decoration decoration
Stories

GROKLAW
When you want to know more...
decoration
For layout only
Home
Archives
Site Map
Search
About Groklaw
Awards
Legal Research
Timelines
ApplevSamsung
ApplevSamsung p.2
ArchiveExplorer
Autozone
Bilski
Cases
Cast: Lawyers
Comes v. MS
Contracts/Documents
Courts
DRM
Gordon v MS
GPL
Grokdoc
HTML How To
IPI v RH
IV v. Google
Legal Docs
Lodsys
MS Litigations
MSvB&N
News Picks
Novell v. MS
Novell-MS Deal
ODF/OOXML
OOXML Appeals
OraclevGoogle
Patents
ProjectMonterey
Psystar
Quote Database
Red Hat v SCO
Salus Book
SCEA v Hotz
SCO Appeals
SCO Bankruptcy
SCO Financials
SCO Overview
SCO v IBM
SCO v Novell
SCO:Soup2Nuts
SCOsource
Sean Daly
Software Patents
Switch to Linux
Transcripts
Unix Books
Your contributions keep Groklaw going.
To donate to Groklaw 2.0:

Groklaw Gear

Click here to send an email to the editor of this weblog.


To read comments to this article, go here
Microsoft's Appeal Brief (i4i), as text
Wednesday, August 26 2009 @ 11:49 PM EDT

Microsoft's Appeal Brief [PDF] in the i4i patent litigation is now available. It's 101 pages. And it concludes like this:
The injunction should be vacated. The judgment should be reversed and the case remanded with instructions to enter judgment of noninfringement and invalidity. In the alternative, the judgment should be vacated and the case remanded for a new trial on all issues as to which Microsoft is not entitled to judgment.
They sound like it's a command, don't they? So that's what they want. No injunction. They want the court to rule that they didn't infringe and that i4i's patent is not valid. Either that or they want the lower court's ruling tossed off a cliff, and then a new trial.

They address the extra damages issues for trial misconduct, and here's how they put it:
Based on the jury’s finding of willful infringement, Microsoft’s size and profitability, and the court’s extraordinary conclusion that, in advocating for his client, Microsoft’s trial counsel had committed “trial misconduct” (A47), the district court decided to enhance i4i’s already outrageous $200,000,000 damage award by an additional $40,000,000. Microsoft was entitled to judgment as a matter of law on the issue of willfulness. But even if that were not the case, the Court’s enhancement analysis is premised on errors of law and accordingly must be vacated.
i4i will answer by September 9th, and then Microsoft gets to respond by the 14th.

Update: Todd Bishop at TechFlash has a response from i4i:

In a statement issued this afternoon, i4i Chairman Loudon Owen called Microsoft's appeal brief an "extraordinary document." Owen added, "It captures the hostile attitude of Microsoft toward inventors who dare to enforce patents against them. It is also blatantly derogatory about the Court system." He continued:

“We sought and received the protection of the court so we can compete on a level playing field, and run our business without infringement by the Defendant Microsoft. Microsoft was proven to have willfully infringed on i4i’s patent. ... We do not have the gargantuan financial resources of Microsoft, but i4i has the protection of fairness under the U.S. justice system. Microsoft is not above the law. It cannot privately expropriate i4i’s patented invention ... We firmly believe the decision of the jury and judge in the United States District Court for the Eastern District of Texas was correct on the facts and we shall prevail on appeal.”

Update 2: Law.com has an article from American Lawyer about the tone of Microsoft's brief, In Yer Face, Judge! Microsoft Files Federal Circuit Brief in i4i Case, so I am by no means alone in noticing:
Right back at ya, Judge Davis. That's the message we took away from the 102-page brief Microsoft filed Tuesday with the U.S. Court of Appeals, seeking to overturn a $290 million patent infringement judgment entered earlier this month against Microsoft by Tyler, Texas, federal district court judge Leonard Davis....

As he promised us last week, Powers also addressed Judge Davis's finding that he made improper arguments at trial when he compared i4i to a bank seeking a bailout. Partly because of those arguments, Judge Davis enhanced the jury's award to i4i by $40 million. But Powers and his colleagues asert in the Federal Circuit brief that alleged trial misconduct has traditionally not justified enhanced awards. Further, they write that there was nothing improper about what Powers told the jury about i4i--because it was true.

"Indeed, when the assertedly 'improper arguments' were made at trial, i4i found no cause to object," Powers and his colleagues write. "Perhaps that is because Microsoft’s argument that i4i and its litigation investors had brought the litigation to achieve liquidity was obviously and inescapably true in light of i4i’s claim for $200 million in damages and the belated request for injunctive relief....The district court might not have liked the implications of those facts, but careful examination of the record will show that, contrary to what the out-of-context statements cited by the district court may suggest, Microsoft never argued that simply asserting a patent in these circumstances was improper."

Here it is as text, very rough text, which I'll try to perfect later, but I wanted you to have it handy even in rough form, and as always, for anything important, go by the PDF:

***********************************

No. 2009-1504

________________________________________________

IN THE UNITED STATES COURT OF APPEALS
FOR THE FEDERAL CIRCUIT

________________________________________________

I4I LIMITED PARTNERSHIP AND
INFRASTRUCTURES FOR INFORMATION INC.,

Plaintiffs-Appellees,

v.

MICROSOFT CORPORATION,

Defendant-Appellant.

__________________________________________________________________

On Appeal from the United States District Court
For the Eastern District of Texas in Case No. 07-CV-113,
Judge Leonard Davis

___________________________________________________

BRIEF FOR DEFENDANT-APPELLANT
MICROSOFT CORPORATION

___________________________________________________

Matthew D. Powers
Principal Attorney
WEIL, GOTSHAL &
MANGES LLP
[address, phone]

Kevin Kudlac
Amber H. Rovner
WEIL, GOTSHAL &
MANGES LLP

Matthew D. McGill
Minodora D. Vancea
GIBSON, DUNN &
CRUTCHER LLP

Counsel for Defendant-Appellant

__________________________________________________________________

Additional counsel listed on inside front cover

ISABELLA FU
MICROSOFT CORPORATION
[address, phone, fax]

Counsel for Defendant-Appellant Microsoft Corporation

ii

CERTIFICATE OF INTEREST
UNITED STATES COURT OF APPEALS FOR THE FEDERAL CIRCUIT

I4I V. MICROSOFT
2009-1504

Counsel for the Defendant-Appellant certifies the following:

1. The full name of every party or amicus represented by me is: Microsoft Corporation.

2. The name of the real party in interest (if the party named in the caption is not the real party in interest) represented by me is: N/A.

3. All parent corporations and any publicly held companies that own 10 percent or more of the stock of the party or amicus curiae represented by me are: N/A.

4. X  There is no such corporation as listed in paragraph 3.

5. The names of all law firms and the partners or associates that appeared for the party or amicus now represented by me in the trial court or agency or are expected to appear in this court are:

Matthew D. Powers
[email]
WEIL GOTSHAL & MANGES LLP
[address, phone, fax]

David J. Lender
[email]
Paul E. Torchia
[email]
Steven Kalogeras
[email]
Kevin Kudlac
[email]
Amber H. Rovner
[email]
Todd S. Patterson
[email]
WEIL GOTSHAL & MANGES LLP
[address, phone, fax]

i

Ariane N. Newell
[email]
WEIL GOTSHAL & MANGES LLP
[address, phone, fax]

Eric H. Findlay
[email]
Roger B. Craft
[email]
FINDLAY CRAFT
[address, phone, fax]

David J. Healey
[email]
Norma N. Bennett
[email]
FISH & RICHARDSON PC
[address, phone, fax]

Matthew D. McGill
[email]
Minodora D. Vancea
[email]
GIBSON, DUNN & CRUTCHER LLP
[address, phone, fax]

T. Andrew Culbert [email]
Isabella Fu
[email]
Microsoft Corporation
[address, phone fax]

Cabrach J. Connor
[address]

Lucy Muzzy
formerly with
WEIL GOTSHAL & MANGES LLP
[address]

August 25, 2009
Date

[signature]
Minodora D. Vancea

ii

TABLE OF CONTENTS

PRELIMINARY STATEMENT ...............................................................................1

JURISDICTIONAL STATEMENT ..........................................................................5

STATEMENT OF THE ISSUES...............................................................................5

STATEMENT OF THE CASE..................................................................................7

STATEMENT OF FACTS ........................................................................................7

SUMMARY OF ARGUMENT ...............................................................................16

STANDARDS OF REVIEW...................................................................................21

ARGUMENT ...........................................................................................................22

I. Claim Construction Errors Require Reversal Of The Judgment Of
Infringement ........................................................................................22
A. The District Court's Construction Reads "Distinct" Out
Of The Claims' Requirement Of Storage Of The Metacode Map..........................................................................22
1. The Specification Expressly Defines “Distinct” Map Storage Means As Separated From Mapped Content Storage Means..................................................24

2. The Applicants Represented To The PTO That “Distinct” Storage Requires Separate Files, Not Just Different Memory Locations..................................26

B. “Distinct” Storage Requires The Ability To Independently Manipulate The Metacode Map And Mapped Content.......................................................................27
1. The Specification Defines “The Invention” As Allowing Independent Manipulation.............................28

2. The Applicants Repeatedly Distinguished The Prior Art On The Basis That Their Invention Allows Independent Manipulation................................29

ii

C. There Is No Infringement Under A Correct Claim Construction .............................................................................30
1. Word Does Not Store The Metacode Map "Distinctly" From The Mapped Content .......................30

2. Word Does Not Allow A Metacode Map And Content To Be Manipulated Independently ..................32

II. The Asserted Claims Are Invalid ........................................................33
A. The Claimed Invention Was Obvious......................................33
1. Each Limitation Was Disclosed In Obvious Combinations Of The Prior Art .....................................34

2. Secondary Considerations Do Not Suffice....................38

B. The On-Sale Bar Applies .........................................................39

C. The Judgment Of Validity Cannot Stand.................................44

III. The Jury's Finding Of Indirect Infringement Cannot Be Sustained .............................................................................................46
A. The District Court Erroneously Instructed The Jury That Microsoft Committed Contributory Infringement If It Sold A "Component" Of i4i's Claimed Methods ....................47

B. Microsoft Is Entitled To Judgment As A Matter Of Law On Contributory Infringement Because The Software It "Sold" --Word--Indisputably Has Substantial Noninfringing Uses ..................................................................49

C. i4i's Own Evidence Demonstrates That Even The Accused Functionality Of Word Has Substantial Noninfringing Uses ..................................................................51

D. The Inferences Of Intent Are Legally Insufficient To Support Either Induced Or Contributory Infringement ...........52

IV. The $200,000,000 Damages Award Cannot Be Sustained .................54

iii

A. The Results Of The Wecker Survey Should Have Been Excluded...................................................................................55

B. Wagner's Georgia-Pacific Analysis Should Have Been Excluded...................................................................................60

C. Two Hundred Million Dollars Is Not A Reasonable Royalty .....................................................................................66

