1292 632 federal reporter, 3d series - …...1292 632 federal reporter, 3d series [2] because mr....

1292 632 FEDERAL REPORTER, 3d SERIES

[2] Because Mr. Knox is proceedingpro se, we liberally construe his pleadings.See Braxton v. Zavaras, 614 F.3d 1156,1159 (10th Cir.2010). Nevertheless, weconclude that the relief he seeks is notavailable to him in federal court.

[3, 4] To begin with, Mr. Knox’s claimsappear to be barred by the Rooker/Feld-man abstention doctrine, because he isessentially seeking review and reversal ofthe state-court denial of his name-changerequest. See Rooker v. Fidelity Trust Co.,263 U.S. 413, 44 S.Ct. 149, 68 L.Ed. 362(1923); D.C. Court of Appeals v. Feldman,460 U.S. 462, 103 S.Ct. 1303, 75 L.Ed.2d206 (1983). Under that doctrine, ‘‘a partylosing in state court is barred from seekingwhat in substance would be appellate re-view of the state judgment in a UnitedStates district court, based on the losingparty’s claim that the state judgment itselfviolates the loser’s federal rights.’’ John-son v. De Grandy, 512 U.S. 997, 1005–06,114 S.Ct. 2647, 129 L.Ed.2d 775 (1994).

[5] But even if Mr. Knox’s federal-court claims can survive, at least in part,Rooker–Feldman, he is not entitled to re-lief. To the extent that he is seeking reliefin the nature of mandamus, ordering De-fendants to take action in their capacitiesas state judges, ‘‘[w]e have no authority toissue such a writ to direct state courts ortheir judicial officers in the performance oftheir duties.’’ Van Sickle v. Holloway, 791F.2d 1431, 1436 n. 5 (10th Cir.1986) (inter-nal quotation marks omitted); see Olson v.Hart, 965 F.2d 940, 942 (10th Cir.1992)(‘‘Federal courts have no authority to issuea writ of mandamus to a state judge.’’).

[6] Likewise, he cannot obtain injunc-tive relief against Defendants. Althoughwe have previously said that a plaintiffmay obtain an injunction against a statejudge under 42 U.S.C. § 1983, see Harrisv. Champion, 51 F.3d 901, 905 (10th Cir.1995); Olson, 965 F.2d at 943, those state-ments were abrogated by the Federal

Courts Improvement Act of 1996, whichprovides that ‘‘injunctive relief [against ajudicial officer] shall not be granted unlessa declaratory decree was violated or de-claratory relief was unavailable,’’ 42 U.S.C.§ 1983. He has not shown that eithercondition was satisfied here.

[7] Finally, we note that Mr. Knoxasks this court to set aside the districtcourt’s determination that his suit is frivo-lous and malicious. His concern is that adismissal on that ground under 28 U.S.C.§ 1915A counts as a strike in determiningeligibility for in forma pauperis statusunder 28 U.S.C. § 1915(g). See Hafed v.Fed. Bureau of Prisons, 635 F.3d 1172,1176–77 (10th Cir.2011). We deny Mr.Knox’s request because, as shown by ourdiscussion of the merits of his claims, thedistrict court properly characterized themas frivolous.

CONCLUSIONWe AFFIRM the judgment below. We

deny Mr. Knox’s motion to proceed informa pauperis on appeal and order himto immediately remit the unpaid balancedue of the filing fee.

,

UNILOC USA, INC. and UnilocSingapore Private Limited,

Plaintiffs–Appellants,

v.

MICROSOFT CORPORATION,Defendant–Cross

Appellant.

Nos. 2010–1035, 2010–1055.

United States Court of Appeals,Federal Circuit.

Jan. 4, 2011.

Rehearing Denied March 22, 2011.Background: Patentee brought action forinfringement of patent for a software reg-

1293UNILOC USA, INC. v. MICROSOFT CORP.Cite as 632 F.3d 1292 (Fed. Cir. 2011)

istration system directed towards reducingthe unauthorized use of software. TheUnited States District Court for the Dis-trict of Rhode Island, William E. Smith, J.,447 F.Supp.2d 177, granted summaryjudgment as to noninfringement. Patenteeappealed. The Court of Appeals, 290 Fed.Appx. 337, reversed and remanded. Fol-lowing jury trial on remand, jury returnedverdict awarding patentee $388,000,000 indamages. The District Court, Smith, J.,640 F.Supp.2d 150, granted alleged in-fringer judgment as matter of law (JMOL)in part, and granted new trial on infringe-ment and willfulness. Patentee appealed.Holdings: The Court of Appeals, Linn,Circuit Judge, held that:(1) whether additional structural compo-

nents in message digest and securehash algorithms contained in allegedinfringer’s software products were‘‘summation algorithms’’ as claimed inpatent was fact issue for the jury;

(2) whether product activation system con-tained in alleged infringer’s softwareproducts used a ‘‘registration system’’with a ‘‘mode switching means,’’ asclaimed in patent was fact issue forjury;

(3) alleged infringer did not act despite anobjectively high likelihood that its ac-tions constituted infringement, as re-quired to show willful infringement ofpatent;

(4) evidence relying on ‘‘25 percent rule ofthumb,’’ which was tool used to approx-imate the reasonable royalty rate themanufacturer of patented productwould be willing to offer to pay to thepatentee during a hypothetical negotia-tion, was inadmissible under Daubert;and

(5) District Court did not abuse its discre-tion in granting conditional new trialon damages.

Affirmed in part and reversed in part.See also 447 F.Supp.2d 177.

1. Patents O312(1.1)To prove infringement of patent, the

plaintiff bears the burden of proof to showthe presence of every element or its equiv-alent in the accused device.

2. Patents O324.55(5)The underlying issue of infringement

of a patent is a question of fact reviewedby the Court of Appeals for substantialevidence.

3. Federal Courts O765, 776Court of Appeals reviews a district

court’s denial of judgment as a matter oflaw (JMOL) after a jury verdict de novo,asking whether the evidence points sostrongly and overwhelmingly in favor ofthe moving party that no reasonable jurycould have returned a verdict adverse tothat party.

4. Patents O314(5)Whether additional structural compo-

nents in message digest and secure hashalgorithms contained in alleged infringer’ssoftware products were ‘‘summation algor-ithms,’’ as claimed in patent for a softwareregistration system, was a fact issue forthe jury in patent infringement action. 35U.S.C.A. § 271(a).

5. Evidence O508, 555.2Under Daubert, the district court

must exercise its ‘‘gatekeeper’’ function inensuring that scientific testimony is rele-vant, reliable and admissible. Fed.RulesEvid.Rule 702, 28 U.S.C.A.

6. Evidence O574It is decidedly the jury’s role to evalu-

ate the weight to be given to the testimonyof dueling qualified experts.

7. Patents O314(5)Whether product activation system

contained in alleged infringer’s softwareproducts used a ‘‘registration system’’ witha ‘‘mode switching means,’’ as claimed in


patent for a software registration system,was fact issue for jury in infringementaction; use of product during ‘‘grace peri-od’’ after agreement to end user licenseagreement (EULA) and before product ac-tivation in accused products did not consti-tute full use in accordance with EULA,and patent was not limited to situationwhere activation and licensing were con-current. 35 U.S.C.A. § 112.

8. Federal Civil Procedure O2339A new trial is only appropriate when

the outcome is against the clear weight ofthe evidence such that upholding the ver-dict will result in a miscarriage of justice.

9. Federal Civil Procedure O2373In contrast to ruling on a motion for

judgment as a matter of law (JMOL), inconsidering a motion for a new trial, thedistrict court may independently weigh theevidence.

10. Federal Courts O825.1District court is entitled to deference

in granting a new trial motion, and theCourt of Appeals only overturns grant ofnew trial if district court has abused itsdiscretion.

11. Patents O312(4, 8)To establish willful infringement of a

patent, a patentee must show by clear andconvincing evidence that infringer acteddespite an objectively high likelihood thatits actions constituted infringement of avalid patent; in addition, a patentee mustshow the risk was either known or soobvious that it should have been known tothe accused infringer.

12. Patents O227Alleged infringer did not act despite

an objectively high likelihood that its ac-tions constituted infringement, as requiredto show willful infringement of patent for asoftware registration system, absent anyevidence at trial or on appeal from patentholder showing why alleged infringer, at

time it began infringement, could not havereasonably determined that product’s al-gorithms did not meet patent’s limitations.

13. Patents O319(1)

A reasonable royalty, as part of anaward of damages upon finding infringe-ment of valid patent, is often determinedon the basis of a hypothetical negotiation,occurring between the parties at the timethat infringement began. 35 U.S.C.A.§ 284.

14. Evidence O555.9

Evidence relying on ‘‘25 percent ruleof thumb,’’ which was tool used to approxi-mate the reasonable royalty rate the man-ufacturer of patented product would bewilling to offer to pay to the patenteeduring a hypothetical negotiation, was in-admissible under Daubert, since it failed totie a reasonably royalty base to facts ofcase at issue. Fed.Rules Evid.Rule 702,28 U.S.C.A.

15. Patents O312(1.7)

The patentee bears the burden ofproving damages in infringement action.35 U.S.C.A. § 284.

16. Evidence O555.9

To be admissible under Daubert, ex-pert testimony opining on a reasonableroyalty rate in patent infringement casemust carefully tie proof of damages to theclaimed invention’s footprint in the marketplace. Fed.Rules Evid.Rule 702, 28U.S.C.A.

17. Patents O318(4.1)

In patent infringement action, the‘‘entire market value rule’’ allows a paten-tee to assess damages based on the entiremarket value of the accused product onlywhere the patented feature creates thebasis for customer demand or substantially


creates the value of the component parts.35 U.S.C.A. § 284.

See publication Words and Phras-es for other judicial constructionsand definitions.

18. Patents O323.3District Court did not abuse its discre-

tion in granting conditional new trial ondamages in action for infringement of apatent for a software registration system,where patentee’s references to total dollarvolume of sales of accused products toshow that its claimed royalties were rea-sonable violated the ‘‘entire market valuerule.’’ Fed.Rules Civ.Proc.Rule 59, 28U.S.C.A.

19. Patents O314(5)Issue of whether patent for a software

registration system was invalid due to ob-viousness in light of prior art, even ifsummation algorithm as used in patentwould have been an obvious or inferiorsubstitute for the algorithms used in priorart patents, was fact issue for jury inpatent infringement action.

Patents O328(2)4,658,093. Cited as Prior Art.

Patents O328(2)5,474,555, 6,263,432. Cited.

Patents O328(2)5,490,216. Infringed.

Donald R. Dunner, Finnegan,Henderson, Farabow, Garrett and Dunner,LLP, of Washington, DC, argued for plain-tiff-appellant. With him on the brief wereDon O. Burley; Erik R. Puknys and AaronJ. Capron, of Palo Alto, CA. Of counsel onthe brief were Paul J. Hayes and Dean G.Bostock, Mintz, Levin, Cohn, Ferris, Glov-sky & Popeo PC, of Boston, MA.

Frank E. Scherkenbach, Fish and Rich-ardson P.C., of Boston, MA, argued for

defendant-cross appellant. With him onthe brief were Kurt L. Glitzenstein; JohnW. Thornburgh, of San Diego, CA andLaura R. Braden, of Washington, DC.

Before RADER, Chief Judge, LINNand MOORE, Circuit Judges.

LINN, Circuit Judge.

Uniloc USA, Inc. and Uniloc SingaporePrivate Limited (collectively, ‘‘Uniloc’’) ap-peal from the decision of the United StatesDistrict Court for the District of RhodeIsland granting Microsoft Corporation’s(‘‘Microsoft’’) motion for judgment as amatter of law (‘‘JMOL’’) of non-infringe-ment and no willful infringement of assert-ed claims of Uniloc’s U.S. Patent No.5,490,216 (‘‘8216 patent’’), and, in the alter-native, granting a new trial on infringe-ment and willfulness. Uniloc USA, Inc. v.Microsoft Corp., 640 F.Supp.2d 150 (D.R.I.Sept.29, 2009) (‘‘Uniloc II ’’). Uniloc alsoappeals the district court’s alternativegrant of a new trial on damages. Micro-soft cross-appeals the district court’s deni-al of its motion for JMOL of invalidity ofthe 8216 patent. Id. at 179–83.

Because the jury’s verdict on infringe-ment was supported by substantial evi-dence, this court reverses the districtcourt’s grant of JMOL of non-infringe-ment; this court also reverses the districtcourt’s alternative grant of a new trial oninfringement as an abuse of discretion.Because the jury’s verdict on willfulnesswas not supported by substantial evidence,this court affirms the district court’s grantof JMOL of no willfulness; the districtcourt’s alternative grant of a new trial forwillfulness is thus rendered moot. Be-cause the jury’s damages award was fun-damentally tainted by the use of a legallyinadequate methodology, this court affirmsthe grant of a new trial on damages. Fi-nally, because the district court did notabuse its discretion in determining that thejury verdict of no invalidity of the 8216


patent was supported by substantial evi-dence, we affirm the district court’s denialof Microsoft’s motion for JMOL of invalidi-ty.

I. BACKGROUND

Commercial software manufacturers likeMicrosoft lose significant sales as a resultof the ‘‘casual copying’’ of software, whereusers install copies of a software programon multiple computers in violation of appli-cable software license conditions. Uniloc’s8216 patent was an early attempt to com-bat such software piracy. There is nodispute as to the actual functioning of Uni-loc’s patented invention and Microsoft’saccused products. The following back-ground information is taken from the dis-trict court’s opinion. Uniloc II, 640F.Supp.2d 150.

