANTO, CHEN, HUGHES, and STOLL, Circuit Judges. MARK E. MILLER, O'Melveny & Myers LLP, San Fran-cisco, CA, filed a response to the petition for plaintiff-cross-appellant. Also represented by ANNE E. HUFFSMITH, LUANN LORAINE SIMMONS. E. JOSHUA ROSENKRANZ, Orrick, Herrington & Sut-cliffe LLP, New York, NY, filed a petition for rehearing en banc for defendants-appellants. Also represented by DANIEL A. RUBENS, ANDREW D. SILVERMAN; ROBERT M. LOEB, ERIC SHUMSKY, Washington, DC; VICKI L. FEEMAN, TRAVIS JENSEN, SCOTT T. LONARDO, Menlo Park, CA; WILLIAM H. WRIGHT, Los Angeles, CA; INDRA NEEL CHATTERJEE, Goodwin Procter LLP, Menlo Park, CA. ERIC E. BENSEN, Garden City, NY, as amicus curiae. PHILLIP R. MALONE, Stanford Law School, Stanford, CA, for amici curiae Public Knowledge and Jeremy W. Bock, et al. Also represented by JEFFREY THEODORE PEARLMAN.


Appellants EVE-USA, Inc., Synopsys Emulation and Verification S.A.S., and Synopsys, Inc. filed a petition for rehearing en banc. A response to the petition was invited by the court and filed by the cross-appellant Mentor Graphics Corporation. Two motions for leave to file amici curiae briefs were also filed and granted by the court.

The petition, response, and briefs of amici curiae were referred to the panel that heard the appeal, and thereaf-ter were referred to the circuit judges who are in regular active service. A poll was requested, taken, and failed.

Upon consideration thereof,


The petition for panel rehearing is denied.

The petition for rehearing en banc is denied.

The mandate of the court will be issued on September 8, 2017.


Peter R. Marksteiner Clerk of Court

Because the panel's decision is consistent with longstanding patent law damages principles,1 I concur in the denial of rehearing en banc.

We have consistently held that where an infringing product is a multi-component product with patented and unpatented components, apportionment is required. See VirnetX, Inc. v. Cisco Sys., Inc., 767 F.3d 1308, 1326 (Fed. Cir. 2014) (“No matter what the form of the royalty, a patentee must take care to seek only those damages attributable to the infringing features.”); Commonwealth Sci. & Indus. Research Org. v. Cisco Sys., Inc., 809 F.3d 1295, 1301 (Fed. Cir. 2015) (“[D]amages awarded for patent infringement ‘must reflect the value attributable to the infringing features of the product, and no more.’ ”) (quoting Ericsson, Inc. v. D–Link Sys., Inc., 773 F.3d 1201, 1226 (Fed. Cir. 2014)); Ericsson, 773 F.3d at 1226 (“[A]pportionment is required even for non-royalty forms of damages.”). The apportionment requirement dates back to Garretson v. Clark, 111 U.S. 120 (1884), where the Supreme Court held that “[t]he patentee ․ must in every case give evidence tending to separate or apportion the defendant's profits and the patentee's damages between the patented feature and the unpatented features.” Garretson, 111 U.S. at 121.

Garretson, however, also holds that damages for patent infringement may be based on the value of the entire infringing product if the patentee can show that “the entire value of the whole machine ․ is properly and legally attributable to the patented feature.” Id. In other words, “[i]f it can be shown that the patented feature drives the demand for an entire multi-component product, a patentee may be awarded damages as a percentage of revenues or profits attributable to the entire product.” LaserDynamics, Inc. v. Quanta Comput., Inc., 694 F.3d 51, 67 (Fed. Cir. 2012).