D. The District Court's Award Of $40 Million In Enhanced Damages Is Unsupportable. .....................................................68

1. The District Court’s Willfulness Analysis Cannot Be Reconciled With Seagate.........................................68

2. The District Court’s Enhancement Analysis Fails On Its Own Terms.........................................................72

V. The Award Of Injunctive Relief In This Case Conflicts With eBay............................................................................................75 CONCLUSION........................................................................................................80

TABLE OF AUTHORITIES

Cases

Page(s)

Agrizap, Inc. v. Woodstream Corp.,
520 F.3d 1337 (Fed. Cir. 2008)............................................................... 21, 33, 38

Albert v. Warner-Lambert Co.,
234 F. Supp. 2d 101 (D. Mass. 2002) ........................................................... 56, 59

AquaTex Indus., Inc. v. Techniche Solutions,
419 F.3d 1374 (Fed. Cir. 2005)............................................................................49

Aro Mfg. Co. v. Convertible Top Replacement Co.,
377 U.S. 476 (1964) .............................................................................................52

Bank of Texas v. Commerce Southwest, Inc.,
741 F.2d 785 (5th Cir. 1984)................................................................................56

Baumstimler v. Rankin,
677 F.2d 1061 (5th Cir. 1982) .............................................................................45

Black & Decker, Inc. v. Robert Bosch Tool Corp.,
260 Fed. Appx. 284 (Fed. Cir. 2008)............................................................ 70, 72

Black v. Food Lion, Inc.,
171 F.3d 308 (5th Cir. 1999)................................................................................65

Boston Sci. Scimed, Inc. v. Cordis Corp.,
554 F.3d 982 (Fed. Cir. 2009)................................................................. 33, 38, 39

Cardiac Pacemakers, Inc. v. St. Jude Medical, Inc.,
--- F.3d ----, 2009 WL 2516346 (Fed. Cir. 2009) ................................................48

Cardinal Chem. Co. v. Morton Int'l,
508 U.S. 83 (1993) ...............................................................................................79

Cohesive Techs., Inc. v. Waters Corp.,
543 F.3d 1351 (Fed. Cir. 2008)............................................................................71

v

Computer Docking Station Corp. v. Dell, Inc.,
519 F.3d 1366 (Fed. Cir. 2008)............................................................................27

Daubert v. Merrell Dow Pharmaceuticals, Inc.,
509 U.S. 579 (1993) .............................................................................................55

Depuy Spine, Inc. v. Medtronic Sofamor Danek, Inc.,
567 F.3d 1314 (Fed. Cir. 2009)............................................................................71

Dombrowski v. Pfister, 380 U.S. 479, 485 (1965) ..................................................28

eBay Inc. v. MercExchange, L.L.C.,
547 U.S. 388 (2006) ...................................................................................... 75, 79

Ecolab, Inc. v. FMC Corp.,
569 F.3d 1335 (Fed. Cir. 2009)............................................................................33

Finisar Corp. v. DirecTV Group, Inc.,
523 F.3d 1323 (Fed. Cir. 2008)............................................................................69

General Elec. Co. v. Joiner,
522 U.S. 136 (1997) .............................................................................................65

Harris Corp. v. Ericsson Inc.,
417 F.3d 1241 (Fed. Cir. 2005)............................................................................32

Hodosh v. Block Drug Co.,
833 F.2d 1575 (Fed. Cir. 1987)............................................................................49

Hoffmann-La Roche Inc. v. Promega Corp.,
33 U.S.P.Q. 2d 1641 (N.D. Cal. 1994) ................................................................51

Honda Motor Co. v. Oberg,
512 U.S. 415 (1994) .............................................................................................73

In re Seagate Technologies, LLC,
497 F.3d 1360 (Fed. Cir. 2007)..................................................................... 68, 72

Integra Lifesciences I, Ltd. v. Merck KGaA,
331 F.3d 860 (Fed. Cir. 2003)....................................................................... 61, 67

vi

Isaksen v. Vermont Castings,
825 F.2d 1158 (7th Cir. 1987) .............................................................................66

Jackson v. Nat'l Action Fin. Servs., Inc.,
441 F. Supp. 2d 877 (N.D. Ill. 2006) ...................................................................58

Juicy Whip, Inc. v. Orange Bang, Inc., 2
92 F.3d 728 (Fed. Cir. 2002)................................................................................44

Jurgens v. CBK, Ltd.,
80 F.3d 1566 (Fed. Cir. 1996)..............................................................................73

Kearns v. Chrysler Corp.,
32 F.3d 1541 (Fed. Cir. 1994)..............................................................................21

King v. Ames,
179 F.3d 370 (5th Cir. 1999)................................................................................66

Koon v. United States,
518 U.S. 81 (1996) ...............................................................................................22

KSR Int'l Co. v. Teleflex, Inc.,
550 U.S. 398 (2007) .......................................................................... 33, 34, 39, 46

Kumho Tire Co., Ltd. v. Carmichael,
526 U.S. 137 (1999) .............................................................................................55

Kyocera Wireless Corp. v. Int'l Trade Comm'n,
545 F.3d 1340 (Fed. Cir. 2008)............................................................................52

Leapfrog Enters., Inc. v. Fisher-Price, Inc.,
485 F.3d 1157 (Fed. Cir. 2007)............................................................................34

Lear, Inc. v. Adkins,
395 U.S. 653 (1969) .............................................................................................79

Lindemann Maschinenfabrik GmbH v. Am. Hoist & Derrick Co., Harris
Press & Shear Div
.,
895 F.2d 1403 (Fed. Cir. 1990)............................................................................66

vii

Mahurkar v. C.R. Bard, Inc.,
79 F.3d 1572 (Fed. Cir. 1996)..............................................................................21

Manufacturing Research Corp. v. Graybar Electric Company,
679 F.2d 1355 (11th Cir. 1982) ...........................................................................45

MEMC Elec. Materials, Inc. v. Mitsubishi Materials Silicon Corp.,
420 F.3d 1369 (Fed. Cir. 2005)............................................................................53

Merck & Co. v. Teva Pharms. USA, Inc.,
395 F.3d 1364 (Fed. Cir. 2005)............................................................................22

Microsoft Corp. v. AT&T Corp.,
550 U.S. 437 (2007) .............................................................................................48

Muniauction, Inc. v. Thomson Corp.,
532 F.3d 1318 (Fed. Cir. 2008)..................................................................... 33, 39

Oiness v. Walgreen Co.,
88 F.3d 1025 (Fed. Cir. 1996)..............................................................................67

Olympia Equipment Leasing Co. v. Western Union Tel. Co.,
797 F.2d 370 (7th Cir. 1986)................................................................................66

Omega Eng'g, Inc. v. Raytek Corp.,
334 F.3d 1314 (Fed. Cir. 2003)............................................................................27

Osram GmbH v. International Trade Comm'n,
505 F.3d 1351 (Fed. Cir. 2007)............................................................................24

Paice LLC v. Toyota Motor Corp.,
504 F.3d 1293 (Fed. Cir. 2007)............................................................................78

Piper Aircraft Co. v. Reyno,
454 U.S. 235 (1981) .............................................................................................80

Pope Mfg. Co. v. Gormully,
144 U.S. 224 (1892) .............................................................................................79

Praxair, Inc. v. ATMI, Inc.,
479 F. Supp. 2d 440 (D. Del. 2007).....................................................................78

viii

Price v. Symsek,
988 F.2d 1187 (Fed. Cir. 1993)............................................................................44

R.R. Comm'n v. Pullman Co.,
312 U.S. 496 (1941) .............................................................................................79

Read Corp. v. Portec, Inc.,
970 F.2d 816 (Fed. Cir. 1992)..............................................................................73

Ricoh Co., Ltd. v. Quanta Computer, Inc.,
550 F.3d 1325 (Fed. Cir. 2008)............................................................... 48, 50, 53

Riles v. Shell Exploration and Production Co.,
298 F.3d 1302 (Fed. Cir. 2002)..................................................................... 61, 65

Rizzo v. Goode,
423 U.S. 362 (1976) .............................................................................................75

Roper Corp. v. Litton Sys., Inc.,
757 F.2d 1266 (Fed. Cir. 1985)............................................................................77

Rotondo v. Keene Corp.,
956 F.2d 436 (3d Cir. 1992).................................................................................21

Rutherford v. Harris County,
197 F.3d 173 (5th Cir. 1999)................................................................................47

Sampson v. Murray,
415 U.S. 61 (1974) ...............................................................................................78

Schering Corp. v. Pfizer Inc.,
189 F.3d 218 (2d Cir. 1999).................................................................................59

Schiller & Schmidt, Inc. v. Nordisco Corp.,
969 F.2d 410 (7th Cir. 1992)................................................................................60

Scott Fetzer Co. v. House of Vacuums, Inc.,
381 F.3d 477 (5th Cir. 2004)......................................................................... 55, 57

Seachange Int'l, Inc. v. C-COR Inc.,
413 F.3d 1361 (Fed. Cir. 2005)............................................................................32

ix

Silverstein v. Penguin Putnam, Inc.,
368 F.3d 77 (2d Cir. 2004)...................................................................................75

Spectrum Sports, Inc. v. McQuillan,
506 U.S. 447 (1993) .............................................................................................47

Stumbo v. Eastman Outdoors, Inc.,
508 F.3d 1358 (Fed. Cir. 2007)............................................................................23

Sundance, Inc. v. DeMonte Fabricating Ltd.,
No. 02-73543, 2007 WL 37742 (E.D. Mich. Jan. 4, 2007) .................................76

U.S. Envt'l Prods. Inc. v. Westall,
911 F.2d 713 (Fed. Cir. 1990)..............................................................................43

Unisplay, S.A. v. American Electronic Sign Co., Inc.,
69 F.3d 512 (Fed. Cir. 1995)................................................................................67

United States v. Carlock,
806 F.2d 535 (5th Cir. 1986)................................................................................57

United States v. Oregon State Med. Soc.,
343 U.S. 326 (1952) .............................................................................................77

Verizon Servs. Corp. v. Vonage Holdings Corp.,
503 F.3d 1295 (Fed. Cir. 2008)............................................................................24

Virginian Ry. Co. v. Sys. Fed'n,
300 U.S. 515 (1937) .............................................................................................79

Voda v. Cordis Corp.,
536 F.3d 1311 (Fed. Cir. 2008)............................................................................54

Water Techs. Corp. v. Calco, Ltd.,
850 F.2d 660 (Fed. Cir. 1988)..............................................................................53

Statutes

35 U.S.C. § 102(b) ...................................................................................................39

35 U.S.C. § 103........................................................................................................33

x

35 U.S.C. § 271(c) ................................................................................ 47, 49, 50, 51

35 U.S.C. § 284................................................................................................. 54, 66

Rules

FED. R. EVID. 701 .....................................................................................................58

Constitutional Provisions

U.S. Const. art. I, § 8..................................................................................................5

Other Authorities

REFERENCE MANUAL ON SCIENTIFIC EVIDENCE, FEDERAL JUDICIAL CENTER 245 (2d ed. 2000) .................................................................................................59

Richard S. Toikka, Patent Licensing Under Competitive and Non- Competitive Conditions, 82 J. Pat. & Trademark Off. Soc'y 279 (2000) ...........61

xi

TABLE OF ABBREVIATIONS

Parties

i4i - Plaintiffs-Appellees i4i Limited Partnership and Infrastructures for Information, Inc.