A. The 8216 Patent

Uniloc’s 8216 patent is directed to asoftware registration system to detercopying of software. The system allowsthe software to run without restrictions(in ‘‘use mode’’) only if the system deter-

mines that the software installation is le-gitimate. A representative embodimentfunctions as follows. First, a user intend-ing to use the software in ‘‘use mode’’enters certain user information whenprompted, which may include a softwareserial number and/or name and addressinformation. An algorithm on the user’scomputer (a ‘‘local licensee unique ID gen-erating means’’) combines the inputted in-formation into ‘‘a registration numberunique to an intending licensee’’ (a ‘‘locallicensee unique ID’’). 8216 patent, Ab-stract. The user information is also sentto the vendor’s system, which performsthe identical algorithm (a ‘‘remote licenseeunique ID generating means’’) to create a‘‘remote licensee unique ID’’ for the user.When the application boots again, a ‘‘modeswitching means’’ compares the local andremote licensee unique IDs. If theymatch, the program enters into ‘‘usemode.’’ If they do not match, the pro-gram enters into ‘‘demo mode,’’ whereincertain features are disabled. Figure 8from the 8216 patent shows the fifth pre-ferred embodiment:


8216 patent, Fig. 8.

Uniloc asserts only independent claim19:

19. A remote registration station incor-porating remote licensee unique ID gen-erating means, said station forming partof a registration system for licensingexecution of digital data in a use mode,said digital data executable on a plat-form, said system including local licen-see unique ID generating means, saidsystem further including mode switch-ing means operable on said platformwhich permits use of said digital data insaid use mode on said platform only if alicensee unique ID generated by saidlocal licensee unique ID generatingmeans has matched a licensee unique IDgenerated by said remote licenseeunique ID generating means; andwherein said remote licensee unique ID

generating means comprises softwareexecuted on a platform which includesthe algorithm utilized by said local licen-see unique ID generating means to pro-duce said licensee unique ID.

8216 patent, col. 15 l.21—col. 16 l.9 (empha-sis added).

B. The Accused Product

The accused product is Microsoft’s Prod-uct Activation feature that acts as a gatek-eeper to Microsoft’s Word XP, Word 2003,and Windows XP software programs.Upon receipt of Microsoft’s retail softwareprogram, the user must enter a 25–charac-ter alphanumeric product key containedwithin the packaging of Microsoft’s retailproducts. If the Key is valid, the user isasked to agree to the End User LicenseAgreement (‘‘EULA’’), by which the li-censor-licensee relationship is initiated.


At about this time, the software createsa Product ID (‘‘PID’’) and a Hardware ID(‘‘HWID’’) on the user’s computer. ThePID is formed from the combination of theProduct Key, information from the soft-ware CD, and a random number from theuser’s computer. The HWID is generatedfrom information about the user’s comput-er. The user may use the software with-out initiating Product Activation, but suchuse is temporally limited (50 start-ups ofOffice and 30 days use of Windows untilbasic functions like saving and printing aredeactivated) and functionally limited (noupdates can be downloaded and installed).If the user elects to initiate Product Acti-vation, the software sends a digital licenserequest to Microsoft over the internet,which includes: the PID, the HWID, andadditional activation information. At Mi-

crosoft’s remote location, this informationis entered into one of two software algor-ithms: the MD5 message digest algorithm(‘‘MD5’’) for Office products and the SHA–1 secure hash algorithm (‘‘SHA–1’’) forWindows products.1

The functionality of the MD5 and SHA–1 algorithms is at the heart of this case.As the district court noted, Microsoft’s ex-pert, Dr. Wallach, provided a demonstra-tive animation, which gives a clear, de-tailed, and uncontested explanation of thefunctionality of these algorithms. The ani-mation and the explanation of it given byDr. Wallach at the trial (Trial Tr. 157:21–166:3, Mar. 31, 2009) may be downloadedat http://oralarguments.cafc.uscourts.gov/animation/Uniloc.2010–1035.Animation.pdf. The first frame is explained below.

First, A, B, C, and D are 32–bit numbers and F, G, H, and I are logical opera-

1. As discussed in the animation referred to,infra, the only differences between MD5 andSHA–1 are the added logical operation groupand shifting step in SHA–1. Throughout thislitigation, the two algorithms have been treat-

ed as functionally identical for infringementpurposes. For ease of presentation, this opin-ion discusses only the MD5 algorithm, but itis uncontested that the same analysis appliesto both.


tions into which the numbers B, C, and Dare entered. For example, F(x, y, z) = (xAND y) OR ((NOT x) AND z). Thus, if B,C, and D are 0, 1, 1, respectively, F(0, 1, 1)= 1. The result is that three 32–bit num-bers are compressed into a single 32–bitnumber.

Second, the resulting number is addedto A using modular addition. Modularaddition is a way of adding that resets thecount of a sum after a certain thresholdnumber is reached. The most familiarexample is the American A.M./P.M. clocksystem. If a three hour movie begins at11:00 A.M., it will end at 2:00 P.M. This isan example of mod12 addition: one firstadds 11v3=14 then subtracts 12 to get 2.Modular addition, or modulo-addition, isused throughout the MD5 algorithm.

Third, M1, the first component of thesource message being hashed, is modulo-added to the result from step 2.

Fourth, additive constant Ki is modulo-added to the result from step 3.

Fifth, the resulting number is ‘‘circularshifted’’ (according to Microsoft) or ‘‘leftshifted’’ (according to Uniloc); the shift isdepicted by the green box with the threearrows in the diagram. Because the actu-al functionality is not disputed, we willrefer to it as ‘‘circular shifting’’ for thepurposes of this opinion. Inserting intothe shifter the binary number 0100 10112

(which is the binary equivalent of 75), andshifting it by 1 place would yield 1001 0110(which is the binary equivalent of 150).The result of this operation is a multiplica-tion by two for each single unit shift. Ifthe number is again shifted (or if theoriginal shift was by 2 places), the outputbecomes 0010 1101 (which is the binaryequivalent of 45). Because a single 8–bitstring cannot represent numbers larger

than 255 (1111 1111), such numbers arerepresented in mod255; thus the 300 ex-pected from multiplying 150 by 2, becomes45 in mod255 (300–255).

Sixth, the resulting number is then mo-dulo-added to initial value B, which finalnumber becomes the new value B’. Initialvalue C becomes new value D’, D becomesA’, and A becomes B’. The hashing algor-ithm is then run again using these newvalues (A’, B’, C’, D’) in place of the old (A,B, C, D) and the second component of themessage (M2) in place of the first (M1).After sixteen rounds of this, a differentlogical function, G(x, y, z) is used, and thesame message string is input in a differentorder. The function G is used for sixteenrounds, followed by sixteen rounds of func-tion H and sixteen rounds of function I.The end result is a ‘‘license digest,’’ i.e. ‘‘ashortened fixed-bit output,’’ Uniloc II, 640F.Supp.2d at 157, derived from the originalmessage.

Microsoft encrypts this digest, andsends it along with the original data backto the user’s computer. The software onthe user’s computer decrypts the messageand recovers the ‘‘license digest.’’ It theninputs the original data (i.e. the PID,HWID, and additional activation informa-tion) and enters it into the same MD5 orSHA–1 algorithm used by Microsoft’s com-puters, resulting in a local ‘‘license digest.’’Microsoft’s Product Activation softwarecompares the local license digest and theremote license digest; if they match, thesoftware product is activated. If they donot, the software returns to pre-ProductActivation mode.

C. Procedural History

In the first iteration of this case, thedistrict court issued a claim construction

2. For demonstration purposes, we use an 8–bit number, though the MD5 algorithm uses a32–bit number. Four of the characters are

italicized to demonstrate the effect of the cir-cular shifter.


ruling, construing several terms that arerelevant to the instant appeal. UnilocUSA, Inc. v. Microsoft Corp., 447F.Supp.2d 177 (D.R.I.2006) (‘‘Uniloc IClaim Construction ’’). These appear be-low.‘‘Licensee ‘‘A unique identifier associatedunique ID’’ with a licensee.’’ Id. at 183.‘‘Local licensee Means plus function. ‘‘Func-unique ID gen- tion: to generate a local orerating means’’ remote licensee unique ID’’and ‘‘Remote and ‘‘Structure: a summationlicensee unique algorithm or a summer andID generating equivalents thereof.’’ Id. atmeans’’ 190.‘‘Use mode’’ ‘‘A mode that allows full use of

the digital data or software inaccordance with the license.’’Id. at 196.

‘‘Mode switching Means plus function. ‘‘Func-means’’ tion: to permit the digital data

or software to run in a usemode if the locally generatedlicensee unique ID matcheswith the remotely generated li-censee unique ID.’’ and ‘‘Struc-ture: program code which per-forms a comparison of twonumbers or a comparator andequivalents thereof.’’ Id. at198 (synonyms in the claimconstruction not relevant tothe claim at issue are removedfor simplicity).

‘‘Registration ‘‘A system that allows digitalsystem’’ data or software to run in a

use mode on a platform if andonly if an appropriate licensingprocedure has been followed.’’Id. at 202.

The district court granted summary judg-ment of non-infringement, Uniloc USA,Inc. v. Microsoft Corp., No. 03–440 (D.R.I.Oct. 19, 2007), concluding that the algor-ithm used at Microsoft’s remote station togenerate a licensee unique ID was notidentical to the algorithm used on theuser’s local station as required by the lastlimitation in the claim. Id. at 24.

On appeal, this court reversed and re-manded the finding of non-infringement,holding that Uniloc had put forth ‘‘exten-sive and by no means conclusory’’ evidencethat Microsoft’s Product Activation usedthe same algorithm at the local and remotesites (respectively, the ‘‘local licensee

unique ID generating means’’ and ‘‘remotelicensee unique ID generating means’’),and that the issue of whether the accusedproducts met this limitation should havegone to the jury. Uniloc USA, Inc. v.Microsoft Corp., 290 Fed.Appx. 337, 343(Fed.Cir.2008) (non-precedential) (‘‘UnilocI ’’). In that appeal, Microsoft presentedseveral alternative grounds for affirmance,including several arguments centeredaround the lack of any information in thatis ‘‘uniquely associated with the person’’that results in a ‘‘licensee unique ID.’’ Br.of Microsoft Corp. at 37–53, Uniloc I (Mar.19, 2008). This court held that ‘‘the licen-see unique ID does not require personalinformation about the user,’’ so long as it is‘‘unique,’’ and not ‘‘based solely on plat-form-related user information.’’ Uniloc I,290 Fed.Appx. at 342–43. We explicitlynoted that the specification of the 8216patent ‘‘leave[s] open the possibility thatvendor-provided information, like Micro-soft’s Product Key, could be the basis for a‘licensee unique ID.’ ’’ Id. at 344. Micro-soft also argued that Product Activationlacked a ‘‘licensee unique ID generatingmeans.’’ This court summarily rejectedthat argument, noting that ‘‘[w]e have con-sidered these arguments [for affirmanceon alternative grounds] and conclude theyare without merit.’’ Id. at 342.

On remand, the district court first re-jected several in limine motions, includinga motion by Microsoft to exclude any testi-mony by Uniloc’s damages expert, Dr.Gemini, under Daubert v. Merrell DowPharmaceuticals, Inc., 509 U.S. 579, 589,113 S.Ct. 2786, 125 L.Ed.2d 469 (1993) andFederal Rule of Evidence 702, for his useof an allegedly arbitrary baseline rate of$10–per–activation, and the use of a 25percent rule of thumb. Uniloc USA, Inc.v. Microsoft Corp., 632 F.Supp.2d 147,150–51 (D.R.I. 2009) (‘‘In Limine ’’). Af-ter a full trial, the jury returned a verdictof infringement and no invalidity of claim


19 of the 8216 patent, and found Micro-soft’s infringement to be willful. The juryawarded Uniloc $388 million in damages.In post trial motions, Microsoft asked for:(1) JMOL of invalidity due to anticipationand obviousness; (2) JMOL of non-in-fringement of the ‘‘licensee unique ID gen-erating means’’ and ‘‘registration sys-tem’’/‘‘mode switching means’’ limitations;(3) JMOL of non-infringement because Mi-crosoft could not have directly infringedthe system because claim 19 requires actsto be taken on the user’s local computerover which Microsoft has no control; (4)JMOL of no willfulness; (5) a new trial ondamages for the improper use of the 25%rule of thumb and the entire market valuerule; and (6) in the alternative, a new trialon infringement and willfulness.

The district court, in a comprehensiveand well-reasoned opinion, denied JMOLof invalidity, granted JMOL of non-in-fringement on the basis of both contestedclaim limitations, granted JMOL of no will-fulness, granted a new trial on damages onthe improper use of the entire marketvalue rule, rejected Microsoft’s argumentsregarding the 25 percent rule of thumb ashaving been previously decided, Uniloc II,640 F.Supp.2d at 184 and n. 42, and grant-ed in the alternative a new trial on in-fringement and willfulness. The districtcourt also considered and rejected Micro-soft’s contentions that it could not directlyinfringe the asserted claims. The detailsof the district court’s opinion are discussedmore fully below.

Uniloc appeals all but the denial ofJMOL of invalidity, which Microsoft cross-appeals. We have jurisdiction under 28U.S.C. § 1295(a)(1).