In this case, on the question of lost profits, the jury was instructed to consider the Panduit factors, including “demand for the patented product” (factor one) and an “absence of acceptable noninfringing substitutes” (factor two). See Panduit Corp. v. Stahlin Bros. Fibre Works, Inc., 575 F.2d 1152, 1156 (6th Cir. 1978). As the panel recognized, these two factors together “consider[ ] demand for the patented product as a whole” and “consider [ ] demand for particular limitations or features of the claimed invention.” Mentor Graphics Corp. v. EVE-USA, Inc., 851 F.3d 1275, 1285 (Fed. Cir. 2017) (citing DePuy Spine, Inc. v. Medtronic Sofamor Danek, Inc., 567 F.3d 1314, 1330–31 (Fed. Cir. 2009)). As the panel noted, the jury found that “Intel would not have purchased the [infringing] Synopsys emulator system without the two patented features and that there were no other alternatives available.” Id. at 1287. This undisputed fact finding established that Mentor proved that the patented features were what imbued the combined features that made up the emulator with marketable value. Under these cir-cumstances, further apportionment is unnecessary. See Ericsson, 773 F.3d at 1227 (“[W]here the entire value of a machine as a marketable article is ‘properly and legally attributable to the patented feature,’ the damages owed to the patentee may be calculated by reference to [the entire value of the machine.]” (quoting LaserDynamics, 694 F.3d at 67); VirnetX, 767 F.3d at 1326. Whether one views this in terms of what imbues value to the ultimate combination of features or what is a driver of demand for those combined features, the result is the same: the apportionment required by Garretson is satisfied.

In my view, the dissent mischaracterizes the panel's holding in this case, suggesting the panel held that in all cases where lost profits are awarded, apportionment is not required. Dissent Op. 1. To the contrary, the panel made clear that apportionment is typically necessary in both reasonable royalty and lost profits analyses. See Mentor Graphics, 851 F.3d at 1287–88. Under the narrow facts of this case, however, the panel determined that because the Panduit factors are satisfied, the damages award properly accounted for apportionment. I do not read the panel's decision to apply broadly to all lost profits analyses.

Accordingly, based on the jury's undisputed fact findings on the Panduit factors in this case, I agree with the panel that Mentor properly accounted for apportionment of lost profits between the patented and unpatented features of the infringing emulator system. For this reason, I concur in the denial of rehearing en banc.

I respectfully dissent from the court's decision to not rehear this case en banc. In my view, the panel decision here improperly holds that when lost profits are awarded for patent infringement, there is no requirement for apportionment between patented and unpatented features, contrary to longstanding Supreme Court authority.

For over a century, it has been established by both the decisions of the Supreme Court and this court that awards of lost profits or reasonable royalties for patent infringement must be apportioned between patented and unpatented features. See, e.g., Dowagiac Mfg. Co. v. Minn. Moline Plow Co., 235 U.S. 641, 646–48 (1915); Dobson v. Hartford Carpet Co., 114 U.S. 439, 443–44 (1885); Blake v. Robertson, 94 U.S. (4 Otto) 728, 733–34 (1876); Garretson v. Clark, 111 U.S. 120, 121 (1884); Seymour v. McCormick, 57 U.S. (16 How.) 480, 489–90 (1853); Ericsson, Inc. v. D-Link Sys., Inc., 773 F.3d 1201, 1226 (Fed. Cir. 2014). Indeed, the panel opinion acknowledges that “apportionment is ․ necessary in both reasonable royalty and lost profits analysis.” Mentor Graphics Corp. v. EVE-USA, Inc., 851 F.3d 1275. 1287 (Fed. Cir. 2017). At the same time, the case law is also clear that any award of lost profits is not appropriate unless the patentee estab-lishes that it would have sold the item but for the infringement. See, e.g., Grain Processing Corp. v. Am. Maize-Prods. Co., 185 F.3d 1341, 1349 (Fed. Cir. 1999). This but-for requirement is encapsulated in the first and second Panduit factors, which are “demand for the patented product” and the “absence of acceptable noninfringing substitutes,” i.e., alternatives that could have prevented the patentee from itself making the sale.1

The panel here holds that applying the first and second Panduit factors results in the required apportionment. See Mentor, 851 F.3d at 1285 (“[T]he absence of non-infringing alternatives ties lost profit damages to specific claim limitations and ensures that damages are commensurate with the value of the patented features.”). But calling the first and second Panduit factors apportionment “ignore[s] the ancient wisdom that calling a thing by a name does not make it so.” City of Madison, Joint Sch. Dist. No. 8 v. Wis. Emp. Relations Comm., 429 U.S. 167, 174 (1976). In my view, the panel opinion simply does not apportion—even though it purportedly recogniz-es apportionment's importance.