Microsoft - Defendant-Appellant Microsoft Corporation

Cites

A__ Joint Appendix at page(s) ___

Terms

'449 Patent - U.S. Patent No. 5,787,449

Asserted Claims - Claims 14, 18 and 20 of the '449 Patent

DeRose - U.S. Patent No. 6,101,512

HTML - Hypertext Markup Language

Kugimiya - U.S. Patent No. 5,587,902

Rita - A2890-2915; A2916-2926; A2927-3043; A3044-3226; A3227-3412

SGML - Standard Generalized Markup Language

XML - Extensible Markup Language

Notes

**Unless otherwise indicated herein, all emphases have been added.

xii

INDEX TO APPENDED MATERIALS

1. Memorandum Opinion and Order dated August 11, 2009 (A5-69)

2. U.S. Patent No. 5,787,449 (A238-256)

STATEMENT OF RELATED CASES

Pursuant to FED. CIR. RULE 47.5(a), Microsoft states that no appeal, other than the current appeal by Microsoft, has been taken in or from the same civil action or proceeding in the lower court. Pursuant to FED. CIR. RULE 47.5(b), Microsoft states that no other cases are pending between the same parties.

xiv

PRELIMINARY STATEMENT

In patent cases, even more than most, the trial judge's role as a gatekeeper is crucial. As gatekeeper, the judge must define the metes and bounds of a patent through claim construction and then ensure that the evidence presented by the parties' numerous experts is both reliable and rooted in the facts of the case at hand. And after the jury has rendered its verdict, it is the judge who, before allowing that verdict to become an enforceable judgment, must ensure that the verdict is adequately supported by the evidence and supportable under the law. This gatekeeping function is especially important in patent cases because of the delicate balance struck by patent law to achieve its objective of promoting, rather than stifling, innovation. That balance can be lost if the district court does not protect the process, and patent litigation then becomes a tax on innovation rather than its guardian.

This case stands as a stark example of what can happen in a patent case when a judge abdicates those gatekeeping functions.

The district court's errors in this case began when it accorded a critical claim term no meaning whatsoever, effectively erasing it from the patent. That impermissible claim construction, in turn, enabled the plaintiffs to present to the jury a theory of infringement that contradicted not just the asserted patent's specification, but also its prosecution history, and, indeed, its very title.

1

The court's failure to act as a gatekeeper extended to Microsoft's validity defenses as well. After Microsoft invoked the on-sale bar with the inventor's pre- litigation statements that showed his invention to have been "implemented" and sold before the critical date, the inventor insisted that he had lied to investors about his pre-sale invention date and that the invention really came later, and that Microsoft could not prove otherwise because he had "discarded" the source code of that product. In the absence of corroboration, an inventor's manifestly self- interested testimony should be insufficient as a matter of law to save his patent. But the district court demurred and sent the question to the jury.

By the time the plaintiffs presented their damages case, the district court had abandoned even the semblance of gatekeeping. Plaintiffs sought to quantify instances of indirect infringement--their royalty base--through a survey of 988 businesses. That survey garnered just 46 individual respondents, who then were paid to guess how many other persons in their place of employment used Microsoft Word in a manner similar to the alleged infringing implementation. Of those 46 respondents, 19--less than half--were aware of any such use of Word. And even they supplied inconsistent or confused responses, so plaintiff's expert simply changed the data --all according to so-called principles of "data imputation," of course. When his manipulations were done, he testified that the 19 responses to his survey proved that 1.85 million Word users infringed plaintiffs'

2

patent--nearly 100,000 for every usable response. Plaintiffs then extrapolated that the 1.85 million infringing uses supposedly demonstrated by the survey had grown to 2.1 million in the six months before trial. When Microsoft challenged the survey as inherently unreliable, the district court offered the clearly erroneous response that fundamental admissibility considerations such as whether a survey respondent has personal knowledge of facts surveyed and whether the results were manipulated went only to the weight of the evidence. If this survey passes muster, there are no limits to what "evidence" can be presented to lead juries astray, and patent law suffers.

A similar fate befell Microsoft's challenge to plaintiffs' Georgia-Pacific analysis. Plaintiffs' expert concluded that, in a hypothetical negotiation, Microsoft would have agreed to a royalty rate of $98--more than half of the current retail price of Word 2007. Wagner reached this conclusion by applying the much- maligned "25 percent rule of thumb" to the hypothetical profit one might have received if Microsoft's profit margin applied to an unrelated third-party's radically dissimilar product never shown to practice the patent and that costs (at retail) nearly three times as much as Word and ten times as much as the maximum possible value of the accused functionality. But the district court had no difficulty concluding that this analysis adequately "fit" the facts of this case. It permitted the expert to testify that a "reasonable royalty" was $200,000,000.

3

Lacking any guidance from the district court as to when a royalty might be unreasonable, the jury awarded plaintiffs exactly what they asked for: a finding of willful infringement and an award of $200,000,000.

At this point, if the district court had been more faithful to its role as gatekeeper, it should have recognized a trial run amok and interceded to prevent a miscarriage of justice--a judicial act that would have been particularly appropriate given that, by the time of the post-trial motions, the PTO had provisionally invalidated the patent-in-suit on a re-examination that considered key prior art presented at trial but never to the examiner during prosecution. But the district court chose another path: It sustained the verdict and improperly awarded plaintiffs $40,000,000 in "enhanced" damages--$15,000,000 more than even the plaintiffs had asked for--and entered a permanent injunction that, on October 10, 2009, will prohibit Microsoft from selling any version of Word or Office currently on the market, even though the plaintiffs do not now compete with Word.

This is not justice. If district courts are free to admit theories of infringement that nullify a patent's claim terms, specification, prosecution history, and title; if they will allow an inventor to validate his patent by testifying without corroboration that he lied about the date of conception; if they will not intercede to preclude manifestly unreliable--indeed, concededly manipulated--surveys of infringing use, or Georgia-Pacific analyses based on "benchmarks" bearing no

4

rational relationship to the accused product, then patent litigation will be reduced to a free-for-all, unbounded by the requirements of the substantive law or the rules of evidence or trial procedure. While that mode of dispute resolution might enrich some plaintiffs and their investors, it hardly can be said to "promote the Progress of Science and the useful Arts." U.S. Const. art. I, § 8.

JURISDICTIONAL STATEMENT Pursuant to 28 U.S.C. § 1295(a)(1), this Court has jurisdiction over Microsoft's timely appeal from the final judgment of the district court. (A1-2.)

STATEMENT OF THE ISSUES 1. Whether the judgment of infringement should be reversed, or at least vacated, where the verdict was premised on a legally erroneous claim construction?

2. Whether the verdict of validity should be vacated where:

a. prior-art combinations not considered during prosecution render

the claimed invention obvious, as confirmed by the PTO's rejection in a

pending reexamination; and

b. undisputed evidence showed that the patentee sold a product

more than a year prior to the critical date that it represented, to the Canadian

government and others, embodied the claimed invention, and where the only

5

opposing evidence is the uncorroborated, conclusory trial testimony of the

inventors as to a later conception date?

3. Whether the verdict of indirect infringement should be vacated as premised on a jury instruction that contradicts the Patent Act, a legally erroneous view of substantial noninfringing uses, and a finding of scienter that cannot be supported as a matter of law?

4. a. Whether a jury verdict of $200,000,000 can be sustained as a "reasonable royalty" where the royalty base of 2.1 million is predicated on an inherently unreliable survey, and the royalty rate of $98--from a product whose retail price is between $97 and $229--is based on application of the "25 percent rule" to a third-party "benchmark" product not alleged to embody the patent and whose retail price is $499?

b. Whether enhanced damages of $40,000,000 can be sustained when the defendant offered numerous good-faith defenses, where the PTO has provisionally invalidated the patent-in-suit, and where the only evidence of a "willful" state of mind is the fact that the patent number was included in marketing materials delivered to a handful of the defendant's employees?

6

5. Whether the traditional principles of equity permit a court to issue an injunction to plaintiffs who delayed four years before asserting infringement and have shown no harm in the past several years?

STATEMENT OF THE CASE Plaintiffs i4i L.P. and i4i, Inc. alleged that certain versions of Microsoft's Word software infringed specified claims of the '449 Patent. (A5632-37.) After a trial, the jury found the Asserted Claims valid and infringed. (A236-37.) The district court denied all of Microsoft's post-trial motions, awarded enhanced damages, and granted i4i's request for a permanent injunction. (A5-69.) This appeal followed, in which Microsoft immediately moved this Court to stay the injunction pending appeal. Microsoft's motion to stay is pending.

STATEMENT OF FACTS This case concerns a technology called markup languages. At its most basic level, a markup language is a way of indicating how text should be displayed-- which words are in boldface, for example, or what should be centered, or where line breaks should appear. In general, a markup language inserts "tags" that indicate how the text between those tags should look on a computer screen. For example, is a computer markup code that indicates the start of a paragraph and indicates the end of a paragraph. The '449 Patent calls these markup codes "metacodes." HTML, SGML, and XML are examples of markup languages.

7

The '449 Patent does not claim to have invented any markup languages. (See e.g., A249 at 1:65-2:54.) Rather, markup languages for computers have existed and, indeed, have been standardized for decades. (A5355-67; A5525-27.) Computer programs for creating and editing these markup languages also have existed for decades, including one sold by the inventors of the '449 Patent before the critical date--called SEMI S4--as well as programs disclosed in two other prior-art references, Rita and DeRose. None of this prior art was before the Patent Office at the time of the prosecution, but the latter two references are now before the Patent Office on reexamination, which has resulted in an initial rejection of the '449 Patent. (A5622-31.)

In June 1994, i4i, Inc. filed its patent application, entitled "Method and System for Manipulating the Architecture and the Content of a Document Separately from Each Other." (A239.) Michel Vulpe, one of the named inventors and i4i, Inc.'s founder, promptly touted the pending patent in a funding application to the Canadian government, boasting that he was patenting the existing SEMI S4 product by applying for the '449 Patent and noting that "[t]he initial implementation is embedded into [i4i's] S4 product" (A3759) and "single metacode model implemented in i4i flagship product S4." (A3770; see also A1978-81; A3759-60; A3769-72.) By the time the '449 Patent issued in 1998, i4i, Inc. had

8

been attempting to capitalize on its so-called invention for over five years, but had lost money every year. (A907-13; A5649-53; A5814-5929.)