II. DISCUSSION

A. Infringement

[1, 2] To prove infringement, the plain-tiff bears the burden of proof to show thepresence of every element or its equivalent

in the accused device. Lemelson v. UnitedStates, 752 F.2d 1538, 1551 (Fed.Cir.1985).The underlying infringement issue is aquestion of fact reviewed for substantialevidence. Finisar Corp. v. DirecTVGroup, Inc., 523 F.3d 1323, 1332 (Fed.Cir.2008).

[3] This court’s review of a districtcourt’s grant of JMOL is governed byregional circuit law. Union CarbideChems. & Plastics Tech. Corp. v. Shell OilCo., 425 F.3d 1366, 1372 (Fed.Cir.2005).The First Circuit reviews a district court’sdenial of JMOL after a jury verdict denovo, asking whether ‘‘the evidence pointsso strongly and overwhelmingly in favor ofthe moving party that no reasonable jurycould have returned a verdict adverse tothat party.’’ Keisling v. SER–Jobs forProgress, Inc., 19 F.3d 755, 759–60 (1stCir.1994). This court may not evaluate‘‘the credibility of witnesses, resolve con-flicts in testimony, or evaluate the weightof the evidence,’’ but must view the evi-dence in the light most favorable to Uniloc.Gibson v. City of Cranston, 37 F.3d 731,735 (1st Cir.1994).

Microsoft argues that because there isno dispute about how the accused productswork, infringement should be reviewed denovo. Br. of Microsoft Corp. at 18, 22–23(citing Athletic Alts., Inc. v. Prince Mfg.,Inc., 73 F.3d 1573, 1578 (Fed.Cir.1996)(‘‘Where, as here, the parties do not dis-pute any relevant facts regarding the ac-cused product but disagree over which oftwo possible meanings of Claim 1 is theproper one, the question of literal infringe-ment collapses to one of claim construc-tion, and should thus be reviewed denovo.’’) and General Mills, Inc. v. Hunt–Wesson, Inc., 103 F.3d 978, 983 (Fed.Cir.1997) (similar)). It is well-settled that in-fringement is a factual issue, reviewed forsubstantial evidence. E.g. Finjan, Inc. v.Secure Computing, Corp., 626 F.3d 1197


(Fed.Cir.2010); Connell v. Sears, Roebuck& Co., 722 F.2d 1542, 1546 (Fed.Cir.1983).The cases cited by Microsoft involve aprocedural posture not present in thiscase. As this court noted in InternationalRectifier Corp. v. IXYS Corp., 361 F.3d1363, 1374 (Fed.Cir.2004), the infringe-ment issue in General Mills collapsed intoclaim construction because ‘‘the partiesagreed with each other and the districtcourt about how each of two competingclaim constructions would apply to the un-disputed structure of the accused inven-tion.’’ In other words, the parties con-ceded that under one claim constructionthere was infringement and under the oth-er there was none, and were arguing onlyover which claim construction was appro-priate. The infringement issue in AthleticAlternatives also came to this court withthe same posture. See 73 F.3d at 1581(‘‘We conclude that Claim 1 of the 8097patent includes the limitation that thesplay-creating string end offset distancetake on at least three values, i.e., a mini-mum, a maximum, and at least one inter-mediate value. We thus affirm the districtcourt’s conclusion that Claim 1 does notliterally read on the Vortex racket.’’). Asdiscussed below, this case presents theopposite procedural posture; the claimconstruction itself is not contested, but theapplication of that claim construction tothe accused device is. Thus, this courtapplies the traditional rule for review ofjury verdicts of factual issues discussedabove.

This case presents three primary in-fringement issues: 1) whether the accusedproducts contain ‘‘licensee unique ID gen-erating means’’; 2) whether the accusedproducts contain a ‘‘registration system’’with a ‘‘mode switching means’’ that pre-cludes full use of the software unless theoutputs of the local and remote algorithmsmatch; and 3) whether Microsoft can beliable for direct infringement when it hasno control over the user’s computer.

1. ‘‘Licensee Unique IDGenerating Means’’

The 8216 patent specification describesthe licensee unique ID generating meansas an algorithm that functions by ‘‘com-bin[ing] by addition the serial number 50with the software product name 64 andcustomer information 65 and previous useridentification 22 to provide registrationnumber 66.’’ Id. col. 11 ll.53–56. Thedistrict court’s construction of ‘‘licenseeunique ID generating means’’ is undisput-ed on appeal: it is a means plus functionclaim, with the function being ‘‘to generatea local or remote licensee unique ID’’ andthe structure being ‘‘a summation algor-ithm or a summer and equivalents there-of.’’ Uniloc I Claim Construction, 447F.Supp.2d at 190.

The district court determined that noreasonable jury could find that the accusedproducts were summation algorithms, andgranted JMOL of non-infringement. Thedistrict court gave seven reasons for itsdecision: (1) the ‘‘circular shifting and mix-ing functions fundamentally create a moresecure result compared to an algorithmbased in summation as the specificationdiscloses,’’ Uniloc II, 640 F.Supp.2d at 170;(2) summation is reversible and MD5 isirreversible and much more complicated,id. (citing Business Objects, S.A. v. Mi-crostrategy, Inc., 393 F.3d 1366, 1370 (Fed.Cir.2005)); (3) ‘‘MD5 achieves its functionin a way an algorithm based in summationcould not,’’ id.; (4) the 8216 patent con-tained only a narrow structural disclosurethat is not entitled to a broad scope, id. at171; (5) the documentary evidence pre-sented by Uniloc did not show what ‘‘thecomplex hashes in this case actually do,and whether that is equivalent to the ‘byaddition’ structure Uniloc disclosed,’’ id. at172; (6) Uniloc did not put forth expertopinion interpreting the documents, exceptfor Klausner’s presentation of ‘‘factual in-


formation under the guise of opinion,’’ id.at 172 and n. 25 (citing Centricut, LLC v.Esab Grp., Inc., 390 F.3d 1361, 1369–70(Fed.Cir.2004), but noting that case is ‘‘nota perfect fit’’); and (7) ‘‘[t]he jury ‘lacked agrasp of the issues before it,’ ’’ id. at 173(citing Tex. Instruments Inc. v. CypressSemiconductor Corp., 90 F.3d 1558, 1570(Fed.Cir.1996)) because it ‘‘ignored Dr.Wallach’s admittedly complex explanationand embraced Mr. Klausner’s’’ ‘‘incom-plete, oversimplified and frankly inappro-priate explanation,’’ id. at 170 n. 21.

Uniloc argues that a reasonable jurycould have concluded that MD5 and SHA1were summation algorithms within themeaning of the 8216 patent, and that thedistrict court erred in granting JMOL ofnon-infringement. The jury heard twosets of evidence in favor of Uniloc’s conten-tion that MD5 and SHA1 were summationalgorithms. First, Dr. Klausner, Uniloc’sexpert, testified that MD5 makes a digestof the message it receives

by doing addition and multiplication in aseries of rounds over and over again. Ittakes a piece of the input, adds andshifts it, takes another piece of the in-put, adds and shifts it. It does a num-ber of other operations, what are calledlogical operations in mathematics. Butthe essence is it eventually adds each ofthe results of these piece-wise opera-tions into a bucket or a hash, and thathash becomes the output of the algor-ithm.

He also testified that MD5 uses ‘‘two pri-mary kinds of operations to do its work.One is addition, summing; and the otheris what we call left shifting TTT [which is]actually nothing more than multiplicationTTT [which] is nothing more than additiondone over and over again.’’ Klausner thenidentified the source code that was thebasis of his understanding that MD5 per-formed addition, noting that ‘‘I’m not say-ing that that’s all that MD5 does, but

that’s a significant portion of the MD5algorithm.’’ Second, the jury saw docu-mentary evidence identifying MD5 as, orequating it to, a summation algorithm.For example, in Microsoft’s Windows Pro-tocols Master Glossary, one entry reads‘‘checksum: A value that is the summa-tion of a byte stream. By comparing thechecksums computed from a data item attwo different times, one can quickly assesswhether the data items are identical.’’That same document equates ‘‘hashes’’and ‘‘checksums,’’ and notes that ‘‘[w]ell-known hash algorithms for computer hash-es include MD4, MD5, and SHA1.’’ OtherMicrosoft documents also refer to the out-come of the MD5 algorithm as a ‘‘check-sum.’’ See Kenneth Pfeil, Data Securityand Data Availability in the Administra-tive Authority, Microsoft TechNet, avail-able at http://technet.microsoft.com/en-us/library/cc722918.aspx (‘‘Hashing. Hash-ing is also referred to as MD5 check-sum.’’). See also Windows Driver Kit:Network Devices and Protocols:NDIS TASK IPSEC, MSDN, updateddocument available at http://msdn.microsoft.com/en-us/library/ff558990.aspx(‘‘MD5 Set by a miniport driver to indi-cate that its NIC can use the keyed MD5algorithm for computing and/or validatinga cryptographic checksum for an AH pay-load and/or ESP payload.’’). In addition,Uniloc relied on U.S. Patent No. 6,263,432(‘‘8432 patent’’), which, in describing theprocedure for generating a secure e-ticket,includes the following step:

In this example, each of the four fieldsin the ‘eticket’ framework 302 and userextension 304 include data representedby the number ‘1’ in step S1. The mes-sage Digest/Hash is represented by asummation (S) algorithm (equated to, orexemplary of, the MD5 protocol or otherhashing algorithm). Hence, to calculatethe Message Digest/Hash, a summation


algorithm is implemented using all eightfields of data in step 2.

8432 patent col. 9 ll.50–57.Uniloc also argues that the district court

improperly narrowed the claim construc-tion on JMOL from ‘‘summation algor-ithm’’ to a ‘‘simple combination of inputsby addition,’’ Uniloc, 640 F.Supp.2d at 170,which was improper under Hewlett–Pack-ard Co. v. Mustek Sys., Inc., 340 F.3d1314, 1320 (Fed.Cir.2003).

Microsoft counters with three argu-ments. First, Uniloc’s interpretation of‘‘summation algorithm’’ would be so broadas to cover any algorithm with a plus sign,and would be akin to adopting the rejectedclaim construction of licensee unique IDgenerating means as simply ‘‘an algor-ithm.’’ Microsoft argues that such a broadreading is inconsistent with the patent,which disclosed only the specific structurewhere ‘‘the algorithm, in this embodiment,combines by addition,’’ 8216 patent, col. 11ll.53–56. Second, Microsoft notes that itsexpert, Dr. Wallach, established that thealgorithms as a whole are not ‘‘summationalgorithm[s]’’ because neither circularshifting nor the logical operations of MD5and SHA1 are addition-based, and that histestimony was unrebutted, because the dis-trict court prevented Uniloc’s expert, Dr.Klausner, from opining that MD5 andSHA1 were summation algorithms withinthe scope of claim 19 of the 8216 patent.Finally, Microsoft contrasts the purpose ofMD5 and SHA1—to irreversibly scramblethe data so that the inputs cannot be de-rived—with the purpose of the summationalgorithms in the 8216 patent, to put datatogether by addition.

[4] As this court held in Uniloc I, 290Fed.Appx. at 342, there was substantialevidence for a jury to conclude that theoutput of the MD5 and SHA1 algorithmswas a licensee unique ID. Thus, bothMD5 and the summation algorithm in the8216 patent perform the same function of

generating a licensee unique ID. It is alsoundisputed that MD5 and SHA1 use someaddition to perform this function. UnilocII, 640 F.Supp.2d at 168. Thus, the issueis whether the additional structural compo-nents of MD5 and SHA1 preclude a rea-sonable jury from finding that they are‘‘summation algorithm[s].’’ This courtagrees with Uniloc that they do not.

First, the breadth of claim 19 is not asnarrow as Microsoft argues and the dis-trict court concluded. ‘‘The literal scope ofa properly construed means-plus-functionlimitation does not extend to all means forperforming a certain function. Rather, thescope of such claim language is sharplylimited to the structure disclosed in thespecification and its equivalents.’’ J & MCorp. v. Harley–Davidson, Inc., 269 F.3d1360, 1367 (Fed.Cir.2001). Nevertheless,in determining equivalence under § 112¶ 6, ‘‘the range of permissible equivalentsdepends upon the extent and nature of theinvention.’’ IMS Tech., Inc. v. Haas Auto-mation, Inc., 206 F.3d 1422, 1436 (Fed.Cir.2000) (citing Tex. Instruments, Inc. v. ITC,805 F.2d 1558, 1563 (Fed.Cir.1986)).‘‘More particularly, when in a claimed‘means’ limitation the disclosed physicalstructure is of little or no importance tothe claimed invention, there may be abroader range of equivalent structuresthan if the physical characteristics of thestructure are critical in performing theclaimed function in the context of theclaimed invention.’’ Id. The structural dis-closure in the 8216 patent is not limited tosimple addition in the colloquial sense ofadding numbers together and nothingmore. In the sixth embodiment, fromwhich the summation structure was de-rived, the algorithm ‘‘combines by additionthe serial number 50 with the softwareproduct name 64 and customer information65 and previous user identification 22 toprovide registration number.’’ 8216 patentcol.11 ll.54–57. This ‘‘combination by addi-


tion’’ necessarily incorporates an initialstep of converting the information into acommon format to be added, which re-quires more than simple addition. More-over, there is no indication that the sum-mation structure was critical to the 8216patent’s licensee unique ID generatingmeans algorithm’s function of generating alicensee unique ID. In fact, the 8216 pat-ent repeatedly refers to the licenseeunique ID generating means by the gener-ic phrase, ‘‘an algorithm,’’ e.g. 8216 patentcol.2 ll.65–66, and makes clear that theimportance of the algorithm is only that itbe ‘‘adapted to generate a registrationnumber which is unique to an intendinglicensee.’’ Id. col.2 ll.66–67. It may wellbe that the structural disclosure of thelicensee unique ID generating means limi-tation is minimal because of the relativeunimportance of the particular structure ofthat element. This does not, as Microsoftargues, result in pure functional claiming,nor expand the claim construction to Uni-loc’s proposed and rejected one of ‘‘analgorithm.’’ Declining to limit the con-struction to simple addition does not alsoextend the claims to any algorithm thatincludes a plus sign; the construction re-tains its explicit limitation that the algor-ithm used be fairly capable of categoriza-tion as ‘‘a summation algorithm.’’