The panel cites cases from the Supreme Court and other circuits holding that but-for causation is necessary both for lost profits generally and for an award of lost profits damages in the patent area. See Mentor, 851 F.3d at 1283-84 (citing, inter alia, Aro Mfg. Co. v. Convertible Top Replacement Co., 377 U.S. 476 (1964) and Livesay Window Co. v. Livesay Indus., Inc., 251 F.2d 469 (5th Cir. 1958)). Cases outside the patent area shed little light on this issue, since contracts and tort cases involving lost profits generally do not require apportionment. Although the patent cases relied upon by the panel hold that but-for causation is required, none remotely suggests that but-for causation and apportionment are the same thing. And they are not.

Rather, the Supreme Court's patent cases make quite clear that more than but-for causation is required for apportionment. The claimed damages must be apportioned between patented and unpatented features. This principle was established by Supreme Court cases involving both the disgorgement of the defendant's profits (allowed before 1946) and the recovery of the patentee's own lost profits (the current rule). As the panel recogniz-es, both types of cases are pertinent because “the basic principle of apportionment which they espouse applies in all of patent damages.” Mentor, 851 F.3d at 1283 n.3.

In Seymour v. McCormick, the Court held that

one who invents some improvement ․ could not claim that the profits of the whole [invention] should be the measure of damages for the use of his improvement. ․ [Likewise,] [w]hen he has himself established the market value of his improvement, ․ he can have no claim ․ to make the profits of the whole machine the measure of his demand.

57 U.S. (16 How.) at 489–90. In Garretson v. Clark the Court further explained that “[w]hen a patent is ․ not for an entire[ ] ․ machine or contrivance, the patentee must ․ give evidence tending to separate or apportion the defendant's profits ․ between the patented feature and the unpatented features.” 111 U.S. at 121 (emphasis added) (internal quotation marks omitted). And in Dobson v. Hartford Carpet Co., the Court held that it was error to conclude “that the price per yard allowed as damages was the entire profit to the plaintiffs, per yard, in the manu-facture and sale of carpets of the patented designs, and not merely the value which the designs contributed to the carpets.” 114 U.S. at 443–44. Finally, in Blake v. Robertson, the Court held that

[t]he complainant made a profit of forty dollars ․ [on] the numerous machines he had sold. But in-ventions covered by other patents were embraced in those machines. It was not shown how much of the profit was due to those other patents ․ The complainant was, therefore, entitled only to nomi-nal damages. ․ It would have been error to give more.

94 U.S. (4 Otto) at 733–34.2

So too does our own case law require apportionment for lost profits recovery—particularly where, as here, the patented feature is only part of the infringing product. In Ericsson, we held that “apportionment is required even for non-royalty forms of damages,” including lost profits. 773 F.3d at 1226 (citing Garretson, 111 U.S. at 121).

The panel decision is therefore directly contrary to these cases, the logical foundation for which is readily apparent. Thus, for example, even if “but for” a patented feature the item would not have been purchased, it could be equally true that but for an unpatented feature (or a feature covered by another patent) the item would not have been purchased. Apportionment between features covered by the asserted patents and other features makes eminent sense. The panel makes no such apportionment. Nor do the Panduit factors.

Four other issues reinforce the appropriateness of en banc review.

First, EVE-USA properly raised the issue of apportionment in district court by seeking to present evidence of apportionment, which the district court excluded. See Appellant Br. 19–20; Appellee Br. 29–30 n.3. Specifically, the district court sustained an objection to exclude the appellant's damages expert from presenting “a slide about how other features of the sold products were important to customers ․ [because] it is really just an apportionment argument when it comes to lost profits.” J.A. 42,241. The district court held that it is “not appropriate ․ to say that 90 percent of the purchase price really had nothing to do with the patented feature. That's apportionment. You can't do that ․ on lost profits.” Id.