In late 2002, Microsoft released a beta version of Word 2003 that contained all of the functionality accused of infringement. (A1701-05.) i4i, Inc., which was a member of Microsoft's Developer Network, got an advance copy and evaluated it. (A1040-41; A1700-01.) But i4i, Inc. did not accuse Microsoft of infringing, even though it knew in fall 2002 that Word did exactly what it later accused of infringement. (A1703-09 (Vulpe); A802 (Owens) (admitting "couldn't say that Word infringed"); A2154-59 (Sweet).) Far from it, Mr. Vulpe congratulated Microsoft on its introduction of the now accused custom XML functionality into Word 2003. A5930. More than four years passed before i4i Inc.'s investors, seeking to "achieve liquidity," got additional investors to fund this lawsuit and sold the patent to a newly minted patent-enforcement vehicle, called i4i L.P., which subsequently sued Microsoft. (A5595-98; A5755.)

The invention claimed by the '449 Patent is an improved method for editing markup-language documents by "extracting metacodes from an existing document and creating a map of the location of the metacode in the document and then storing the map and the content of the document separately." (A2796.) The patent teaches that the metacodes are stored in a "metacode map"--a "separate entity" from the content (A2796) that matches each metacode to its location. (A250-52.)

9

A user can then look at the metacode map to determine where each metacode belongs in the stream of content. This, as the patent's title suggests, allows the '449 invention to achieve the key goal of "independent manipulation," in which the user may edit the structure of the document (i.e., the metacodes) by accessing only the metacode map, without ever needing to access (or have access to) the content. (A252 at col. 7:6-10; A2796.)

i4i agrees that the '449 Patent did not "invent the concept of the metacode map" (A1728 (Vulpe); see also A1725-28 (Vulpe); A2337-39 (Rhyne); A2841- 42.) Instead, as the district court's original claim construction recognized, independent manipulation was a key to the inventionthat order specifically stated that the specification and prosecution history "require the claimed ... method to differentiate between the stored metacode map and mapped content" such that "different process and users could edit the metacode map and mapped content independently and without access to both the metacode map and the mapped content." (A101.) Shortly before trial, however, the court "clarified" its claim construction, finding that independent manipulation was no longer a requirement but merely "one benefit of the invention" (A72), thus eliminating one of Microsoft's central noninfringement arguments.

10

At i4i's urging, the court also construed the claim terms "distinct map storage means" and "mapped content distinct storage means" 1 to mean merely "a portion of memory for storing a metacode map" and "a portion of memory for storing mapped content," respectively, thereby reading the key claim term "distinct" out of the claims. (A116.) The court adopted this countertextual construction of "distinct," even though i4i, when faced with the examiner's rejection of the patent because "storage i[s] always distinct" (A2786), assured the examiner that the architecture of the document (i.e., the metacode map) is "an entity having distinct storage" because the metacode map is stored separately from the mapped content. (A2796.)

Armed with a newly broadened claim construction at trial, i4i asserted infringement of three method claims. i4i did not allege that use of Word necessarily infringes, or even that use of Word's XML development platform necessarily infringes. Instead, i4i alleged that Word users infringe the '449 Patent only when they use Word to open files of certain formats (.xml, .docx, or .docm) that contain custom XML, asserting that when used in this manner, Word separates the custom XML tags from content and stores them in the manner claimed by the

1 Further confirming the effective deletion of "distinct," the district court construed "mapped content distinct storage means," as in claim 2, identically to "mapped content storage means," as used in the Asserted Claims. (A116.)

11

'449 Patent. There is no dispute, however, that opening files in the more familiar ".doc" and ".dot" formats--even if such files contain custom XML--is not an infringing act. (A4584-86.)

Relatively few Microsoft employees opened custom XML documents in Word. (A1435-36.) i4i thus rested its nine-figure damages case upon allegations of indirect infringement. i4i's proof of the requisite scienter rested on two marketing documents, prepared by i4i and provided to certain Microsoft employees, that referenced the '449 Patent merely by number. The fact that a handful of Microsoft employees received the patent number was sufficient, i4i argued (and the court agreed), to establish that Microsoft had knowledge not only of the contents of i4i's patent, but also specific knowledge that use of Word's custom XML development platform actually infringed that patent. According to i4i's expert, this was because the law required a duty to investigate. (A1210 (explaining that Microsoft "should have looked at ... the patent"); A1297 (explaining that he "would direct somebody to investigate"); see also A1294-97.)

Because the only asserted claims were method claims, i4i had to prove actual use of the method. To attempt to quantify alleged damages based on the number of actual users of the accused functionality, i4i commissioned Dr. William Wecker to perform a telephone survey. Wecker contacted 988 businesses, but even offering money was able to entice only 46 companies to respond. (A1602;

12

A4245.) The survey did not ask about respondents' own use of custom XML, but asked respondents to "estimate" how many other employees in their companies used Word and custom XML over a 5-year period. (A1578-83; A4244-58.) And even then, of the 46 respondents, just 19 reported any knowledge of any use of Word's custom XML functionality within their organizations. Not surprisingly, the handful of responses (all oral) had many internal inconsistencies. Undeterred, Wecker "corrected" the inconsistent answers himself using "data imputation" based on his assumptions as to what the respondents meant. (A1602-13.) Wecker then used the adjusted responses to conclude that, as of November 2008, about 1.85 million installations of Word 2003 and Word 2007 had been used to open and save custom XML documents in the pertinent file formats. (A1533-35; A1561- 66.) i4i's damages expert, Michael Wagner, increased this number to 2.1 million users through trial, claiming that every day some 1,525 new users infringe the '449 Patent for the first time. (A1394-95; A4020; A5931-36.)

With this base of users premised on the Wecker survey, Wagner went about positing a "reasonable royalty." However, he did not use the price of any accused version of Word or even i4i's own products (supposedly covered by the '449 Patent) as the basis for his analysis. Instead, he turned to a third-party product, XMetaL, that has never been accused of infringement and has a list price of $499--roughly triple the entire retail price of most versions of Word. (A1467-69;

13

A3938.) He applied Microsoft's profit margin to the XMetaL list price, and then applied the supposed "25% rule of thumb" to end up with a per user "reasonable" royalty of $96 (which he then increased to $98 based on his analysis under Georgia-Pacific). (A1469-70.) Thus, Wagner claimed that in a hypothetical negotiation, Microsoft would have agreed to pay i4i approximately 60 percent of the current retail price of the professional edition Word 2007 and more than the entire retail price of some other editions of Word. (A3928.)

Microsoft moved to exclude both the results of Wecker's survey and Wagner's Georgia-Pacific analysis arguing that they lacked the reliability required of expert evidence and did not fit the facts of the case. (A5944-45; A6018-19; A415; A439.) The district court summarily denied those motions (A415; A439), and further denied Microsoft's subsequent motion at trial to strike the opinions. (A1619-20.)

With no gatekeeper in sight, Wagner told the jury that damages of approximately $200,000,000 would constitute a reasonable royalty for Microsoft's alleged indirect infringement. Soon thereafter, the jury rendered a verdict that Microsoft had willfully infringed and awarded exactly the $200,000,000 that Wagner had suggested was a "reasonable" royalty. (A236-37.)

After a hearing, the district court denied all of Microsoft's post-trial motions. (A3-69.) Based on the jury's finding of willful infringement, the district court

14

enhanced the $200,000,000 damage award by $40,000,000. (A48.) The court concluded that enhanced damages were warranted in part because "Microsoft's 2008 revenue was $60.42 billion" and the jury award is but "a small fraction of the profit that Microsoft has gained from sales of its WORD products." (A44.) Also weighing in favor of enhancement, in the court's view, were Microsoft's counsel's "legally improper" arguments that this non-practicing patent owner might better be analogized to a banker seeking a bailout than a patentee protecting a valuable product--arguments to which i4i did not object when they were made, yet the court felt compelled "to temper ... with a specific instruction to the jury." (A46- 47.) But the court did not similarly offer a curative instruction when i4i's expert trumpeted Microsoft's supposed "monopoly power" (A1452), against the in limine order on this issue (A5643), which i4i played up by repeatedly referring to Microsoft's supposed "power" and the ability to "dictate" the terms of a negotiation in the real world. (A1503.)

The district court also issued a permanent injunction that, as of October 10, 2009, effectively will prohibit Microsoft from selling any Word products that can open .xml, .docx, or .docm files containing custom XML--which is to say, all versions of Word and Office currently available. (A3-4.) The court ruled that even though i4i's product is currently only an add-on to Word--and thus, as i4i's own witnesses agreed, "neither Microsoft Word 2003 [n]or Microsoft Word 2007 today

15

competes directly with i4i's product" (A1028; see also A1026-28)--an injunction was nonetheless warranted because of past harm. (A53.) The court found damages for such injuries to be incalculable despite the fact that the court (at i4i's strong urging) imposed post-verdict damages of $144,060 per day for the period between verdict and judgment. (A48-50.)

Microsoft appeals.

SUMMARY OF ARGUMENT The judgment below rests on several errors of law. It should be reversed or vacated.

A fundamental error underlying the entire judgment is the district court's claim construction, which stripped a key limitation from the claims and allowed a finding of infringement where none is warranted. The district court's construction rendered the limitation "distinct map storage means" meaningless by extending it to any "portion of memory," just as it did with the related term "mapped content distinct storage means." As the applicants' exchanges with the patent examiner made clear, however, "distinct" storage in the '449 Patent cannot mean merely storing the metacode map in a separate memory location from the mapped content, because all storage is separate in that sense. Rather, the applicants repeatedly stated in the intrinsic record that "distinct" storage is sufficiently separate that it allows for independent manipulation. The district court's construction violated

16

this Court's precedent by failing to give those representations limiting effect. Because i4i's own experts admitted that Word does not store a metacode map in a separate file from content, and does not allow for independent manipulation of a metacode map and mapped content, the jury's infringement verdict should be reversed.

The district court also erred in failing to set aside the jury's verdict of validity on grounds of obviousness. In obtaining its patent, i4i persuaded the PTO that the prior art failed to teach the elements of "persistent" metacodes and a "menu of metacodes." The district court disagreed with Microsoft's arguments that the claims require "persistence," and recognized only a "menu" as a limitation. Microsoft presented unrebutted evidence that the allegedly missing "menu" was taught by two prior art references never considered by the PTO, and thus their combination with the previously cited art renders the '449 Patent obvious. Indeed, when confronted with these additional references in the pending reexamination of the '449 Patent, the PTO admitted its mistake and the claims now stand provisionally rejected. The district court's denial of JMOL on invalidity should be reversed.

Legal error also undermined the district court's rejection of the on-sale bar. The entire contemporaneous evidentiary record--consisting of i4i's own documents and the testimony of a disinterested witness--confirms that, well before

17

the critical date, i4i sold a product (S4) that embodied the invention later claimed in the '449 Patent. i4i's only opposition at trial was the conclusory, self-serving testimony of the inventors that the S4 product could not have embodied their invention because they allegedly did not conceive it until later. In dismissing the inherent incongruity between the inventors' trial testimony and the pre-litigation documents as merely a credibility dispute, the district court erroneously assumed that the corroboration rule--which this Court applies to preclude inventors from salvaging their patents with uncorroborated oral testimony--is limited to priority disputes. The policies underlying the corroboration rule apply equally to prevent an inventor from lying about a post-sale conception to avoid the on-sale bar.