Second, a jury could reasonably deter-mine that MD5 and SHA1 were not asradically different from the summation al-gorithm disclosed in the 8216 as Microsoftand the district court determined them tobe. Klausner testified that the ‘‘essence’’of MD5 is that it ‘‘adds each of the resultsof [the logical operations and shifts] into abucket or hash,’’ and that addition is one ofthe ‘‘two primary kinds of operations [thatMD5 performs],’’ and the second is leftshifting, which he equated to multiplica-tion, which he testified ‘‘is nothing morethan addition done over and over again.’’To be sure, Microsoft’s Dr. Wallach disa-greed with Klausner’s testimony, choosing

to focus on the logical functions, which hecalled the ‘‘heart and soul that makes MD5what it is,’’ and the circular shifter, both ofwhich he opined were not ‘‘summation.’’However, Microsoft has not explained whyall the steps of an algorithm must be sum-mation steps in order for the algorithm toqualify as a summation algorithm. Thejury could reasonably have believed thatMD5 is a summation algorithm. As thiscourt noted in IMS Tech., ‘‘though twostructures arguably would not be consid-ered equivalent structures in other con-texts, e.g., if performing functions otherthan the claimed function,’’ they may nev-ertheless be equivalent under § 112 ¶ 6when performing the same function. 206F.3d at 1436. Here, the claimed functionis the generation of a licensee unique ID,see infra section I.C, and if as Klausnertestified, MD5 uses addition to performthis function, the enhanced functionality ofMD5 in making the output more secureshould not prevent it from being consid-ered an equivalent structure. Microsoft’sargument that because MD5 is irreversibleit cannot be a summation algorithm—suchthat even if ‘‘you know the output of thealgorithm, it is impossible even to guessany one input that would create the out-put,’’ Br. of Microsoft at 21—is likewiseunconvincing, because the same is true ofthe most basic simple addition algorithm(e.g., it is impossible to identify the twonumbers whose sum is 23).

[5, 6] Third, the district court improp-erly rejected Klausner’s testimony as ‘‘in-complete, oversimplified and frankly inap-propriate,’’ justifying its rejection byKlausner’s failure to discuss hashing, sum-mation, or left-shifting in his expert report,and his analogizing of an MD5 digest to aReader’s Digest book. Uniloc II, 640F.Supp.2d at 170 n. 21. In common withthe other circuits, First Circuit law doesnot allow the district court in a jury trial to


evaluate ‘‘the credibility of witnesses, re-solve conflicts in testimony, or evaluate theweight of the evidence.’’ Gibson, 37 F.3dat 735. Under Daubert, the district courtmust exercise its ‘‘gatekeeper’’ function inensuring that scientific testimony is rele-vant and reliable. Kumho Tire Co. v.Carmichael, 526 U.S. 137, 137, 119 S.Ct.1167, 143 L.Ed.2d 238 (1999) (discussingDaubert v. Merrell Dow Pharm., Inc., 509U.S. 579, 589, 113 S.Ct. 2786, 125 L.Ed.2d469 (1993)). Here, the district court ex-plicitly noted that Klausner was ‘‘quali-fied.’’ Uniloc II, 640 F.Supp.2d at 172 n.25. It is decidedly the jury’s role to evalu-ate the weight to be given to the testimonyof dueling qualified experts. i4i Ltd.P’ship v. Microsoft Corp., 598 F.3d 831,856 (Fed.Cir.2010), cert. granted, 562 U.S.––––, 131 S.Ct. 647, 178 L.Ed.2d 476 (2010)(‘‘[I]t is not the district court’s role underDaubert to evaluate the correctness offacts underlying an expert’s testimony.’’).The district court’s criticism of Klausner’suse of the analogy of a digest to a ‘‘read-er’s digest’’ is also improper, because Mi-crosoft did not object at trial and has usedthe same analogy in describing the outputof SHA1 as a ‘‘hash digest, where digestindicates a shortened size, similar to Read-er’s Digest condensed books.’’

Klausner’s testimony was certainly asimplification of the functioning of MD5,but neither the district court nor Microsoftdemonstrate why it was ‘‘oversimplified,’’Uniloc II, 640 F.Supp.2d at 171 n. 21, oreven why it was inaccurate. Klausner rec-ognized that summation was not all thatMD5 did, but opined that it was ‘‘a signifi-cant portion of the MD5 algorithm.’’

Finally, Microsoft’s attacks on the docu-mentary evidence presented by Uniloc areunwarranted. In particular, the contem-poraneous Microsoft documents that definethe output of MD5 as a ‘‘checksum,’’ or‘‘cryptographic checksum,’’ or indicate that

MD5 is indicative of a summation algor-ithm, 8432 patent col. 9 ll.50–57, help toassociate the MD5 procedure within thereasonable bounds of the word ‘‘summa-tion.’’ It is reasonable to consider MD5 asummation algorithm where those skilledin the art refer to its output as a ‘‘hash-sum ’’ or an ‘‘MD5 sum.’’ Dr. Wallach hadthe opportunity to respond at trial. Forexample, in discussing Microsoft’s Tech-Net document, which included an entry,‘‘Hashing. Hashing is also referred to asMD5 checksum,’’ Dr. Wallach and Uniloc’sattorney engaged in the following colloquy:

Q. Thank you. And so, then, appar-ently, you disagree that it would be fairto say that a—that an MD5 is a check-sum?A. MD5 is a cryptographic checksum.It’s a specialized kind of checksum.Q. No, no, I didn’t ask that question.A. Yes, you did.Q. I said would you agree that it wouldbe fair to refer to the MD5 as just achecksum, as specifically done in Micro-soft’s document here?A. Computer scientists would call MD5a cryptographic checksum to distinguishit from other kinds of checksums.

Microsoft has failed to show why a reason-able jury could not have rejected this dis-tinction. Moreover, the district court’s re-sponse to these documents is puzzling.The district court acknowledged that‘‘[s]ome of these documents no doubt saythat MD5 and SHA–1 are a type of hash,or checksum,’’ but noted that the docu-ments did not show what ‘‘the complexhashes in this case actually do,’’ Uniloc II,640 F.Supp.2d at 172. However, MD5 andSHA–1 are the complex hashes in thiscase.

For the above reasons, this court con-cludes that a reasonable jury could rely onKlausner’s testimony and the documentaryevidence to conclude that MD5 and SHA1


were ‘‘summation algorithm[s]’’ as thatphrase is used in the context of the 8216patent.

2. ‘‘Registration System’’ and‘‘Mode Switching Means’’

In the alternative, the district courtgranted JMOL of non-infringement be-cause of Product Activation’s failure toincorporate a ‘‘registration system’’ or‘‘mode switching means’’ using the follow-ing constructions:‘‘Mode switching Means plus function. ‘‘Func-means’’ tion: to permit the digital data

or software to run in a usemode if the locally generatedlicensee unique ID matcheswith the remotely generated li-censee unique ID.’’ and ‘‘Struc-ture: program code which per-forms a comparison of twonumbers or a comparator andequivalents thereof.’’ Uniloc IClaim Construction, 447F.Supp.2d at 198 (synonyms inthe claim construction not rele-vant to the claim at issue areremoved for simplicity).

‘‘Registration ‘‘A system that allows digitalsystem’’ data or software to run in a

use mode on a platform if andonly if an appropriate licensingprocedure has been followed.’’Id. at 202.

‘‘Use mode’’ ‘‘A mode that allows full use ofthe digital data or software inaccordance with the license.’’Id. at 196.

These constructions are undisputed on ap-peal. It is also undisputed that the rele-vant ‘‘license’’ in the definition of ‘‘usemode’’ is the EULA, to which the useragrees prior to initiation of Product Acti-vation in the accused product. The crux ofthe question is whether the use of theaccused products before Product Activa-tion constitutes full use in accordance withthe EULA.

Microsoft argues that the legal licensingoccurred at the time the EULA was ac-cepted by the user, and that whatever usethis permitted was full use in accordancewith the license. In other words, becausethe terms of the EULA only give the user

the right to use the accused products withcertain temporal and functional restric-tions, such restricted use is ‘‘full use’’ un-der the terms of the EULA, and ‘‘fulfill[s]the seller’s/licensor’s obligations in relationto the sale or license of the right to exe-cute the digital data or software in the usemode.’’ 8216 patent col. 2 ll.42–44. Micro-soft contrasts its system with that dis-closed in the 8216 patent, which it says islimited to systems in which legal licensingand registration occur concurrently.

The district court agreed with Microsoft,holding that once the user agrees to theEULA, ‘‘the user becomes a licensee, andcan use the software in accordance withthe terms of the license, and with theprovided functionalityTTTT Activation itselfsimply opens additional doors which werepreviously locked to the licensee.’’ UnilocII, 640 F.Supp.2d at 175–77.

[7] Microsoft’s argument ultimatelyfails because it rests on the false factualpremise that the functionality during the‘‘grace period’’ between the EULA andProduct Activation satisfies Microsoft’s ob-ligations under the EULA. This factualpremise is false for three reasons. First,the EULA accompanying Microsoft Officestates: ‘‘Mandatory Activation. You maynot be able to exercise Your rights to theSoftware Product under this EULA after afinite number of product launches unlessYou activate Your copy of the SoftwareProduct in the manner described duringthe launch sequence.’’ This sentence indi-cates that ‘‘rights TTT under this EULA’’are restricted unless the product is activat-ed, and do not encompass some abstractright to full functionality. Consistently,the Windows EULA, in a clause discussing‘‘Mandatory Activation’’ notes that ‘‘[t]helicense rights granted under this EULAare limited to the first thirty (30) daysafter you first install the Product unlessyou supply information required to acti-


vate your licensed copy.’’ These ‘‘licenserights granted under this EULA’’ arerights that had already been defined bythe EULA without temporal or functionalrestrictions: ‘‘You may install, use, access,display and run one copy of the Product ona single computer,’’ subject only to limita-tions on the number of processors andcomputers that may use the program.Second, both the Windows license (‘‘Micro-soft grants you the following rights provid-ed that you comply with all terms andconditions of this EULA’’) and the Officelicense (‘‘The license rights described inthis Section are subject to all other termsand conditions of this EULA’’) are condi-tional. Both also note that Product Acti-vation is ‘‘[m]andatory.’’ Thus, unless theuser activates the product, she is not enti-tled to the rights granted by the EULA.Finally, Klausner testified that unless theaccused products are activated, they can-not receive product updates or upgrades.However, the Windows EULA allows auser to install ‘‘updates, supplements, add-on components, or Internet-based servicescomponents, of the Product that Microsoftmay provide to you or make available toyou after the date you obtain your initialcopy of the Product.’’ Thus, in order tohave ‘‘full use TTT in accordance with thelicense,’’ the user must have access tothese upgrades. This only occurs uponactivation.

This court thus concludes that use dur-ing the ‘‘grace period’’ after agreement tothe EULA and before Product Activationin the accused product does not constitutefull use in accordance with the EULA. Itis undisputed that Product Activation liftsall the grace period restrictions if and onlyif the information entered indicates a legit-imate copy of Office or Windows.

Moreover, the 8216 patent is not limitedto the situation where activation and li-censing are concurrent. In the preferredembodiment shown in Figures 2a-c, the

registration system requires the user toview the license and to ‘‘continue’’ with theregistration, far upstream of the activation.Until the user inputs confirmed paymentdetails and plugs in a valid registrationnumber, only the ‘‘demo version’’ of thesoftware will run. Once the user performsthese steps, the registration systemswitches the software into the ‘‘full ver-sion.’’ 8216 patent, Figs. 2a-c.

For the above reasons, the jury hadsubstantial evidence to find that Micro-soft’s Product Activation included a ‘‘regis-tration system’’ and ‘‘mode switchingmeans,’’ and thus the district court erredin granting JMOL of non-infringement onthe basis of this limitation.

3. Alternative Ground for Affirmance:Licensee Unique ID

Microsoft also argues as an alternativeground for affirmance of JMOL that theoutput of its MD5 and SHA1 algorithmswas not a licensee unique ID as requiredby claim 19 of the 8216 patent because itwas not ‘‘associated with a licensee.’’ Thisargument was sufficiently addressed anddecided against Microsoft by this court inthe prior appeal, and is thus law of thecase here. Uniloc I, 290 Fed.Appx. at 345(reversing summary judgment of non-in-fringement and holding that the output ofProduct Activation ‘‘generate[s] whatmight qualify as a licensee unique ID, thehash value’’).