Second, in denying a post-trial motion for a new trial, the district court recognized that its failure to apportion was in error. The district court agreed that determining lost profits is “a two-step process,” that satisfying the but-for test of the Panduit factors is only step one, and that “there was an error in the trial on that score” to subse-quently not determine whether the entire value of the product is attributable to the patented feature. J.A. 42,600.3

Third, unlike the panel, the patentee did not suggest that applying the Panduit factors is equivalent to apportionment. Quite to the contrary, the patentee explicitly argued that “the value of the patent is considered as part of the ‘but for’ analysis, not in ․ apportionment,” Appellee Br. 33, that “lost profits should not be apportioned,” id. at 32, and that “apportioned lost profits can never adequately compensate a patentee for sales lost as a result of infringement,” id. at 30. In other words, the panel's effort to equate but-for causation and apportionment is a new theory not even adopted by the patent holder in this case.

Finally, apportionment is an important issue that will likely arise in every future lost profits case.

* * * * *

Respectfully, Judge Stoll's opinion (joined by various others) does nothing to rehabilitate the panel opinion. First, the theory that the panel found the application of the Panduit factors to be the same as the entire market value rule is not tenable. Judge Stoll Concurrence at 3–4. The combination of consumer demand and but-for causation (i.e., that customers would not have purchased the product without the two patented features) is not remotely the same as the entire market value rule. Consumer demand for the patented feature and but-for causation may exist (and satisfy the Panduit factors), but this does not mean that other features do not contribute to consumer demand. The entire market value rule only applies if consumer demand is driven by the patented feature. Consumer demand for the patented feature and but-for causation do not establish that consumer demand is only attributable to the patented feature. Consumer demand may also be driven by other features in the product, which may be just as necessary to purchasing decisions, leading to the required apportionment. The panel never says or even suggests that the Panduit factors and the entire market value rule are the same. And the accused infringer here never had the opportunity to address the entire market value rule on the facts of this particular case.

Second, Judge Stoll suggests that the dissent “mischaracterizes” the majority opinion by suggesting that it does not require apportionment when it does. To be sure, the panel says that it is requiring apportionment. But in fact it does not. Instead, it equates consumer demand and but-for causation with apportionment, contrary to the clear holding of the Supreme Court that the apportionment must be between patented and unpatented features, an apportionment that the panel here rejects. Since the factual findings necessary to satisfy the Panduit factors are a necessary predicate for lost profits, the result here is that true apportionment will never be required for lost profits.

I respectfully dissent from the court's decision to not rehear this case en banc.

Synopsys, Inc., Synopsys Emulation and Verification S.A.S., and EVE-USA, Inc. (“EVE”) (collectively, “Synopsys”)1 petition for rehearing, arguing our court should abolish assignor estoppel. We recognize that assignor estoppel may arise in multiple fact patterns, each of which would result in a unique balancing of the equities. We may be inclined to reconsider the breadth of the doctrine of assignor estoppel, but this case is not a proper vehicle to do so. Synopsys devoted approximately one page of its brief to this court to the issue of assignor estoppel where it argued nothing other than we should eliminate the doctrine because in Lear, Inc. v. Adkins, 395 U.S. 653 (1969), the Supreme Court “demolished the doctrinal underpinnings of assignor estoppel.” Synopsys Br. 42. Synopsys' petition for rehearing was no more detailed. In a little over two pages it argues that the doctrine of assignor estoppel should be abolished. Synopsys PFR Br. 13–15.

To be clear, Synopsys has made no arguments to this court regarding the scope of the assignor estoppel doctrine or its applicability to this particular case. For example, it makes no arguments in any briefing that applying the doctrine to this case would be improper because of rules of privity, the facts of this case, the nature of the employee/employer relationship, the nature of the compensation conveyed for the assignment, or that the assignment was not knowing or voluntary. It has asked this court to take the case en banc to answer a binary question: abolish or not. It has not argued that the doctrine is too broad as applied in this case. Because we do not believe we can or should eliminate the doctrine in its entirety, we decide not to rehear this case.