The district court's errors on indirect infringement include a fundamental misapplication of the Patent Act. As this Court recently made clear, method claims are not indirectly infringed by selling "components." The proper test under Section 271(c) as applied to method claims requires a sale of a "material or apparatus," and the software licensed by Microsoft qualifies as neither. Contributory infringement was also fatally undermined by the substantial noninfringing uses for both Word and the accused functionality shown by i4i's own evidence. And as to both forms of indirect infringement, the district court's conclusions on scienter rested on the legally insupportable notion that, merely upon receiving marketing literature with i4i's patent number, Microsoft must have

18

known that Word would infringe and intended to infringe--even though the record was devoid of evidence that Microsoft ever actually saw the patent, much less had knowledge of any infringing use of Word.

The district court's legal errors are even more egregious with regard to damages. The need for effective gatekeeping was never more apparent than when i4i sought to predicate its damages case of $200,000,000 on an inherently unreliable "survey," which posed unanswerable questions, extrapolated millions of alleged users from only 19 paid respondents, and was riddled with inconsistent answers "corrected" (i.e., changed) by the surveyor. Over Microsoft's objections, however, the district court allowed i4i's damages expert to base his entire damages model on this survey. Contrary to the court's conclusion, infirmities in the expert's royalty analysis were not merely issues of weight subject to cross-examination. The survey results were inherently unreliable and, accordingly, the testimony of i4i's damages expert that relied upon the survey was inherently unreasonable and should have been excluded.

The legal errors underlying i4i's damages claim were further compounded by presentation of a $98 royalty rate derived from application of the questionable "25 percent rule-of-thumb" to an expensive third-party "benchmark" product-- triple the price of Word--that was never even alleged to embody the patent. Georgia-Pacific does not countenance results-oriented benchmarks that are not tied

19

to the claimed invention and are transparently chosen solely to inflate the royalty rate. And the resulting $200,000,000 award is not "reasonable" under any measure.

The district court's $40,000,000 enhancement of the $200,000,000 verdict was fatally premised on a distortion of the objective prong under this Court's Seagate test for willful infringement. Although Microsoft succeeded in invalidating 13 of the 20 patent claims before trial and further presented invalidity defenses at trial that the PTO has found sufficient in reexamination, the district court nevertheless concluded that those defenses were "irrelevant" because they would not have been "apparent" to Microsoft when infringement allegedly began. In other words, although the district court imposed omniscience on Microsoft when it came to "knowing" that it infringed upon receipt of merely the patent number, the district court unfairly assumed complete ignorance when it came to objectively reasonable defenses. This Court, however, has consistently recognized that Seagate's objective prong is to be assessed in accord with the full record of the infringement proceeding, which includes credible defenses developed for trial. The verdict of willfulness should have been set aside, thereby precluding any enhancement.

Finally, the district court violated the principles of eBay in entering a permanent injunction. i4i provided no evidence of irreparable harm from

20

Microsoft's continued sales of Word. Whatever losses i4i might have proved, they all occurred in the past and can be adequately remedied by monetary relief. Meanwhile, the losses that Microsoft faces from disruption of its flagship Word and Office products--of which the accused functionality is only a tiny fraction, with few users--would be irreparable and immediate.

STANDARDS OF REVIEW Claim construction, and grants or denials of JMOL, are reviewed de novo. Mahurkar v. C.R. Bard, Inc., 79 F.3d 1572, 1576 (Fed. Cir. 1996). Obviousness is an issue of law reviewed de novo (with any presumed factual findings reviewed for substantial evidence). Agrizap, Inc. v. Woodstream Corp., 520 F.3d 1337, 1343 (Fed. Cir. 2008). A verdict of infringement is reviewed for substantial evidence if it rests on a correct claim construction. Kearns v. Chrysler Corp., 32 F.3d 1541, 1547-48 (Fed. Cir. 1994). The district court's denial of a new trial and grant of an injunction, as well as its evidentiary rulings, are reviewed for abuse of discretion. Rotondo v. Keene Corp., 956 F.2d 436, 438 (3d Cir. 1992). Where a district court "makes an error of law," it "by definition abuses its discretion." Koon v. United States, 518 U.S. 81, 100 (1996).

21

ARGUMENT I. Claim Construction Errors Require Reversal Of The Judgment Of Infringement

The '449 Patent claims distinct means for storing a metacode map so that it can be edited separately from the content. The district court, however, erroneously read these requirements out of the patent. Because no reasonable jury could have found infringement under a correct construction of the Asserted Claims, the erroneous construction requires reversal or vacatur of the infringement judgment.

A. The District Court's Construction Reads "Distinct" Out Of The Claims' Requirement Of Storage Of The Metacode Map

A fundamental principle of claim construction is that claim limitations may not be read out of the claims. See Merck & Co. v. Teva Pharms. USA, Inc., 395 F.3d 1364, 1372 (Fed. Cir. 2005). The Asserted Claims require a "distinct map storage means." The applicants consistently emphasized the importance of "distinct storage" in the patent specification and during prosecution by stressing two core concepts of "the invention": (1) "distinct" storage of a document's metacode map from its content; (2) so each can be accessed and edited independently. (A2796, A2812, A2839, A2853-54.) Consistent with the intrinsic record, Microsoft proposed construing this term as requiring "persistently storing the metacode map separately and distinctly from the mapped content so that the metacode map can be edited directly without having access to the mapped

22

content." (A3910-3911; A3859-65.) Microsoft advanced a similar construction for the related term, "mapped content storage means." (Id.) The district court, however, construed these disputed terms as essentially just portions of memory. (A116.) The district court's unduly broad construction allowed i4i's expert to tell the jury that as long as Word stores the bits of the "metacode map" at merely different memory addresses than the bits of the "mapped content," the storage is "distinct." (A1265-71; A1283-84.)

That construction rendered the term "distinct" superfluous because, as the examiner recognized in the first office action, all storage is "distinct" in the sense that different bits are stored at different memory addresses. (A2786.) This Court "has denounced" such constructions. See, e.g., Stumbo v. Eastman Outdoors, Inc., 508 F.3d 1358, 1362 (Fed. Cir. 2007). Thus, just as the applicants represented to the PTO, "distinct" storage in the '449 Patent must mean something more than bits stored in different (separate) memory locations--and that something more requires storage in separate files to allow for independent manipulation. (A2796.) Under this proper construction as proposed by Microsoft, it is undisputed that Word cannot infringe.

A claim construction that violates the claim language, or clear definitions in the specification, or representations made to the PTO cannot stand; the district court's construction of the disputed terms here violates all three.

23

1. The Specification Expressly Defines "Distinct" Map Storage Means As Separated From Mapped Content Storage Means

Claims are to be construed consistently with the "purpose[] of the invention." Osram GmbH v. ITC, 505 F.3d 1351, 1358 (Fed. Cir. 2007). Where, as here, particular characteristics are described throughout the patent as essential to "the invention," they limit the claims. Verizon Servs. Corp. v. Vonage Holdings Corp., 503 F.3d 1295, 1308 (Fed. Cir. 2007) ("When a patent thus describes the features of the `present invention' as a whole, this description limits the scope of the invention").

The '449 Patent defines the purpose of "the invention" as separating out the metacodes from the content and placing them into a "metacode map," which is placed in storage "distinct" from the mapped content. (A250 at 4:5-10.) Distinct storage is uniformly depicted in the patent figures by showing the metacode map and mapped content as separately accessible units in "primary storage":

24

(A240.)

According to the patent, separating the metacode map and the mapped content into these "separately stored and protected" units allows changes to be made "solely on the metacodes" or "solely on the content." (A252 at 7:6-25.) Indeed, the patent declares that "[m]ost of the benefits flow from the fact that the invention recognizes the separateness of content and structure." (A251 at 6:18- 22.) This is not merely an embodiment or benefit of the invention; it is an essential characteristic of the invention.

25

2. The Applicants Represented To The PTO That "Distinct" Storage Requires Separate Files, Not Just Different Memory Locations

The applicants relied heavily on "distinct" storage to distinguish the prior art during prosecution, insisting that distinct storage of the metacodes in a "separate entity" from the content is a required aspect of "the invention":

In this invention, a document's architecture can be treated as a separate entity from the document's content.... This separation is achieved by extracting metacodes from an existing document (or from a document being created) and creating a map of the location of the metacodes in the document and then storing the map and the content of the document separately. (A2812; see also A2796.) Similarly, in response to the examiner's Final Rejection, the applicants distinguished the Kugimiya prior-art patent because it, inter alia, stored "metacodes intermixed with the content." (A2853.) In contrast, according to applicants, their invention required metacodes to be separated out and "managed separately" from the content. (A2854.)

The separate-file requirement is confirmed by the applicants' response to a rejection based on the Mizuta reference. Characterizing Mizuta as storing "all document information ... in one file--the document file," the applicants insisted that Mizuta "lacks any notion of a metacode map, let alone such a map in a `metacode map distinct storage means' as presently taught and claimed." (A2816.)

By construing "distinct map storage means" as merely "a portion of memory for storing a metacode map," without any requirement that the metacode map be a

26

"separate entity" (that is, separate file) from the "mapped content," the district court effectively read "distinct" out of the claims. This construction contradicts not only the language of the claims, but also clear statements in the intrinsic record. See, e.g., Omega Eng'g, Inc. v. Raytek Corp., 334 F.3d 1314, 1326-27 (Fed. Cir. 2003) (holding that patentee's repeated statements during prosecution regarding the invention created an unmistakable surrender of subject matter); Computer Docking Station Corp. v. Dell, Inc., 519 F.3d 1366, 1375-77 (Fed. Cir. 2008).

B. "Distinct" Storage Requires The Ability To Independently Manipulate The Metacode Map And Mapped Content

The element of "distinct" storage ties into another key requirement of the claimed invention--allowing a user to edit the structure of the document (i.e., the metacode map) by accessing this one data structure alone, without needing to access the content. This requirement--referred to by the parties as "independent manipulation"--is identified in the specification as another fundamental aspect of the alleged invention:

The present invention provides the ability to work solely on metacodes. The process allows changes to be made to the structure of a document without requiring the content [sic]. A metacode map could be edited directly without the mapped content. (A252 at 7:6-10.)

27

The district court's original claim construction order agreed with Microsoft that the '449 Patent requires differentiating between the metacode map and the mapped content to allow for independent manipulation:

In total, the statements [in the file history], in light of the specification, require the claimed computer system or method to differentiate between the stored metacode map and mapped content. In such a situation, different processes and users could edit the metacode map and mapped content independently and without access to both the metacode map and the mapped content. (A101.) A few months before trial, however, i4i submitted a technical expert report in which the expert took the surprising position that independent manipulation was merely a "benefit" of the invention but not a "requirement" of the claims. Microsoft raised this inconsistency at the pretrial conference, and the court allowed supplemental briefing. (A462-64, A466-71, A485, A491-92.)

In a supplemental opinion issued just before trial, the district court revised its earlier construction and adopted i4i's arguments that independent manipulation is merely "one benefit" of the invention, not a limitation on the claims. (A72.)