4. Alternative Ground for Affirmance:Direct Infringement

Microsoft presents an alternativeground for affirmance of JMOL of non-infringement, on the basis that Unilocfailed to prove direct infringement becauseMicrosoft did not supply or use the end-users’ computers that implemented the lo-cal licensee unique ID generating meansand mode switching means. Microsoft re-


lies primarily on Cross Medical Products,Inc. v. Medtronic Sofamor Danek, Inc.,424 F.3d 1293 (Fed.Cir.2005), and a line ofcases including Muniauction, Inc. v.Thomson Corp., 532 F.3d 1318 (Fed.Cir.2008) and BMC Res., Inc. v. Paymentech,L.P., 498 F.3d 1373 (Fed.Cir.2007). Thedistrict court rejected this argument in itsJMOL opinion.

Microsoft’s argument is severely ham-pered by the language of claim 19. Claim19 is directed to ‘‘A remote registrationstation incorporating remote licenseeunique ID generating means, said stationforming part of a registration system TTT

including local licensee unique ID generat-ing meansTTTT’’ 8216 patent col.15 ll.21–26. As we noted in BMC, ‘‘[a] patenteecan usually structure a claim to captureinfringement by a single party,’’ by ‘‘fo-cus[ing] on one entity.’’ 498 F.3d at 1381.This is exactly what Uniloc did in claim 19,which focuses exclusively on the ‘‘remoteregistration station,’’ and defines the envi-ronment in which that registration stationmust function. It cannot be disputed thatduring each Product Activation, Microsoft‘‘uses’’ a ‘‘remote registration station’’ thatincorporates a ‘‘remote licensee unique IDgenerating means,’’ and this station formspart of a ‘‘registration system’’ that alsoincludes a ‘‘local licensee unique ID gener-ating means’’ and a ‘‘mode switchingmeans.’’ That other parties are necessaryto complete the environment in which theclaimed element functions does not neces-sarily divide the infringement between thenecessary parties. For example, a claimthat reads ‘‘An algorithm incorporatingmeans for receiving e-mails’’ may requiretwo parties to function, but could neverthe-less be infringed by the single party whouses an algorithm that receives e-mails.

The claim here is thus distinguishablefrom those at issue in Muniauction andBMC, because here, only one party, Micro-soft, makes or uses the remote registration

station. See Muniauction, 532 F.3d at1329; BMC, 498 F.3d at 1373. Nor isclaim 19 analogous to the claim at issue inCross Medical. There, the claim called for‘‘[a] fixation device comprising TTT an an-chor seat means which has a lower boneinterface operatively joined to said bonesegment.’’ 424 F.3d at 1299 (citing U.S.Patent No. 5,474,555 col. 8 ll.33–41). Thiscourt construed ‘‘operatively joined’’ tomean that the interface and the bone mustbe in contact, id. at 1305, and held thatMedtronic did not infringe the claims be-cause ‘‘Medtronic does not itself make anapparatus with the ‘interface’ portion incontact with bone,’’ id. at 1311. Here,however, Microsoft does make and use theremote registration station in the environ-ment required by the claims, when theMD5 and SHA1 generate a remote licen-see unique ID. Moreover, this courtagrees with the district court that ‘‘[a]c-cepting Microsoft’s argument that the localside of Claim 19 requires an end-user’sparticipation, similar to the surgeons’ par-ticipation in Cross Medical, would be akinto importing a method step into this soft-ware system—something the language ofClaim 19 does not support.’’ Uniloc II,640 F.Supp.2d at 162.

5. New Trial on Infringement Issues

Citing the closeness of the questionspresented on JMOL in this case, the dis-trict court also granted in the alternativeMicrosoft’s motion for a new trial on in-fringement.

[8, 9] This court’s standard of reviewover a district court’s grant of a motion fornew trial is governed by regional circuitlaw. WMS Gaming, Inc. v. Int’l GameTech., 184 F.3d 1339, 1361 (Fed.Cir.1999).In the First Circuit, a new trial is onlyappropriate when ‘‘the outcome is againstthe clear weight of the evidence such thatupholding the verdict will result in a mis-


carriage of justice.’’ Ramos v. Davis &Geck, Inc., 167 F.3d 727, 731 (1st Cir.1999).In contrast to JMOL, in considering amotion for a new trial, the district courtmay ‘‘independently weigh the evidence.’’Jennings v. Jones, 587 F.3d 430, 436 (1stCir.2009).

[10] ‘‘[A] district judge cannot dis-place a jury’s verdict merely because hedisagrees with it or because a contraryverdict may have been equally support-able. As we have repeatedly observed,trial judges do not sit as thirteenth jurors,empowered to reject any verdict withwhich they disagree.’’ Id. (internal cita-tion omitted). Nevertheless, the districtcourt is entitled to deference in granting anew trial motion, and the First Circuitonly overturns the grant of a new trial ifthe district court has abused its discretion.Id. at 435 (citing Gasperini v. Ctr. ForHumanities, Inc., 518 U.S. 415, 435, 116S.Ct. 2211, 135 L.Ed.2d 659 (1996)). Asthe Supreme Court noted, ‘‘[t]rial judgeshave the unique opportunity to considerthe evidence in the living courtroom con-text, while appellate judges see only thecold paper record.’’ Gasperini, 518 U.S.at 438, 116 S.Ct. 2211.

The district court granted Microsoft’smotion for a new trial on the infringementissues in the alternative to its JMOL mo-tion, and did not present any analysisapart from its analysis of the JMOL in-fringement issues discussed above. Thiscourt is convinced that the district court’sgrant of a new trial on infringement has nomore merit than the district court’s grantof JMOL on infringement. Though it is aclose issue, this is not a situation wherethe evidence falls within the zone wheresubstantial evidence supports the verdictand the district court’s discretion in grant-ing a new trial trumps such evidence.This court thus reverses the districtcourt’s grant of a new trial on infringe-

ment for the same reasons as it reversesthe grant of JMOL of non-infringement.

B. JMOL and New Trial for Willfulness

[11] ‘‘[T]o establish willful infringe-ment, a patentee must show by clear andconvincing evidence that the infringer act-ed despite an objectively high likelihoodthat its actions constituted infringement ofa valid patent.’’ In re Seagate Tech., LLC,497 F.3d 1360, 1371 (Fed.Cir.2007) (enbanc). This is an objective inquiry. Id. Inaddition, a patentee must show that thisrisk ‘‘was either known or so obvious thatit should have been known to the accusedinfringer.’’ Id. This is a subjective inquiry.

The district court concluded that no rea-sonable jury could have found that Micro-soft’s conduct fell under either Seagateprong. Uniloc II, 640 F.Supp.2d at 176–77 (objective prong), 177–79 (subjectiveprong). If the accused infringer’s positionis susceptible to a reasonable conclusion ofno infringement, the first prong of Seagatecannot be met. See Cohesive Techs., Inc.v. Waters Corp., 543 F.3d 1351, 1374 (Fed.Cir.2008) (‘‘Because ‘rigid’ was susceptibleto a reasonable construction under whichWaters’s products did not infringe, therewas not an objectively high likelihood thatWaters’s actions constituted infringe-ment.’’).

[12] Uniloc has failed to meet thethreshold objective prong of Seagate. Un-iloc has not presented any evidence attrial or on appeal showing why Microsoft,at the time it began infringement, couldnot have reasonably determined that MD5and SHA1 did not meet the ‘‘licenseeunique ID generating means,’’ ‘‘licenseeunique ID,’’ or ‘‘registration sys-tem’’/‘‘mode switching means’’ limitations.Specifically, infringement of the ‘‘licenseeunique ID generating means’’ limitation isa complicated issue, made more so because‘‘equivalence requires an intensely factual


inquiry,’’ DePuy Spine, Inc. v. MedtronicSofamor Danek, Inc., 567 F.3d 1314, 1337(Fed.Cir.2009). Uniloc’s argument aboutcopying is largely inapposite. See id. at1336 (‘‘[E]vidence of copying in a case ofdirect infringement is relevant only to Sea-gate’s second prong.’’). As the districtcourt noted, the facts here presented are‘‘hardly the stuff of which objectively reck-less unreasonable conduct is made.’’ Uni-loc II, 640 F.Supp.2d at 177.

Given this court’s conclusion that Unilocfailed to show that a reasonable jury couldfind Microsoft’s conduct objectively reck-less on the evidence presented, this courtneed not address the subjective prong ofSeagate. This court thus affirms the dis-trict court’s grant of JMOL of no willful-ness, and need not address the districtcourt’s alternative grant of a new trial onwillfulness.

C. New Trial on Damages

The jury here awarded Uniloc $388 mil-lion, based on the testimony of Uniloc’sexpert, Dr. Gemini. Dr. Gemini opinedthat damages should be $564,946,803.This was based on a hypothetical negotia-tion between Uniloc and Microsoft and theGeorgia–Pacific factors. See Georgia–Pa-cific Corp. v. U.S. Plywood Corp., 318F.Supp. 1116 (S.D.N.Y.1970). Gemini be-gan with an internal pre-litigation Micro-soft document that stated:

Product Keys are valuable for two majorreasons. First, since Product Keys canbe used to install a product and create avalid Product ID, you can associate amonetary value to them. An appraisalprocess found that a Product Key isworth anywhere between $10 and$10,000 depending on usage. Secondly,Product Keys contain short digital sig-nature technology that Microsoft Re-search created. For these reasons, it iscrucial that Product Keys are handledwith maximum security.

In Limine, 632 F.Supp.2d at 150 n. 2.Gemini took the lowest value, $10, andtestified that this is ‘‘the isolated value ofProduct Activation.’’ Gemini then appliedthe so-called ‘‘25 percent rule of thumb,’’hypothesizing that 25% of the value of theproduct would go to the patent owner andthe other 75% would remain with Micro-soft, resulting in a baseline royalty rate of$2.50 per license issued. Gemini justifiedthe use of the rule of thumb because it has‘‘been accepted by Courts as an appropri-ate methodology in determining damages,in [his] experience, in other cases.’’ Hethen considered several of the Georgia–Pacific factors, with the idea being ‘‘toadjust this 25% up or down depending onhow [the Georgia–Pacific factors] favor[ ]either party.’’ At bottom, he concludedthat the factors in favor of Uniloc andMicrosoft generally balanced out and didnot change the royalty rate. He thenmultiplied the $2.50 royalty rate by thenumber of new licenses to Office and Win-dows products, 225,978,721, to get a finalreasonable royalty of $564,946,803. Gemi-ni then ‘‘did kind of a check to determinewhether that number was reasonable. It’sobviously, you know, a significant amountof money. I wanted to check to make sureit was a reasonable number.’’ The ‘‘check’’was performed by ‘‘estimating the grossrevenues for the accused products’’ bymultiplying the 225,978,721 licenses by theaverage sales price per license of $85.The resulting gross revenue value was$19.28 billion. Gemini then calculated thathis damages calculation resulted in a royal-ty rate over the gross revenue of Officeand Windows of approximately 2.9%.Gemini presented this information in a de-monstrative pie chart to accompany histestimony. In response to Uniloc’s attor-ney’s question: ‘‘And have you prepared achart or a graph or a pie chart to show usthis comparison?’’ Uniloc’s attorney, Mr.Cronin stated, ‘‘Your honor, there’s no ob-


jection,’’ and Microsoft attorney Mr.Scherkenbach stated, ‘‘Right, there is noobjection.’’ Gemini then opined that ‘‘inmy experience, and data I’ve seen as far asindustry royalty rates for software, whichare generally above—on average, above10% or 10, 11%, I felt that this royalty wasreasonable and well within that range.’’

Microsoft had challenged the 25% rulein limine and attempted to exclude Mr.Gemini’s testimony. The district courtnoted that ‘‘the concept of a ‘rule of thumb’is perplexing in an area of the law wherereliability and precision are deemed para-mount,’’ but rejected Microsoft’s positionbecause the rule has been widely accepted.The district court thus considered the useof the rule of thumb to be reasonable. InLimine, 632 F.Supp.2d at 151. Microsoftcontested Gemini’s use of the entire mar-ket value rule ‘‘check’’ because ProductActivation was not the basis of the con-sumer demand for Microsoft’s Office andWindows products. The district courtagreed with Microsoft, and granted a newtrial on damages, because the ‘‘$19 billioncat was never put back into the bag’’ andthe jury may have ‘‘used the $19 billionfigure to ‘check’ its significant award of$388,000,000.’’ Uniloc II, 640 F.Supp.2dat 185.

On appeal, the parties present the courtwith three damages issues: 1) the propri-ety of using the 25 percent rule; 2) appli-cation of the entire market value rule as a‘‘check’’; and 3) excessiveness of damages.Because this court affirms the districtcourt’s conditional grant of a new trial ondamages, this court need not reach the lastissue.

1. 25 Percent Rule

[13] Section 284 of Title 35 of the Unit-ed States Code provides that on findinginfringement of a valid patent, damagesshall ‘‘in no event [be] less than a reason-able royalty for the use made of the inven-

tion by the infringer, together with inter-est and costs as fixed by the court.’’ Inlitigation, a reasonable royalty is often de-termined on the basis of a hypotheticalnegotiation, occurring between the partiesat the time that infringement began.Wang Labs. Inc. v. Toshiba Corp., 993F.2d 858, 869–70 (Fed.Cir.1993). A rea-sonable royalty is the predominant meas-ure of damages in patent infringementcases. William C. Rooklidge and MarthaK. Gooding, When Hypothetical Turns toFantasy: The Patent Reasonable RoyaltyHypothetical Negotiation, BNA InsightsVol. 80:1983, at 701 n. 10 (‘‘HypotheticalNegotiation ’’) (citing PriceWaterhouse-Coopers, A Closer Look: Patent LitigationTrends and the Increasing Impact ofNonpracticing Entities at 5 (2009)).