An amicus brief (the “Professors Brief”) argues we should rehear this case en banc because “there is no basis for expanding assignor estoppel outside the narrow cir-cumstances addressed in the Supreme Court's cases.” Professors Br. 7. The Professors Brief does not argue that we can or should abrogate assignor estoppel but rather that our court has expanded it in ways which undermine important public policy goals and interfere with employee mobility. Id. at 5–7; see also Mark Lemley, Rethinking Assignor Estoppel, 54 HOUS. L. REV. 513, 540–42 (2016) (recognizing the wisdom of a continued role for the doctrine). There may someday be a case which argues these points to this court (which Synopsys has not) and presents a record which would allow us to consider these issues in the context of an actual case, but this is not such a case.

The Professors Brief complains that the Federal Circuit has expanded privity notions beyond appropriate bounds,2 Professors Br. 5–7, and “has expanded the doctrine beyond cases where an inventor knowingly and voluntarily transfers a patent.” Id. at 6. The Professors Brief does not suggest, nor does the party briefing suggest, that any such expansion has occurred in this case. Synopsys made arguments related to privity before the district court, which it lost. And it chose not to appeal those issues. As already explained, there is no dispute presented to this court which causes us to consider whether the equities in this case warrant application of the assignor estoppel doctrine to Synopsys. We are not an academic tribunal. Our job is to decide cases based on their facts. And in this case, Synopsys did not argue applying assignor estoppel to the facts of this case is unjust, unfair, or in any manner inequitable.

Synopsys' argument that we should eliminate assignor estoppel is foreclosed by Westinghouse, where the Supreme Court held that an assignor could be estopped from challenging the validity of a patent in certain situa-tions. See Westinghouse Elec. & Mfg. Co. v. Formica Insulation Co., 266 U.S. 342, 353 (1924). We are bound by this precedent. Cf. Rodriguez de Quijas v. Shearson/Am. Express, Inc., 490 U.S. 477, 484 (1989) (“If a precedent of this Court has direct application in a case, yet appears to rest on reasons rejected in some other line of decisions, the Court of Appeals should follow the case which directly controls, leaving to this Court the prerogative of overrul-ing its own decisions.”).

In particular, the Supreme Court analogized assignor estoppel to estoppel by deed in Westinghouse. 266 U.S. at 349–50; see also Gottfried v. Miller, 104 U.S. 21 (1881). Estoppel by deed prevents the grantor from later denying the truth of the deed. The Supreme Court applied this principle from the law of real property to intellectual property in Westinghouse:

The analogy between estoppel in conveyances of land and estoppel in assignments of a patent right is clear. If one lawfully conveys to another a patented right to exclude the public from the making, using and vending of an invention, fair dealing should prevent him from derogating from the title he has assigned, just as it estops a grantor of a deed of land from impeaching the effect of his solemn act as against his grantee. The grantor purports to convey the right to exclude others, in the one instance, from a defined tract of land, and in the other, from a described and limited field of the useful arts.

Westinghouse, 266 U.S. at 350. Thus, the Supreme Court has endorsed at least one application of assignor estoppel, and we are therefore precluded from doing away with the doctrine in its entirety.

Synopsys' primary criticism of assignor estoppel is that fifty years ago the Supreme Court eliminated licensee estoppel in Lear, and according to Synopsys, the same logic ought to result in the elimination of assignor estoppel. In Lear, the Supreme Court ultimately held that the public interest in the free use of ideas in the public do-main outweighed the unfairness of potentially depriving the licensor of the full value of his contractual rights. Id. at 670–71. However, the issue of assignor estoppel was not squarely before the Court in Lear.