1. The Specification Defines "The Invention" As Allowing Independent Manipulation

Contrary to the district court's reasoning underlying its new construction, the applicants did draft the claims to require the "benefit" of independent manipulation--by modifying "map storage means" in the claims with the term "distinct." The intrinsic record explains that "distinct map storage means" includes

28

the ability to manipulate a metacode map independently from mapped content. Indeed, the notion of independent manipulation is embedded throughout the patent, beginning with the title of the patent: "Method and System for Manipulating the Architecture and the Content of a Document Separately from Each Other." (A239.) The Abstract likewise states that the patent is for a "system and method for the separate manipulation of the architecture and content of a document." (Id.) This concept is further emphasized in the Summary of the Invention, as noted above. (A252 at 7:6-25.)

2. The Applicants Repeatedly Distinguished The Prior Art On The Basis That Their Invention Allows Independent Manipulation

To obtain their patent, the applicants emphasized independent manipulation as a key requirement of their invention and in fact argued against the very construction the district court ultimately adopted. The examiner initially rejected as obvious the claims requiring "distinct storage," stating that "[s]torage [is] always distinct, even if at distinct addresses." (A2786.) Applicants rebutted that assertion by explaining that "distinct storage" was much more than simply distinct memory addresses because "using the present invention, one could change the architecture (layout, structure, or presentation format) of a document without even having access to the actual content of the document." (A2796; see also A2839, A2853-54.)

29

In this litigation, however, i4i contradicted these representations by asserting that the "metacode map distinct storage means" is simply a "portion of memory for storing a metacode map." The district court's adoption of i4i's empty construction cannot stand in the face of contrary and repeated representations to the PTO.

C. There Is No Infringement Under A Correct Claim Construction

1. Word Does Not Store The Metacode Map "Distinctly" From The Mapped Content

As noted, the claims require distinct storage of the "metacode map" from the "mapped content." It is not disputed that Word does not store the alleged metacode map and content in separate files; rather, Word stores the information for each document in a single file (in one of various formats, such as ".docx"). (A1356-57.) i4i argued, however, that it needed to prove only that different bits were stored at different addresses (A1179-80; A1265-71; A1283-84), which is of course always true in memory and is exactly what the Examiner said was meaningless. (A2786.)

Indeed, i4i's own infringement analysis confirms the lack of separateness. A "metacode map," as defined by the district court, is a "data structure." (A116.) But i4i was unable to identify in Word any single data structure as a "metacode map." Rather, based on the notion of "logical" relationships, i4i's expert (Rhyne) pieced together selected portions of seven different data structures in Word 2003 and Word 2007 to comprise what he opined to be a "metacode map." (A1249-78.)

30

i4i's strained identification of a "metacode map" as an amalgam of data structures is fatal to its infringement case under a proper construction of "distinct storage." As Rhyne agreed, there is no objective way to differentiate the linking of the structures he declared to be the "metacode map" from other structures, including the structure he declared to be the "mapped content" in Word. (A1262- 71.) Following Rhyne's "logical" relationships line of reasoning, therefore, one could equally combine many other data structures in Word to "identify" a "data structure" that contains both the alleged mapped content and the alleged metacode map--thus destroying the alleged distinctiveness of the "metacode map distinct storage means."

Indeed, i4i's expert did not even disagree with this conclusion, but insisted that it did not matter under the district court's construction because the only thing that had to be "distinct" was where the metacode map and mapped content were stored in memory--in other words, merely different addresses in memory. (A1266.) That very argument was expressly disclaimed by the applicants during prosecution, however, in response to the examiner's first rejection, as noted supra. (A2796.) Because such distinct storage within the meaning ascribed by the inventors during prosecution is not possible in Word (as further discussed below), correcting the court's claim construction to give meaning to the term "distinct" consistent with the intrinsic record requires reversal of the infringement verdict.

31

2. Word Does Not Allow A Metacode Map And Content To Be Manipulated Independently

The metacode map and mapped content purportedly identified by i4i in Word cannot satisfy the disputed limitations (properly construed) for another reason--there is no way to manipulate the structure of the document through the alleged metacode map independently from the content. This point was conceded at trial by one of i4i's technical experts. (A1374-75.) Because a change to one requires a change to the other, the alleged "metacode map" and "mapped content" in Word 2003 and Word 2007 cannot be independently manipulated. Therefore, because the intrinsic record requires incorporation of the independent manipulation requirement into the construction of the disputed terms, the infringement verdict must be reversed.

* * *

Where, as here, an infringement verdict rests on incorrect claim construction, and no reasonable jury could have found infringement under the proper construction, this Court may reverse a district court's denial of JMOL without remand. Harris Corp. v. Ericsson Inc., 417 F.3d 1241, 1255-57 (Fed. Cir. 2005); see also Seachange Int'l, Inc. v. C-COR Inc., 413 F.3d 1361, 1381 (Fed. Cir. 2005). At a minimum, remand for a new trial is warranted.

32

II. The Asserted Claims Are Invalid Every element of the Asserted Claims is disclosed in prior-art combinations, and thus the claims are invalid as obvious. The claims are also invalid under the on-sale bar.

A. The Claimed Invention Was Obvious

The district court erred in refusing to vacate the jury's verdict of validity in light of clear and convincing evidence that every step of the Asserted Claims is met by the combination of either the Rita or DeRose prior art with Kugimiya. Indeed, this has been confirmed by the PTO's rejection of the Asserted Claims in reexamination based on these references. (A4464-82.)

The ultimate judgment of obviousness under 35 U.S.C. § 103 "is a legal determination" and is therefore ripe for resolution in this case as a matter of law. See KSR Int'l Co. v. Teleflex, Inc., 550 U.S. 398, 416 (2007). This Court reviews "a jury's conclusions on obviousness, a question of law, without deference." Muniauction, Inc. v. Thomson Corp., 532 F.3d 1318, 1326-28 (Fed. Cir. 2008) (reversing denial of JMOL in favor of defendant on obviousness); see also Boston Sci. Scimed, Inc. v. Cordis Corp., 554 F.3d 982, 991 (Fed. Cir. 2009) (same); Ecolab, Inc. v. FMC Corp., 569 F.3d 1335, 1349 (Fed. Cir. 2009) (same); Agrizap, 520 F.3d at 1342-44 (same).

33

As the Court recognized in KSR, "[t]he combination of familiar elements according to known methods is likely to be obvious when it does no more than yield predictable results." 550 U.S. at 416. Such is the case here.

1. Each Limitation Was Disclosed In Obvious Combinations Of The Prior Art

1. The applicants' only bases for distinguishing the Kugimiya patent (A3721-41) before the PTO was that Kugimiya allegedly did not "persist" the metacode map long enough for the user to manipulate the metacodes, and that Kugimiya did not disclose a menu of metacodes. (A2842-43; see also A1727-28.) Because the district court did not require "persistent" storage in its claim construction--even though requested by Microsoft (A98-99)--the question of obviousness therefore boils down to whether it would have been obvious to merely add a "menu of metacodes" to Kugimiya.

i4i itself has already answered this question in the affirmative--a declaration from i4i's own expert submitted in opposition to summary judgment described metacode menus as "ubiquitous to software systems, and well-known in the art" at the time of the '449 Patent application. (A4427 at ¶ 23.) There can be no question that adding a ubiquitous feature (menu) to a known technology (an SGML document processor/translator that used a map of metacodes) would have been obvious to provide the benefits of that feature. See, e.g., Leapfrog Enters., Inc. v. Fisher-Price, Inc., 485 F.3d 1157, 1162 (Fed. Cir. 2007).

34

2. Moreover, i4i never rebutted Microsoft's evidence that both of these features--persistent metacodes and a menu of metacodes--are present in both the Rita and DeRose prior art references. Like the '449 Patent, the Rita SGML editor enabled users to create and edit tagged documents, such as SGML documents. (A2890-2915.) It provided a user interface in which SGML tags were displayed in a window separate from the document content, allowing independent access to either the SGML tags or the document text. (A2895-96; A2051; A3244.) The Rita SGML editor is prior art under 35 U.S.C. § 102(b), as it was disclosed in several references published in 1987, 1990, and 1991, all well before the critical date. (A2890-2915; A2916-26; A2927-3043; A3044-3226; A3227-3412.) Although the district court noted a dispute concerning precisely what version of Rita source code was sold prior to the critical date, the testimony of Microsoft's expert, Mr. Gray, established that the Rita publications themselves--which are all indisputably prior art--disclose each and every limitation of the claims. (A2051-60.) Any dispute over source code is thus immaterial to this issue on appeal.

DeRose, which has a priority date of July 19, 1991, is prior art under 35 U.S.C. § 102(e). (A3431-68.) DeRose disclosed a system for generating, analyzing, and navigating electronic documents containing markup, in particular SGML documents. (A3457 at 3:18-32.) The processing of SGML documents

35

taught by DeRose entailed the separation of SGML tags from the content of the document into separate data structures. (See, e.g., A3460 at 10:36-13:2; A2061.)

According to the district court, i4i asserted that three limitations are absent from the Rita and DeRose references: the "metacode map," "addresses of use," and "mapped content." (A31.) 2 In particular, i4i insisted that these references did not disclose "metacode maps" because they stored tag information in a "tree" data structure, and that accordingly there is no corresponding "mapped content." (A2282-92.) i4i also insisted that Rita and DeRose did not have "addresses of use" because their tree data structures used pointers (which are memory addresses). (A2286-88.)

But these arguments fail because they are not based on the district court's claim construction. Under the court's construction, both references undeniably satisfy the "metacode map" and "address of use" limitations. The district court defined the "metacode map" as "a data structure that contains a plurality of metacodes and their addresses of use corresponding to a mapped content." (A116.) i4i's technical expert acknowledged that trees, like the metacode map described in the '449 specification, separate metacodes from the content. (A2286.) He also 2 Gray systematically compared every step of the Asserted Claims to the Rita and DeRose prior-art publications and software. (A2051-67.) i4i's expert acknowledged that Rita and DeRose both separated metacodes from content and stored those codes in data structures with pointers to the content. (A2283-85.)

36

acknowledged that the data structures containing the SGML tags contained pointers from those tags to the content. (A2283-85.) Thus, it is undisputed that the trees contain the metacodes (SGML tags) and addresses that, according to i4i's expert, "identif[y] the place in memory where the text stream [marked by the SGML tag] lives." (A2287.) Such a structure meets the construction of "metacode map." Indeed, i4i admitted that its own products, which purportedly embody the '449 Patent, also use a tree structure. (A2320.)

3. Even if i4i's argument were accepted, however, it cannot avoid invalidity. There is no dispute that either Rita or DeRose, when taken together with the teachings of the Kugimiya patent, disclose all of the limitations of the Asserted Claims. The Kugimiya patent undeniably disclosed a metacode map (and thus, by the very definition of "metacode map," "mapped content" and "addresses of use" as well). This fact was acknowledged by the applicants during prosecution (A2842-44), as well as at trial by Vulpe and by two i4i experts. (A1727-28 (Vulpe); A2337-38 (Rhyne); see also A2561(Rappaport).)