The 25 percent rule of thumb is a toolthat has been used to approximate thereasonable royalty rate that the manufac-turer of a patented product would be will-ing to offer to pay to the patentee during ahypothetical negotiation. Robert Gold-scheider, John Jarosz and Carla Mulhern,Use Of The 25 Per Cent Rule in ValuingIP, 37 les Nouvelles 123, 123 (Dec. 2002)(‘‘Valuing IP ’’). ‘‘The Rule suggests thatthe licensee pay a royalty rate equivalentto 25 per cent of its expected profits forthe product that incorporates the IP atissue.’’ Id. As explained by its leadingproponent, Robert Goldscheider, the ruletakes the following form:

An estimate is made of the licensee’sexpected profits for the product thatembodies the IP at issue. Those profitsare divided by the expected net salesover that same period to arrive at aprofit rate. That resulting profit rate,say 16 per cent, is then multiplied by 25per cent to arrive at a running royaltyrate. In this example, the resulting roy-alty rate would be 4 per cent. Goingforward (or calculating backwards, inthe case of litigation), the 4 per cent


royalty rate is applied to net sales toarrive at royalty payments due to the IPowner.

Id. at 124. The underlying ‘‘assumption isthat the licensee should retain a majority(i.e. 75 percent) of the profits, because ithas undertaken substantial development,operational and commercialization risks,contributed other technology/IP and/orbrought to bear its own development, op-erational and commercialization contribu-tions.’’ Id.

The rule was originally based on Gold-scheider’s observations of commercial li-censes entered into by a ‘‘Swiss subsidiaryof a large American company, with 18licensees around the world, each having anexclusive territory.’’ Id. The rights trans-ferred were a portfolio of patents and oth-er intellectual property apparently relatedto the patented products. Id. The term ofeach of these licenses was for three years,with the expectation that the licenseswould be renewed. Id. at 123. The licen-sees ‘‘faced strong competition,’’ and ‘‘wereeither first or second in sales volume, andprobably profitability, in their respectivemarket.’’ Id.

According to its proponents, the veracityof the 25 percent rule has been ‘‘confirmedby a careful examination of years of licens-ing and profit data, across companies andindustries.’’ John C. Jarosz, Carla S. Mul-hern and Michael Wagner, The 25% RuleLives On, IP Law360, Sept. 8, 2010. Gold-scheider published a further empiricalstudy in 2002, concluding that across allindustries, the median royalty rate was22.6 percent, and that the data supportedthe use of the 25 percent rule ‘‘as a tool ofanalysis.’’ Valuing IP, 37 les Nouvelles at132–33. Additionally, in a 1997 study oflicensing organizations, 25 percent of theorganizations indicated that they use the25 percent rule as a starting point in nego-tiations. Stephen A. Degnan & Corwin

Horton, A Survey of Licensed Royalties,32 les Nouvelles 91, 95 (June 1997).

The 25 percent rule has, however, metits share of criticism that can be broadlyseparated into three categories. First, itfails to account for the unique relationshipbetween the patent and the accused prod-uct. See Gregory K. Leonard and LaurenJ. Stiroh, Economic Approaches to Intel-lectual Property Policy, Litigation, andManagement, 949 PLI/Pat 425, 454–55(Sept.–Nov. 2008) (‘‘[The 25 percent rule]takes no account of the importance of thepatent to the profits of the product sold,the potential availability of close substi-tutes or equally noninfringing alternatives,or any of the other idiosyncrasies of thepatent at issue that would have affected areal-world negotiation.’’); Richard S. Toik-ka, Patent Licensing Under Competitiveand Non–Competitive Conditions, 82 J.Pat. & Trademark Off. Soc’y 279, 292–93(Apr. 2000) (arguing that it fails to ‘‘distin-guish between monopoly and normal prof-itTTTT Thus for narrow patents, the rulemay be overly generous to the patentee,and for broad patents it may be overlystingy’’). Second, it fails to account forthe unique relationship between the par-ties. See Ted Hagelin, Valuation of Pat-ent Licenses, Tex. Intell. Prop. L.J. 423,425–26 (Spring 2004) (noting that the ruleshould not be used in isolation because itfails to ‘‘account[ ] for the different levelsof risk assumed by a licensor and licen-see’’); Hypothetical Negotiations at 702(‘‘[T]he rule is unlikely to have any basisin the accused infringer’s industry, in thetechnology involved in either the patent orthe accused product or service, or in theclaimed invention’s contribution to the in-fringing product or service.’’). Finally, therule is essentially arbitrary and does notfit within the model of the hypotheticalnegotiation within which it is based. SeeRoy J. Epstein and Alan J. Marcus, Eco-nomic Analysis of the Reasonable Royal-


ty: Simplification and Extension of theGeorgia–Pacific Factors, 85 J. Pat. &Trademark Off. Soc’y 55, 574 (July 2003)(‘‘[The 25% and the 5%] rules of thumbare best understood as special cases [ ]that may be appropriate to a given situa-tion only by chance.’’); Roy J. Epstein,Modeling Patent Damages: Rigorous andDefensible Calculations (2003) (paper pre-sented at the AIPLA 2003 Annual Meet-ing) at 22 available at http://www.royepstein.com/epstein aipla 2003 articlewebsite.pdf (last accessed Nov. 19, 2010)(arguing that the 25% rule ‘‘shortcut’’ ‘‘isessentially arbitrary. Because it is basedon ex post results, it does not necessarilyrelate to the results of a negotiation thattook place prior to the infringement’’).

The admissibility of the bare 25 percentrule has never been squarely presented tothis court. Nevertheless, this court haspassively tolerated its use where its ac-ceptability has not been the focus of thecase, see e.g., i4i Ltd., 598 F.3d 831; Fo-nar Corp. v. General Elec. Co., 107 F.3d1543, 1553 (Fed.Cir.1997), or where theparties disputed only the percentage to beapplied (i.e. one-quarter to one-third), butagreed as to the rule’s appropriateness,Finjan, Inc. v. Secure Computing Corp.,626 F.3d 1197, 1210–11(Fed.Cir.2010).Lower courts have invariably admitted evi-dence based on the 25% rule, largely inreliance on its widespread acceptance orbecause its admissibility was uncontested.See In Limine, 632 F.Supp.2d at 151 (‘‘The‘25% Rule’ has been accepted as a properbaseline from which to start [a royalty]analysis.’’ (internal citations omitted));GSI Grp., Inc. v. Sukup Mfg., Co., 641F.Supp.2d 732, 745 (C.D.Ill.2008) (same);i4i Ltd. P’ship v. Microsoft Corp., 670F.Supp.2d 568, 592 (E.D.Tex.2009), aff’don other grounds by 598 F.3d 831 (‘‘[i4i’sexpert] testified that it was customarywithin his field to apply a ‘25% rule ofthumb’TTTT Thus, considering the founda-tion laid by [i4i’s expert’s] testimony, his

application of the 25% rule was relevantand appropriate considered.’’); Static Con-trol Components, Inc. v. Lexmark Int’l,Inc., Nos. 5:02–571, 5:04–84, 2007 WL7083655 at *13–14 (E.D.Ky. May 12, 2007)(‘‘While Lexmark does not believe the ‘ruleof thumb’ approach is the most appropri-ate way to calculate ‘reasonable royalty,’ asSCC correctly notes, case law suggests itis one way of doing so’’ (citing StandardMfg. Co. v. United States, 42 Fed.Cl. 748,766 (1999))); Novozymes A/S v. GenencorInt’l, Inc., 474 F.Supp.2d 592, 606 (D.Del.2007) (‘‘While there is no particular analyt-ical justification for [the rule of thumb], ithas been used to estimate royalties.’’); In-line Connection Corp. v. AOL Time War-ner Inc., 470 F.Supp.2d 424, 432 n. 38(D.Del.2007) (allowing 25% rule because itsuse was not disputed); Bose Corp. v. JBL,Inc., 112 F.Supp.2d 138, 167 (D.Mass.2000)(‘‘Courts have found the 25%/75% ap-proach to be a useful approach to arrivingat a baseline royalty rateTTTT [The oppos-ing expert] conceded that this approach isa common and reasonable one, though hehas never used that approach in negotiat-ing licenses’’ (citing Standard Mfg., 42Fed.Cl. at 764)); Standard Mfg., 42 Fed.Cl. at 766 (‘‘[T]he 25% rule or a closevariant of it has been recognized by anumber of other federal courts as a ‘rule ofthumb’ or ‘typical’ in the licensing field.’’);Procter & Gamble Co. v. Paragon TradeBrands, Inc., 989 F.Supp. 547, 612 (D.Del.1997) (‘‘Although the Court will considerthe Rule–of–Thumb analysis in determin-ing the royalty rate, this approach will notreceive substantial weight.’’); Secure En-ergy, Inc. v. Coal Synthetics, LLC, No.4:08–CV–1719, 2010 WL 1692076 at *1(E.D.Mo. Apr.27, 2010) (‘‘The parties agreethat application of the 25% ’rule of thumb’is acceptable to determine a reasonablyroyalty case such as this.’’). See also Pa-ice LLC v. Toyota Motor Corp., 609F.Supp.2d 620, 629–30 (E.D.Tex.2009) (ap-


plying 25% rule without discussion); EZDock, Inc. v. Schafer Sys., Inc., No. 98–2364, 2003 WL 1610781 (D.Minn. Mar.8,2003) (same). In at least one case, thedistrict court admitted the evidence, butrefused to give it substantial weight be-cause, ‘‘neither expert testified as to thecustomary profit percentage used to setthe royalty rates in licenses in other busi-nesses’’ and because ‘‘[t]here was no testi-mony advocating the use of the [sic ] thisapproach as an appropriate guidepost forthe determination of a royalty rate under aGeorgia–Pacific analysis.’’ Procter &Gamble Co. v. Paragon Trade Brands,Inc., 989 F.Supp. at 612.

In Daubert, 509 U.S. at 589, 113 S.Ct.2786 and Kumho Tire, 526 U.S. 137, 119S.Ct. 1167, the Supreme Court assigned tothe district courts the responsibility of en-suring that all expert testimony must per-tain to ‘‘scientific, technical, or other spe-cialized knowledge’’ under Federal Rule ofEvidence (‘‘FRE’’) 702, which in turn re-quired the judge to determine that thetestimony was based on a firm scientific ortechnical grounding. Daubert, 509 U.S. at589–90, 113 S.Ct. 2786; Kumho Tire, 526U.S. at 148, 119 S.Ct. 1167. ‘‘Expert testi-mony which does not relate to any issue inthe case is not relevant and, ergo, non-helpful.’’ Daubert, 509 U.S. at 591, 113S.Ct. 2786 (citing 3 Weinstein & Berger¶ 702[02], p. 702–18).

[14] This court now holds as a matterof Federal Circuit law that the 25 percentrule of thumb is a fundamentally flawedtool for determining a baseline royalty ratein a hypothetical negotiation. Evidencerelying on the 25 percent rule of thumb isthus inadmissible under Daubert and theFederal Rules of Evidence, because it failsto tie a reasonable royalty base to the factsof the case at issue.

[15] The patentee bears the burden ofproving damages. Lucent Techs., Inc. v.Gateway, Inc., 580 F.3d 1301, 1324 (Fed.

Cir.2009). To properly carry this burden,the patentee must ‘‘sufficiently [tie the ex-pert testimony on damages] to the facts ofthe case.’’ Daubert, 509 U.S. at 591, 113S.Ct. 2786 (‘‘An additional considerationunder Rule 702—and another aspect ofrelevancy—is whether expert testimonyproffered in the case is sufficiently tied tothe facts of the case that it will aid thejury in resolving a factual dispute.’’) (citingUnited States v. Downing, 753 F.2d 1224,1242 (3d Cir.1985)). If the patentee failsto tie the theory to the facts of the case,the testimony must be excluded. For ex-ample, in General Electric Co. v. Joiner,522 U.S. 136, 118 S.Ct. 512, 139 L.Ed.2d508 (1997), the Supreme Court allowed theexclusion of eight of Joiner’s experts whoopined that polychlorinated biphenyls(‘‘PCBs’’) could cause cancer on thestrength of several studies showing thatmice receiving high doses of PCB devel-oped cancer. The Supreme Court notedthat ‘‘[t]he studies were so dissimilar tothe facts presented in this litigation that itwas not an abuse of discretion for theDistrict Court to have rejected the ex-perts’ reliance on them,’’ id. at 144–45, 118S.Ct. 512, and affirmed the exclusion be-cause Joiner had failed to tie the experts’opinions to the ‘‘seemingly far-removed an-imal studies,’’ id. at 144, 118 S.Ct. 512.Likewise, in Kumho Tire, a products liabil-ity case arising out of a blown tire, theSupreme Court affirmed the exclusion ofan expert opinion that argued that thecause of the accident at issue was a defectin the tire, based on the expert’s visual andtactile inspection of the tire. 526 U.S. at153, 119 S.Ct. 1167. The specific issue wasnot whether the visual and tactile inspec-tion methodology was ‘‘reasonable[ ] ingeneral,’’ but whether ‘‘it was [reasonableto] us[e] such an approach TTT to draw aconclusion regarding the particular matterto which the expert testimony was directlyrelevant.’’ Id. at 153–54, 119 S.Ct. 1167.