When an inventor/assignor assigns his patent rights to someone else for value, he may make an implicit representation that what he sold has value. See Diamond Scientific Co. v. Ambico, Inc., 848 F.2d 1220, 1224–25 (Fed. Cir. 1988). This implicit representation may warrant application of assignor estoppel. Id. at 1225. A licensee has not made a similar representation. The licensee did not sell the patent itself to the licensor. The licensee did not sign oaths or declarations or make asser-tions implicitly or explicitly about the patent's validity before inducing another to make an investment based on the perceived worth of the patent. Thus, the Supreme Court's abolition of licensee estoppel in Lear does not directly bear on the continued vitality of assignor estoppel. Diamond Scientific, 848 F.3d at 1224 (“Unlike the licensee, who, without Lear might be forced to continue to pay for a potentially invalid patent, the assignor who would challenge the patent has already been fully paid for the patent rights.”).

Because the record in this case and the arguments put forth by the parties do not convince us that we can or should entirely abolish the doctrine of assignor estoppel, we decide not to rehear this case.


1.   I also believe the panel decision to be consistent with long-standing damages principles in property, tort and contract. I do not agree with the dissent that there should be a special rule for damages in patent cases which is at odds with mainstream damages principles.

1.   In Panduit Corp. v. Stahlin Bros. Fibre Works, Inc., 575 F.2d 1152 (6th Cir. 1978), the Sixth Circuit held that a patentee can recover lost profits only if it can prove “(1) demand for the patented product, (2) absence of acceptable noninfringing substitutes, (3) his manufactur-ing and marketing capability to exploit the demand, and (4) the amount of the profit he would have made.” Id. at 1156. However, Panduit did not deal with apportionment since the patent in Panduit is directed to an electrical wiring duct that constituted the entire product in dispute. Id. at 1155.

2.   To be sure, Dobson and Garretson both also hold that recovery for all of the profits for a product is permit-ted if it can be shown that consumer demand is attributable to the patented feature. See Dobson, 114 U.S. at 444; Garretson, 111 U.S. at 121. This rule has become known as the entire market value rule. For the entire market value rule to apply, “the patentee must prove that ‘the patent-related feature is the “basis for customer demand.” ’ ” Lucent Techs., Inc. v. Gateway, Inc., 580 F.3d 1301, 1336 (Fed. Cir. 2009) (quoting Rite-Hite Corp. v. Kelley Co., 56 F.3d 1538, 1549 (Fed. Cir. 1995)(en banc)). When the entire market value rule applies, no apportionment is required. LaserDynamics, Inc. v. Quanta Comput-er, Inc., 694 F.3d 51, 67–68 (Fed. Cir. 2012). Here, however, the panel opinion does not invoke the entire market value rule, and the accused infringer was precluded by the district court from introducing evidence that the entire market value rule is inapplicable. See J.A. 42,241.

3.   The district court deemed this error harmless. J.A. 42,601.

1.   EVE is a subsidiary of Synopsys. References to Synopsys refer to all the Synopsys and EVE entities unless otherwise noted.

2.   The question of privity and the extent to which assignor estoppel ought to prohibit a future employer of the inventor/assignor from challenging the validity of the patent is an interesting one not raised by the parties on appeal in this case. This court has held that whether the assignor estoppel doctrine ought to apply to a future employer is a case specific determination that depends on the equities and the nature of the employee's role in the new company. See, e.g., Shamrock Tech., Inc. v. Med. Sterilization, Inc., 903 F.2d 789 (Fed. Cir. 1990). Given that Synopsys has made no arguments at any stage of this appeal challenging the extension of the inventor's assignor estoppel to it, this case does not provide a vehicle for consideration of these privity issues.


STOLL, Circuit Judge, with whom NEWMAN, MOORE, O'MALLEY, REYNA, and WALLACH, Circuit Judges, join, concurring in the denial of rehearing en banc. DYK, Circuit Judge, with whom HUGHES, Circuit Judge, joins, dissenting from the denial of rehearing en banc. MOORE, Circuit Judge, with whom CHEN, Circuit Judge, joins, concurring in the denial of panel rehearing.