The district court concluded that the jury was entitled to reject this combination on the ground that Kugimiya "was in a different field"--that is, "translation" instead of "document processing." (A32.) Such holding represents an insupportable view of the field of the invention, and of the ordinary creativity of one of skill in the art. All three references disclosed systems for processing SGML

37

documents, and they operated in a similar manner. (A2890-2915; A3431-3468 at 3:26-29; A3721-41 at 7:23-35.) They were also all directed at a similar problem-- the separation of metacodes and mapped content in electronic documents, irrespective of whether they did so with the aid of a metacode map or a tree. Indeed, even i4i's expert acknowledged that both DeRose and Rita separated content and metacodes. (A2281-85.) Given this problem, and the "finite" number of references addressing it, "a person of ordinary skill has good reason to pursue the[se] known options." KSR, 550 U.S. at 421.

2. Secondary Considerations Do Not Suffice

Even if the Court were to accept that i4i has put forth relevant evidence of secondary considerations, those secondary considerations cannot, as a matter of law, defeat the overwhelming prima facie evidence of obviousness in this case, and JMOL in favor of Microsoft is warranted. Agrizap, 520 F.3d at 1344 ("[T]he objective evidence of nonobviousness simply cannot overcome such a strong prima facie case of obviousness."); Muniauction, 532 F.3d at 1327.

This Court's recent decision in Boston Scientific is instructive. There, as here, the jury returned a verdict of non-obviousness and the district court denied the post-verdict motion for JMOL. 554 F.3d at 992. This Court reversed, concluding that the combination was obvious, and that, "given the strength of the prima facie obviousness showing, the evidence on secondary considerations was

38

inadequate to overcome a final conclusion that [the claim] would have been obvious." Id. (citations omitted; alteration in original). The Court also made clear that "[w]e are free to override the jury's legal conclusion on the ultimate question of obviousness without deference." Id. The jury's verdict of non-obviousness here should similarly be reversed.

B. The On-Sale Bar Applies

The district court's denial of JMOL on anticipation by the SEMI S4 System under the "on-sale bar" of 35 U.S.C. § 102(b) rested on legal error--namely, refusing to require corroboration of inventors' conclusory testimony of an alleged invention date that was offered to contradict the contemporaneous evidentiary record showing the sale of their invention before the critical date. While this Court has not yet addressed whether the corroboration rule applies to a situation in which an inventor seeks to post-date his conception to avoid an on-sale bar, the policies underlying that rule are as compelling here as they are in the typical priority dispute.

The only dispute underlying the on-sale bar was whether the SEMI S4 System--which everyone agreed had been sold to SEMI before the June 2, 1993,

39

critical date 3--embodied the claimed invention. The contemporaneous evidentiary record established a prima facie case that the SEMI S4 System as sold did embody the '449 Patent.

First, the SEMI S4 User Guide showed that the SEMI S4 System as sold before June 6, 1993, was--just as the system described and claimed in the '449 Patent--a system for creating, opening, editing, and storing documents containing SGML metacodes, and allowed the metacodes to be manipulated separately from the content. 4 (See, e.g., A3472-3474, A3572-3573, A3578-3582.) In fact, the same hardware platform described in the SEMI S4 manual is the very hardware platform described as the '449 Patent preferred embodiment. (A251 at 5:3-5; A3480; A829-31 (Owens).)

Second, in a March 1994 letter to potential investors, Vulpe wrote: "I am currently exploring the patenting of some fundamental ideas used in the Infrastructures technology. The basis of the patent and the preliminary work on the vali[d]ation precedes Infrastructures." (A2882.) Vulpe admitted at trial that, because i4i was founded by early 1993, he was telling these potential investors that

3 A26; see also A761, A806 (Owens); A1628-31 (Vulpe); A1678 (Vulpe); A1964- 69 (Young); A2875-77; A2884-89 (SEMI Work Plan). 4 Microsoft also presented the testimony of its technical expert, limitation-by- limitation, to demonstrate anticipation by SEMI S4. (See A2070-80 (Gray).)

40

the basis of the '449 Patent dated back to precisely when the SEMI S4 System was being developed--not after. (A1695-99.)

Third, in August 1994, Canada-based i4i submitted a funding application to the Canadian government, in which it touted its application for the '449 Patent as having an "initial implementation" that was "embedded into Infrastructure's S4 product." (A3759-3768; A1688-91, A1773 (Vulpe).) One month later, i4i submitted another portion of that funding application, which stated that the '449 Patent's "single metacode model" was "implemented in i4i flagship product S4 vertical market product." (A3770; A1691-94 (Vulpe).)

Fourth, Vulpe admitted at trial that the SEMI S4 System was i4i's only product at the time of the Canadian funding application, and both inventors further confirmed that there were no changes to its architecture after being installed in early 1993. (A1679-80, A1687-88, A1691-92, A1776-77; see also A794-95 (Owens).) In addition, the funding application described the patent-pending S4 product as "targeted to the semi-conductor and publishing industries," which Vulpe conceded referred to his customer SEMI. (A1691; A2580-82 (Vulpe).)

Fifth, the SEMI S4 System's embodiment of the '449 Patent was confirmed by a disinterested witness at trial--Scott Young, a former employee of both SEMI and i4i. Young testified that Vulpe told him that the '449 Patent application was filed to cover the SEMI S4 System. (A1977-81.) Young also testified that after he

41

left SEMI to join i4i, he and Vulpe touted the patenting of the SEMI S4 System in attempting to obtain funding for i4i's business. (A1979-81, A1983-85.)

Sixth, although many technical details of the SEMI S4 system were lost before this litigation when i4i destroyed its source code, 5 Young also confirmed that the SEMI S4 System provided a mapping between the content and SGML tags (i.e., metacodes) of an SGML document opened by a user of the SEMI system, as claimed in the '449 Patent. (A1971-74.) At trial, Mr. Vulpe agreed that if Young's unrebutted testimony was accepted, then the '449 Patent should be invalidated. (A1685-86.)

The inventors had little to say against the backdrop of this compelling contemporaneous evidentiary record. In conclusory fashion, and with no corroboration, Vulpe and Owens simply insisted that the prior-art SEMI S4 System could not have embodied the '449 Patent because they had not conceived of that invention at the time the SEMI S4 System was sold. (A1682-84 (Vulpe); A839 (Owens).) Amazingly, Vulpe tried to explain the glaring inconsistency between his conception testimony at trial, and the real evidence of conception found in the

5 Vulpe insisted that the hard drive containing all of the source code for his company's original product was simply "discarded" because they "weren't obligated to keep it anymore." (A1771-72.)

42

contemporaneous evidentiary record, by claiming that he lied in the pre-litigation letter to investors to further his financial interests. (A1696-98.)

2. i4i's mere denials cannot rebut Microsoft's strong prima facie showing of on-sale bar. U.S. Envtl. Prods. Inc. v. Westall, 911 F.2d 713, 716 (Fed. Cir. 1990) ("Once a defendant demonstrates a prima facie case of on-sale or public use, the patent holder must `come forward with convincing evidence to counter that showing.'"). Although the district court accepted the notion that i4i's own destruction of the source code for the S4 System precluded the parties' trial experts from knowing how it worked, 6 the absence of that evidence should strengthen Microsoft's case, not diminish it. Given i4i's many contemporaneous pre- litigation admissions, coupled with key concessions of the inventors at trial-- countered only by uncorroborated testimony of interested inventors--Microsoft has made a prima facie case that S4 embodied the invention.

The district court rejected the need for corroboration of the inventors' conclusory, self-serving testimony on the misplaced notion that the inventors were not affirmatively trying to prove conception, as in a priority dispute, but were instead responding to Microsoft's anticipation defense. (A27.) But the legal requirement for corroboration of an invention date that is asserted to avoid the

6 A27-28; A1771-72 (Vulpe); A2299 (Rhyne); A2512 (i4i Closing).

43

invalidity of a patent is not a mere credibility issue. Rather, the corroboration rule provides a gatekeeping function, recognizing that oral testimony by an interested witness attempting to salvage a patent is inherently untrustworthy because it offers "great temptation to perjury" and "would have the effect of virtually precluding the adverse party from the possibility of rebutting such evidence." Price v. Symsek, 988 F.2d 1187, 1194 (Fed. Cir. 1993); see also Juicy Whip, Inc. v. Orange Bang, Inc., 292 F.3d 728, 740-41 (Fed. Cir. 2002).

Indeed, the importance of this Court's corroboration requirement can hardly be more apparent than in the present case, where an inventor has not only admitted that his self-interested trial testimony is contradicted by pre-suit documents, but has brazenly asked the jury to accept that he was lying in the past, but is telling the truth now. Without corroboration, any inventor with a little-known product could avoid an on-sale bar by destroying the source code and product, waiting years to file suit, and then lying about a post-sale conception date. This Court should hold that this conclusory testimony cannot support the jury's verdict as a matter of law, and grant JMOL of anticipation.

C. The Judgment Of Validity Cannot Stand

For the reasons discussed above, reversal and entry of judgment is warranted. At minimum a new trial is required because the verdict is against the

44

great weight of evidence. 7 In addition, the court's instruction on the clear-and- convincing burden of proof was faulty. As the Supreme Court has acknowledged, the rationale underlying the "clear-and-convincing" standard is "much diminished" when a defense of invalidity rests on prior art that the PTO did not consider. KSR, 550 U.S. at 426. Although this Court has been reluctant to embrace a lower standard in such cases, all the regional courts of appeals uniformly did so before this Court was created. See, e.g., Baumstimler v. Rankin, 677 F.2d 1061, 1066 (5th Cir. 1982) ("[T]hus, the challenger of the validity of the patent need no longer bear the heavy burden of establishing invalidity either `beyond a reasonable doubt' or `by clear and convincing evidence.'"); Mfg. Research Corp. v. Graybar Elec. Co., 679 F.2d 1355, 1364 (11th Cir. 1982). In this case, the clear-and-convincing standard was stressed by i4i's technical expert and its counsel in argument, and was set forth in the court's instructions to the jury, over Microsoft's objection. (A2264; A2447-48; A2279-80.) That was error, since the relevant evidence was never presented to the PTO (which, not coincidentally, has since found the claims 7 i4i's only rebuttal to Microsoft's clear showing, discussed above, that all elements of the claims are practiced by both Rita and DeRose, when combined with Kugimiya, amounted to an illusory distinction between "trees" and "maps" and an unsupported assertion that pointers (i.e., memory addresses) are not "addresses of use," that find no support in the claim construction. And its only rebuttal to the on-sale bar was the inventor's conclusory testimony that he had invented it after the sale, thus lying in contemporaneous admissions to potential investors.

45

invalid on the basis of that very evidence). A new trial is warranted in which the jury is instructed under a preponderance standard.