‘‘The relevant issue was whether the ex-pert could reliably determine the cause ofthis tire’s separation.’’ Id. at 154, 119S.Ct. 1167. The Court held that the ex-pert had failed to reliably opine on thisissue under Daubert because his generaltheory—‘‘that in the absence of at leasttwo of four signs of abuse TTT he concludesthat a defect caused the separation,’’ id.—did not take into account the facts of theparticular tire at issue: that the tire ‘‘hadtraveled far enough so that some of thetread had been worn bald; it should havebeen taken out of service; it had beenrepaired (inadequately) for punctures; andit bore some of the very marks that theexpert said indicated, not a defect, butabuse through overdeflection.’’ Id. In re-sponding to the plaintiff’s argument, ‘‘thata method of tire failure analysis that em-ploys a visual/tactile inspection is a reliablemethod,’’ based on ‘‘its use by other ex-perts and to Carlson’s [the expert in thecase] long experience working for Miche-lin,’’ the Court reaffirmed that ‘‘the ques-tion before the trial court was specific, notgeneral.’’ Id. The trial court had to decidewhether this particular expert had suffi-cient specialized knowledge to assist thejurors ‘in deciding the particular issues inthe case.’ Id. at 156, 119 S.Ct. 1167. TheCourt held that he did not.

The bottom line of Kumho Tire andJoiner is that one major determinant ofwhether an expert should be excluded un-der Daubert is whether he has justified theapplication of a general theory to the factsof the case. Consistent with this conclu-sion, this court has held that ‘‘[a]ny evi-dence unrelated to the claimed inventiondoes not support compensation for in-fringement but punishes beyond the reachof the statute.’’ ResQNet.com, Inc. v.Lansa, Inc., 594 F.3d 860, 869 (Fed.Cir.2010).

In ResQNet, Lucent Technologies, 580F.3d 1301, and Wordtech Systems, Inc. v.

Integrated Networks Solutions, Inc., 609F.3d 1308 (Fed.Cir.2010), this court deter-mined that a patentee could not rely onlicense agreements that were ‘‘radicallydifferent from the hypothetical agreementunder consideration’’ to determine a rea-sonable royalty. Lucent Techs., 580 F.3dat 1327. See also ResQNet, 594 F.3d at870–72 (holding that evidence of royaltyrates from licenses without a relationshipto the claimed invention could not form thebasis of a reasonable royalty calculation).In Lucent Technologies, the patentee’s ex-pert relied in large part on ‘‘eight variedlicense agreements,’’ four of which in-volved ‘‘PC-related patents,’’ but either thespecific subject matter of the patents wasnot explained to the jury or the license was‘‘directed to a vastly different situationthan the hypothetical licensing scenario ofthe present case,’’ and four of which Lu-cent did not describe the relationship be-tween the patented technology licensedtherein and the licensee’s products. See580 F.3d at 1328–31. This court notedthat the ‘‘licenses relied on by the patenteein proving damages [must be] sufficientlycomparable to the hypothetical license atissue in suit,’’ id. at 1325, and that thepatentee’s failure to do so ‘‘weighs stronglyagainst the jury’s award’’ relying on suchnon-comparable licenses, id. at 1332. Sim-ilarly, in ResQNet, the patentee’s expert‘‘used licenses with no relationship to theclaimed invention to drive the royalty rateup to unjustified double-digit levels,’’ look-ing at licenses that did not mention thepatents and had no ‘‘other discernible linkto the claimed technology.’’ 594 F.3d at870. This court rejected the expert’s testi-mony, holding that the district court ‘‘mustconsider licenses that are commensuratewith what the defendant has appropriated.If not, a prevailing plaintiff would be freeto inflate the reasonable royalty analysiswith conveniently selected licenses withoutan economic or other link to the technolo-


gy in question.’’ Id. at 872. This courtheld that on remand, ‘‘the trial courtshould not rely on unrelated licenses toincrease the reasonable royalty rate aboverates more clearly linked to the economicdemand for the claimed technology.’’ Id.at 872–73.

Similarly, in Wordtech, the patentee ‘‘in-troduced thirteen patent licenses that itpreviously granted to third parties forrights to some or all of the patents-in-suit’’to argue to support the jury’s damagesdetermination. 609 F.3d at 1319. Thiscourt rejected eleven of the licenses be-cause they were running royalty licenses(the patentee had only asked for a lumpsum payment) and represented far lowerrates than the jury returned. Id. at 1320–21. This court rejected the remaining twolicenses (both for lump sum payments) be-cause ‘‘[n]either license describe[d] howthe parties calculated each lump sum, thelicensees’ intended products, or how manyproducts each licensee expected to pro-duce.’’ Id. at 1320.

The meaning of these cases is clear:there must be a basis in fact to associatethe royalty rates used in prior licenses tothe particular hypothetical negotiation atissue in the case. The 25 percent rule ofthumb as an abstract and largely theoreti-cal construct fails to satisfy this fundamen-tal requirement. The rule does not sayanything about a particular hypotheticalnegotiation or reasonable royalty involvingany particular technology, industry, orparty. Relying on the 25 percent rule ofthumb in a reasonable royalty calculationis far more unreliable and irrelevant thanreliance on parties’ unrelated licenses,which we rejected in ResQNet and LucentTechnologies. There, the prior licenses atleast involved the same general industryand at least some of the same parties asthe hypothetical negotiations at issue, andin Wordtech even involved licenses to thepatents in suit entered into by the paten-

tee-plaintiff. Lacking even these minimalconnections, the 25 percent rule of thumbwould predict that the same 25%/75% roy-alty split would begin royalty discussionsbetween, for example, (a) TinyCo and IBMover a strong patent portfolio of twelvepatents covering various aspects of a pio-neering hard drive, and (b) Kodak andFuji over a single patent to a tiny improve-ment in a specialty film emulsion.

It is of no moment that the 25 percentrule of thumb is offered merely as a start-ing point to which the Georgia–Pacific fac-tors are then applied to bring the rate upor down. Beginning from a fundamentallyflawed premise and adjusting it based onlegitimate considerations specific to thefacts of the case nevertheless results in afundamentally flawed conclusion. This isreflected in Lucent Technologies, in whichunrelated licenses were considered underGeorgia–Pacific factor 1, but this courtheld that the entire royalty calculation wasunsupported by substantial evidence.

[16] To be admissible, expert testimo-ny opining on a reasonable royalty ratemust ‘‘carefully tie proof of damages to theclaimed invention’s footprint in the marketplace.’’ ResQNet, 594 F.3d at 869. Thiscourt has sanctioned the use of the Geor-gia–Pacific factors to frame the reason-able royalty inquiry. Those factors prop-erly tie the reasonable royalty calculationto the facts of the hypothetical negotiationat issue. This court’s rejection of the 25percent rule of thumb is not intended tolimit the application of any of the Georgia–Pacific factors. In particular, factors 1and 2—looking at royalties paid or re-ceived in licenses for the patent in suit orin comparable licenses—and factor 12—looking at the portion of profit that may becustomarily allowed in the particular busi-ness for the use of the invention or similarinventions—remain valid and importantfactors in the determination of a reason-


able royalty rate. However, evidence pur-porting to apply to these, and any otherfactors, must be tied to the relevant factsand circumstances of the particular case atissue and the hypothetical negotiationsthat would have taken place in light ofthose facts and circumstances at the rele-vant time.

In this case, it is clear that Gemini’stestimony was based on the use of the 25%rule of thumb as an arbitrary, general rule,unrelated to the facts of this case. Whenasked the basis of his opinion that the ruleof thumb would apply here, Gemini testi-fied: ‘‘[i]t’s generally accepted. I’ve usedit. I’ve seen others use it. It’s a widelyaccepted rule.’’ Upon further questioning,Dr. Gemini revealed that he had been in-volved in only four or five non-litigationrelated negotiations, and had recom-mended the 25% rule only once in a caseinvolving a power tool. He did not testifythat the parties here had a practice ofbeginning negotiations with a 25%/75%split, or that the contribution of ProductActivation to Office and Word justifiedsuch a split. He did not base his 25percent baseline on other licenses involv-ing the patent at issue or comparable li-censes. In short, Gemini’s starting pointof a 25 percent royalty had no relation tothe facts of the case, and as such, wasarbitrary, unreliable, and irrelevant. Theuse of such a rule fails to pass musterunder Daubert and taints the jury’s dam-ages calculation.

This court thus holds that Microsoft isentitled to a new trial on damages.

2. Entire Market Value Rule

As discussed above, Gemini performed‘‘a check to determine whether’’ his$564,946,803 royalty figure was reasonableby comparing it to his calculation of Micro-soft’s approximate total revenue for Officeand Windows of $19.28 billion. Duringtrial, Gemini testified that his calculated

royalty accounted for only 2.9% of Micro-soft’s revenue, and accented his point byreference to a prepared pie chart, showingMicrosoft’s $19.28 billion in revenue with a2.9% sliver representing his calculated roy-alty rate. He concluded that 2.9% was areasonable royalty based on his experiencethat royalty rates for software are ‘‘gener-ally above—on average, above 10% or 10,11%.’’

[17] The entire market value rule al-lows a patentee to assess damages basedon the entire market value of the accusedproduct only where the patented featurecreates the ‘‘basis for customer demand’’or ‘‘substantially create[s] the value of thecomponent parts.’’ Lucent Techs., 580F.3d at 1336; Rite–Hite Corp. v. KelleyCo., 56 F.3d 1538, 1549–50 (Fed.Cir.1995).This rule is derived from Supreme Courtprecedent requiring that ‘‘the patentee TTT

must in every case give evidence tendingto separate or apportion the defendant’sprofits and the patentee’s damages be-tween the patented feature and the unpat-ented features, and such evidence must bereliable and tangible, and not conjecturalor speculative,’’ or show that ‘‘the entirevalue of the whole machine, as a market-able article, is properly and legally attrib-utable to the patented feature.’’ Garret-son v. Clark, 111 U.S. 120, 121, 4 S.Ct. 291,28 L.Ed. 371 (1884). See also LucentTechs., 580 F.3d at 1336–37 (tracing theorigins of the entire market value to sever-al Supreme Court cases including Garret-son ).

Microsoft argues that Uniloc employedthe entire market value of Office and Win-dows by virtue of Gemini’s pie chart, hiscomparison of his calculated royalty to thetotal revenue Microsoft earns through theaccused products, and Uniloc’s attorneys’belittlement of Microsoft’s expert’s royaltyfigure as representing only .0003% of totalrevenue. Microsoft argues that Uniloc’s


use of the entire market value rule was notproper because it is undisputed that Prod-uct Activation did not create the basis forcustomer demand or substantially createthe value of the component parts. Micro-soft continues that Gemini’s testimonytainted the jury’s damages deliberations,regardless of its categorization as a‘‘check.’’

Uniloc responds that: (1) Microsoft didnot object at trial and so waived any evi-dentiary argument to Gemini’s testimonyand demonstratives; (2) the entire marketvalue of the product can be used if theroyalty rate is low enough; and (3) the $19billion figure was used only as a ‘‘check,’’and the jury was instructed not to base itsdamages determination on the entire mar-ket value, an instruction it should be pre-sumed to have followed.

The district court agreed with Microsoft,and ordered a conditional new trial ondamages. It noted that ‘‘Uniloc concededcustomers do not buy Office or Windowsbecause of [Product Activation] and said itwould not base a royalty calculation on theentire market value of the products.’’ Un-iloc II, 640 F.Supp.2d at 184–85. As such,the use of the entire market value of Officeand Windows in the form of the $19 billionfigure was ‘‘irrelevant’’ and ‘‘taint[ed]’’ thejury’s damages award. Id. at 185. Thedistrict court also disagreed with Unilocthat Microsoft had waived its arguments tothe entire market value, noting that ‘‘Mi-crosoft objected specifically under the en-tire market value rule to use of a demons-trative pie chart,’’ and that ‘‘[t]he Courtpreliminarily allowed it but after hearingthe testimony instructed counsel to stayaway from the $19 billion figure.’’ Id.

[18] This court agrees with Microsoftand the district court that Uniloc’s use ofthe $19 billion ‘‘check’’ was improper underthe entire market value rule. First, re-garding Uniloc’s assertion that Microsofthas waived the issue, this court will not

second-guess the district court’s explicitrecognition of Microsoft’s objections toGemini’s testimony. FRE 103(a) notesthat ‘‘Error may not be predicated upon aruling which admits or excludes evidenceunless TTT (1) Objection.—In case the rul-ing is one admitting evidence, a timelyobjection or motion to strike appears ofrecordTTTT Once the court makes a defini-tive ruling on the record admitting or ex-cluding evidence, either at or before trial,a party need not renew an objection oroffer of proof to preserve a claim of errorfor appeal.’’ The district court here ex-plicitly noted that Microsoft’s objection fellinto the exception at the last line of FRE103(a): ‘‘Although Microsoft did not con-tinue to repeat an objection, it made itsposition on this evidence sufficiently clearto preserve the instant challenge’’ to Gemi-ni’s use of the entire market value rule.Uniloc II, 640 F.Supp.2d at 184 n. 43.This is supported by Microsoft’s in liminefilings and Uniloc’s response, where Unilocexplicitly said that it would not be relyingon the entire market value of the accusedproducts. This court thus agrees with thedistrict court that Microsoft has notwaived its objection.

Uniloc argues that the entire marketvalue of the products may appropriately beadmitted if the royalty rate is low enough,relying on the following statement in Lu-cent Technologies:

Simply put, the base used in a runningroyalty calculation can always be thevalue of the entire commercial embodi-ment, as long as the magnitude of therate is within an acceptable range (asdetermined by the evidence)TTTT Micro-soft surely would have little reason tocomplain about the supposed applicationof the entire market value rule had thejury applied a royalty rate of .1% (in-stead of 8%) to the market price of theinfringing programs.’’


580 F.3d at 1338–39. Just before thisstatement, however, this court held thatone of the flaws in the use of the entiremarket value in that case was ‘‘the lack ofevidence demonstrating the patentedmethod of the Day patent as the basis—oreven a substantial basis—of the consumerdemand for OutlookTTTT [t]he only reason-able conclusion supported by the evidenceis that the infringing use of the date-pickertool in Outlook is but a very small compo-nent of a much larger software program.’’Id. at 1338. Thus, in context, the passagerelied on by Uniloc does not support itsposition. The Supreme Court and thiscourt’s precedents do not allow consider-ation of the entire market value of accusedproducts for minor patent improvementssimply by asserting a low enough royaltyrate. See Garretson, 111 U.S. at 121, 4S.Ct. 291; Lucent Techs., 580 F.3d at 1336(‘‘In one sense, our law on the entire mar-ket value rule is quite clear. For theentire market value rule to apply, the pat-entee must prove that the patent-relatedfeature is the basis for customer demand’’(emphasis added, internal citations omit-ted)); Rite–Hite, 56 F.3d at 1549 (same);Bose Corp. v. JBL, Inc., 274 F.3d 1354,1361 (Fed.Cir.2001) (same); TWM Mfg.Co. v. Dura Corp., 789 F.2d 895, 901 (Fed.Cir.1986) (‘‘The entire market value ruleallows for the recovery of damages basedon the value of an entire apparatus con-taining several features, when the featurepatented constitutes the basis for customerdemand.’’).

This case provides a good example ofthe danger of admitting consideration ofthe entire market value of the accusedwhere the patented component does notcreate the basis for customer demand. Asthe district court aptly noted, ‘‘[t]he $19billion cat was never put back into the bageven by Microsoft’s cross-examination ofMr. Gemini and re-direct of Mr. Napper,and in spite of a final instruction that thejury may not award damages based on

Microsoft’s entire revenue from all the ac-cused products in the case.’’ Uniloc II,640 F.Supp.2d at 185. This is unsurpris-ing. The disclosure that a company hasmade $19 billion dollars in revenue froman infringing product cannot help but skewthe damages horizon for the jury, regard-less of the contribution of the patentedcomponent to this revenue. Uniloc exacer-bated the situation in colloquies like thefollowing on cross-examination of Micro-soft’s damages expert, in which it implied arelationship between the entire marketvalue of the accused products and the pat-ent:

Q [Uniloc]. You understand that thereare approximately $20 billion in sales ofinfringing product, correct?A [Napper]. That’s the calculation byMr. Gemini, yes, the entire market valueof those products.Q. And you understand your lump-summax theory is $7 million?A. Yes.Q. And that would be an effective roy-alty of approximately .000035%?A. If one were inappropriately puttingthe entire market value of the products,that’s what it would result in.Q. Uniloc invents it, correct?A. They have a patent, yes.Q. And under your theory, Microsoftgoes out and infringes a valid patent,right?A. That’s my assumption.Q. Under your theory, Microsoftbrings in billions in revenue and salesfrom the sales of the infringing product,to wit, approximately 20, correct?A. The entire market value of thoseproducts, that’s correct.Q. And at the end of the day, the in-fringer, Microsoft, who violated the pat-ent law, they get to keep 99.9999% ofthe box and the inventor, whose patent


they infringed, he gets the privilege ofkeeping .00003%?A. When expressed as the entire mar-ket value of the products, that’s correct.Q. And that’s reasonable to you?A. Yes.

This is in clear derogation of the entiremarket value rule, because the entire mar-ket value of the accused products has notbeen shown to be derived from the patent-ed contribution.

Uniloc’s final argument is that the use ofthe $19 billion figure was only as a check,and the jury must be presumed to havefollowed the jury instruction and not basedits damages calculation on the entire mar-ket value rule. This argument attempts togloss over the purpose of the check aslending legitimacy to the reasonableness ofGemini’s $565 million damages calculation.Even if the jury’s damages calculation wasnot based wholly on the entire marketvalue check, the award was supported inpart by the faulty foundation of the entiremarket value. Moreover, Uniloc’s derisionof Microsoft’s damages expert by virtue ofthe .00003% of the entire market valuethat his damages calculation representedmay have inappropriately contributed tothe jury’s rejection of his calculations.Thus, the fact that the entire market valuewas brought in as only a ‘‘check’’ is of nomoment.

For the foregoing reasons, this courtconcludes that the district court did notabuse its discretion in granting a condi-tional new trial on damages for Uniloc’sviolation of the entire market value rule.

3. Excessiveness of Damages

As an alternative ground for affirmanceof the district court’s alternative grant of anew trial on damages, Microsoft arguesthat the damages here were excessive.Because this court is affirming the districtcourt’s grant of new trial on damages, andbecause the two bases on which Uniloc’s

damages case was built have both beenrejected, it would be premature to considerthe excessiveness of damages that couldarise on remand. This court thus express-es no opinion on the excessiveness or rea-sonableness of the damages awarded bythe jury.

D. Cross–Appeal

Microsoft also cross-appeals the districtcourt’s denial of its motion for JMOL ofinvalidity. Microsoft argues that underUniloc’s interpretation of the claim con-struction in its infringement case, claim 19is invalid as anticipated or obvious overU.S. Patent No. 4,658,093 (‘‘8093 refer-ence’’) titled, ‘‘Software Distribution Sys-tem.’’

Before this court addresses the merits,two procedural issues must be addressed.First, Microsoft argues that its burden forboth the new trial and JMOL motions wasto show invalidity simply by a preponder-ance of the evidence, because the 8093reference was not before the PTO. Thisargument is based on a statement in KSRInt’l Co. v. Teleflex Inc., 550 U.S. 398, 426,127 S.Ct. 1727, 167 L.Ed.2d 705 (2007)(‘‘[T]he rationale underlying the presump-tion—that the PTO, in its expertise, hasapproved the claim—seems much dimin-ished here [where the allegedly invalidat-ing prior art was not before the patentoffice].’’). Microsoft has made this argu-ment before, and we held that the statuto-ry presumption of validity can be overcomeonly by showing invalidity by clear andconvincing evidence, even where allegedlyinvalidating prior art was not before thepatent office. See i4i, 598 F.3d at 848,cert. granted 562 U.S. ––––, 131 S.Ct. 647,178 L.Ed.2d 476 (2010); Am. Hoist & Der-rick Co. v. Sowa & Sons, Inc., 725 F.2d1350, 1359 (Fed.Cir.1984). Until changedby the Supreme Court or this court sittingen banc, that is still the law.


Second, Microsoft argues that the dis-trict court’s grant of a new trial extendedto validity even though the district courtdid not mention validity in the new trialsection of its opinion. See Uniloc II, 640F.Supp.2d at 183–86. Microsoft failed toraise this issue in its motion for a newtrial, and this court finds no basis to con-clude that the district court implicitly de-cided the issue. The issue is thus notbefore this court and we do not reach it.

Turning then to the merits. Microsoftframes its validity arguments as follows: ifclaim 19 reaches far enough to read onMicrosoft’s Product Activation, then it nec-essarily extends far enough to read on theprior art 8093 reference. Thus, Microsoftuses Product Activation as a proxy for thescope of claim 19 (under the assumptionthat the jury verdict of infringement isupheld and Product Activation infringes),and compares it to the 8093 reference.According to Microsoft, the prior art 8093reference discloses a software authoriza-tion process and system, which generatesan authorization code from the followinginputs: ‘‘a secret key identifier of the com-puter embodied in the hardware (SK), arandom or nonrepeating number (R), theserial number, the software package name(H), the number of uses (N), and userbilling information.’’ Uniloc II, 640F.Supp.2d at 181.

In the prior appeal, this court held thatthe licensee unique ID must be ‘‘a uniqueidentifier associated with a licensee,’’ butone ‘‘that cannot be based solely on plat-form-related user information.’’ Uniloc I,290 Fed.Appx. at 343–44. The question inthis appeal is whether no reasonable jurycould have concluded that the inputs to thecryptographic hash function in the 8093reference are not sufficient to create an

association with the licensee. The focus ofthe dispute on appeal, just as below, iswhether random number R in the 8093reference is uniquely associated with auser. We think that it is not and that the8093 reference does not anticipate claim 19.

Uniloc argues that R is ‘‘platform-relat-ed’’ because it is generated by the user’scomputer. Microsoft argues that ‘‘plat-form-related’’ must be narrower than‘‘generated by a computer,’’ otherwise itcould not infringe because the output ofProduct Activation (the PID) is also gener-ated by a computer.

It is undisputed that R and SK, the onlyinputs that Microsoft argues are associatedwith the user, are generated by the com-puter.3 In the prior appeal, this courtnoted that the licensee unique ID ‘‘cannotbe based solely on platform related userinformation.’’ Id. at 343. This was basedin part on Uniloc’s distinction made duringprosecution of the 8216 patent between itsinvention and a prior art reference that‘‘relie[d] for its security on a machine iden-tification code unique to the machine.’’ Id.at 343–44. The number R in the 8093reference does just what the distinguishedprior art does: whatever association it cre-ates does not identify the user, but ratherit identifies the machine. In other words,the R in the 8093 reference is platform-related, and thus cannot form the basis ofthe association between the output of thecryptographic hash function and the licen-see. The same is true for SK, which isbased on a computer-generated serialnumber.

Microsoft’s argument that R is not plat-form-related though it is generated by acomputer (without user input) is uncon-vincing for two reasons. First, the proper

3. The ‘‘user billing information’’ in the 8093reference is not an input into the hash func-tion and is thus irrelevant in determiningwhether the 8093 reference discloses the ‘‘li-

censee unique ID’’ and ‘‘licensee unique IDgenerating means’’ elements of the 8216 pat-ent.


framework for challenging the validity of apatent is not for the accused to show thatit is practicing the prior art, but to showthat every element of the patent claimsreads on a single prior art reference. SeeZenith Elecs. Corp. v. PDI Comm. Sys.,Inc., 522 F.3d 1348, 1363 (Fed.Cir.2008)(‘‘[M]ere proof that the prior art is identi-cal, in all material respects, to an allegedlyinfringing product cannot constitute clearand convincing evidence of invalidity. An-ticipation requires a showing that eachelement of the claim at issue, properlyconstrued, is found in a single prior artreference.’’). Second, although the outputof the hash algorithm in Product Activa-tion, the PID, is computer generated, theinformation that is input into the MD5algorithm is associated with a user, in theform of the product key provided by thevendor. This court has held that it isreasonable to determine that such informa-tion creates an association with the user.Uniloc I, 290 Fed.Appx. at 343 n. 4 (‘‘Thespecification [of the 8216 patent] certainlydoes allow for the use of vendor-providedinformation to generate a licensee uniqueID.’’). As noted, supra section II.A.3, thiscourt is bound by the first appeal thatinputs associated with the licensee are suf-ficient to maintain the association in theoutput as well. See also Uniloc I, 290Fed.Appx. at 344 (‘‘Microsoft’s Product Ac-tivation system inputs non-platform-relat-ed information unique to a user, such as aProduct Key, to generate what might qual-ify as a licensee unique ID.’’). Unlike thePID, the output of the hash algorithm inthe 8093 reference is ‘‘based solely on plat-form-related user information,’’ see id. at343, in the form of R and SK, neither ofwhich are ‘‘vendor-provided,’’ and both ofwhich are generated by the user’s comput-er. The PID is thus distinguishable fromthe output of the hash algorithm in the8093 reference, and a reasonable jury couldhave returned a verdict that the 8093 ref-erence does not disclose a ‘‘licensee unique

ID’’ as required by claim 19 of the 8216patent. This court thus affirms the dis-trict court’s denial of JMOL of invaliditybased on anticipation.

[19] Though obviousness is a questionof law, this court gives the jury its usualdeference on the underlying factual ques-tions. What the prior art shows is a ques-tion of fact. Graham v. John Deere Co.,383 U.S. 1, 17, 86 S.Ct. 684, 15 L.Ed.2d 545(1966). As discussed above, a reasonablejury could have found that the 8093 refer-ence failed to disclose a licensee uniqueID. Microsoft has presented no convincingargument for why incorporating an inputinto the hash function that is associatedwith a user would have been obvious toone of ordinary skill in the art. This courtthus affirms the district court’s denial ofJMOL of invalidity on the basis of obvious-ness.

CONCLUSION

For the foregoing reasons, this courtreverses the district court’s grant ofJMOL of non-infringement, affirms thedistrict court’s grant of JMOL of no will-fulness, affirms the district court’s grant ofa new trial on damages, vacates the dis-trict court’s grant of an alternative motionfor new trial on infringement, and affirmsthe district court’s denial of JMOL of in-validity of claim 19 of the 8216. The caseis remanded for proceedings consistentwith this opinion.

AFFIRMED–IN–PART, REVERSED–IN–PART, VACATED–IN–PART, andREMANDED

COSTS

Each party shall bear its own costs.

,

1292 632 federal reporter, 3d series - …...1292 632 federal reporter, 3d series [2] because mr....

Documents