Microsoft is also entitled to a new trial because the Court did not permit Microsoft to present to the jury evidence concerning the PTO's recent grant of reexamination of the '449 Patent. At trial, i4i not only touted the PTO's allowance of the '449 Patent over Kugimiya, but also repeatedly asserted that the PTO had validated i4i's invention through its use of i4i's products. (See A1519- 28; A1664-65.) Microsoft should have been able to explain to the jury that the PTO has granted reexamination in light of Rita and DeRose, the very references that are now at issue. (A5622-5631.) The false impression that the PTO blessed the '449 Patent or that it would have allowed the currently Asserted Claims over these references was highly prejudicial to Microsoft. III. The Jury's Finding Of Indirect Infringement Cannot Be Sustained The district court denied Microsoft's request for a special verdict form (A2271-72; A5749-50), and consequently, when the jury found infringement, it did not specify the theory or theories under which it found Microsoft liable. The $200,000,000 damages award, however, leaves no doubt that the jury relied on a theory of indirect infringement--contributory infringement, induced infringement, or both--because even applying Wagner's monstrous $98 per unit royalty, the 1,306 alleged instances of direct infringement (by Microsoft personnel) would

46

garner a damage award of only $127,988. (A1435-36.) But unless i4i can sustain findings of both contributory infringement and induced infringement, this Court must award Microsoft a new trial on infringement (and damages) because, in the face of a general verdict, "the reviewing court cannot determine whether the jury based its verdict on a sound or unsound theory." Rutherford v. Harris County, 197 F.3d 173, 185 (5th Cir. 1999); see also Spectrum Sports, Inc. v. McQuillan, 506 U.S. 447, 451 (1993) (reversing general verdict where jury was erroneously instructed on one of three theories of liability).

A. The District Court Erroneously Instructed The Jury That Microsoft Committed Contributory Infringement If It Sold A "Component" Of i4i's Claimed Methods Section 271(c) prohibits the sale of a "component" of a patented product, or a "material or apparatus for use in practicing a patented process," knowing "the same" to be not "suitable for substantial noninfringing use." 35 U.S.C. § 271(c).

Over Microsoft's repeated objections (A4649-50, 4657-59; A5692, n.45), the district court insisted upon instructing the jury that it could hold Microsoft liable for contributory infringement it if found that Microsoft sold a "component for use in practicing the patented method." (A213.) But, as the en banc Court recently confirmed, the plain language of Section 271(c) compels the conclusion that, as applied to method patents such as the patent-in-suit, it prohibits only the sale of "a material or apparatus," and "a material or apparatus for use in practicing

47

a patented process is not a component of that process." Cardiac Pacemakers, Inc. v. St. Jude Med., Inc., --- F.3d ----, 2009 WL 2516346, at *13 (Fed. Cir. Aug. 19, 2009). "The components of the process are the steps of the process," and it is a "physical impossibility" to sell or otherwise "supply" such "components." Id. at *13-14.

The district court's instructional error was prejudicial here because the custom XML functionality that i4i and the district court contend that Microsoft "sold" indisputably is not a "material or apparatus." It is software code--a set of instructions that tell a computer how to use Word to provide users with XML editing capability. See Ricoh Co. v. Quanta Computer, Inc., 550 F.3d 1325, 1335 (Fed. Cir. 2008) (software is "instructions to perform a process"); Microsoft Corp. v. AT&T Corp., 550 U.S. 437, 452 (2007) ("Abstract software code is an idea without physical embodiment"). The only circumstance in which the custom XML functionality could ever be reduced to a "material or apparatus" is when, as a part of Word, it is encoded onto a disk. But it is undisputed that a disk encoded with Word has thousands of substantial noninfringing uses; contributory infringement liability therefore could not possibly have been predicated on Microsoft's sale of that "material or apparatus." The jury's verdict thus cannot be permitted to stand on a finding of contributory infringement.

48

B. Microsoft Is Entitled To Judgment As A Matter Of Law On Contributory Infringement Because The Software It "Sold" --Word--Indisputably Has Substantial Noninfringing Uses Under Section 271(c), a substantial noninfringing use precludes a finding of contributory infringement. This Court has recognized that, in determining whether a material or apparatus is "suitable for substantial noninfringing use," one must not limit the inquiry to a "mere ingredient" or functionality of the article, but rather must focus on "what was actually sold" by the defendant. Hodosh v. Block Drug Co., 833 F.2d 1575, 1578-79 (Fed. Cir. 1987); see also AquaTex Indus., Inc. v. Techniche Solutions, 419 F.3d 1374, 1379-80 n.** (Fed. Cir. 2005). This is because, for contributory infringement liability to attach, the "commodity of commerce [not] suitable for substantial noninfringing use" must be the "same" as the "material or apparatus" sold by the defendant. 35 U.S.C. § 271(c).

A straightforward application of these principles to the undisputed facts would compel the conclusion that Microsoft does not violate Section 271(c) because that which it "sells"--Word--has substantial noninfringing uses. Yet, over Microsoft's objection (A5693), the district court instructed the jury that, in evaluating noninfringing uses, it should focus on "the component itself, not the product in which the component is embedded." (A2382.) And later, denying Microsoft's post-trial motion for judgment as a matter of law, the court confirmed its view--quite to the contrary of this Court's view in Hodosh--that noninfringing

49

use must be "analyzed within the context of the accused feature and not the product as a whole." (A15 n.5.)

To justify a narrow focus on Word's custom XML functionality, the district court pointed to Ricoh, but its reliance on that decision was misplaced. Ricoh found an exception to the textual command of Section 271(c) for circumstances where it is conceded that the subpart that is actually used to infringe is "separable" and could, if sold on its own, be used to infringe. This is not the case here; Word's custom XML functionality runs only as part of the Word platform, cannot run separately from Word, and is not sold by itself. Further, Ricoh addressed the inclusion "of a component with substantial noninfringing uses" in a larger "product that contains other components useful only to infringe." 550 F.3d at 1339. Word presents the opposite situation; the accused custom XML functionality is not the larger product to which the rest of Word was appended. Rather, custom XML functionality is a small part of Word.

If, as the district court held, Ricoh necessarily applies to these facts, then it conflicts with the plain text of the statute and should be overruled. Word's custom XML functionality is not the "same" commodity as that sold by Microsoft here, nor is it a "commodity" sold in commerce at all. 35 U.S.C. § 271(c).

50

C. i4i's Own Evidence Demonstrates That Even The Accused Functionality Of Word Has Substantial Noninfringing Uses

The parties' experts agreed that Word's custom XML functionality could be used in three noninfringing ways. This should have been sufficient to establish as a matter of law that Word's custom XML functionality was "suitable for substantial noninfringing use." 35 U.S.C. § 271(c). Yet, the district court rejected that view, based on the testimony of i4i's expert that the noninfringing uses "subverted" what that expert determined to be "essential advantages of XML." (A16.)

This was an error of law. Whether a noninfringing use is "substantial" "depends on how likely and often the use will occur." Hoffmann-La Roche Inc. v. Promega Corp., 33 U.S.P.Q.2d 1641, 1648 (N.D. Cal. 1994) (cited by district court at A16). Here, far from "occasional, farfetched, impractical, experimental, or hypothetical," (A16), i4i's own survey demonstrated that more than 2 million users of Word's custom XML functionality saved their documents in noninfringing file formats (e.g., the familiar ".doc" format). (A4243.) If the survey is legally sufficient evidence of millions of infringing uses (in the context of damages), then it must also be legally sufficient to establish even more noninfringing uses. i4i's expert's suppositions about the "impracticality" of noninfringing uses "given the purpose of the invention" are answered by i4i's own evidence showing millions of noninfringing uses. i4i cannot pick and choose which of the Wecker survey's

51

conclusions are to be believed, and its assertion of the Wecker survey's reliability therefore should have precluded the issue of contributory infringement ever from going to the jury.

D. The Inferences Of Intent Are Legally Insufficient To Support Either Induced Or Contributory Infringement

Both of i4i's indirect infringement theories fail for the additional reason that i4i has failed to adduce substantial evidence of the requisite mental state. See Kyocera Wireless Corp. v. Int'l Trade Comm'n, 545 F.3d 1340, 1354 (Fed. Cir. 2008) (induced infringement requires specific intent to cause another's direct infringement); Aro Mfg. Co. v. Convertible Top Replacement Co., 377 U.S. 476, 488 (1964) (contributory infringement requires knowledge that sale will result in direct infringement).

Here, it is undisputed that i4i did not give Microsoft any pre-suit notice of potential infringement, and there was no evidence that anyone at Microsoft had read the '449 Patent or otherwise understood its contents. (A1021-22 (Thomas); A1294-96 (Rhyne); A1700, A1712-13 (Vulpe).) i4i's only evidence going to Microsoft's knowledge was one marketing document, authored by i4i and sent to a handful of Microsoft employees, that stated that i4i's advertised product was "patented," and referred to the patent number (A5639; A966-67), and one email stating that Word 2003 would make i4i's product (not the patent) "obsolete."

52

(A5640.) Neither of these documents discussed the content of the patent or which features of i4i's advertised product were patented.

Based on just the reference to the patent number, the district court held that the jury could conclude that Microsoft "knew that its WORD products containing XML functionality would infringe the '449 patent." (A15.) 8 This was an error of law.

Although (part of) the requisite intent for induced or contributory infringement can be established by showing that a defendant had "knowledge of the patent," Ricoh, 550 F.3d at 1342, the requisite "knowledge" is of the contents of the patent, not merely its number. The cases cited in Ricoh bear this out: In Water Technologies Corp. v. Calco, Ltd., 850 F.2d 660, 668-69 (Fed. Cir. 1988), the defendant stated that his improvement "skirted" the patent--a statement that necessarily implies knowledge of the contents of the patent. And in MEMC Electronic Materials, Inc. v. Mitsubishi Materials Silicon Corp., 420 F.3d 1369, 1378 n.4(Fed. Cir. 2005), the patentee sent the defendant notice that the defendant's product was covered by the patent. Pre-suit notice is glaringly absent here.

8 Although i4i (and the court) relied on a larger collection of documents, only these two documents both were shared with Microsoft and mentioned the patent number.

53

It is simply not logical to conclude that where, as here, a defendant admittedly did not copy the patentee's product (A44), it nonetheless "knew" that its products would infringe simply because a handful of employees received the number of a patent. A mere patent number says nothing about what the patent covers, and does nothing to inform the defendant that its product is covered by the patent.

In the absence of any evidence demonstrating--even inferentially--that Microsoft had knowledge of the contents of the '449 Patent, the district court's ruling could only have been premised on a duty to investigate. Indeed, that was what i4i's infringement expert testified to: "If I knew there was a patent out there that had similar characteristics to something that I was doing or my company was doing, I think I would direct somebody to investigate." (A1297.) But that theory likewise would fail, because this Court already has held that a defendant's knowledge that a patent "may have relevance" to its activities is insufficient to trigger a "duty to exercise due care and to investigate." Voda v. Cordis Corp., 536 F.3d 1311, 1327-28 (Fed. Cir. 2008).

Microsoft was entitled to JMOL on both theories of indirect infringement. IV. The $200,000,000 Damages Award Cannot Be Sustained The Patent Act authorizes courts to award, and sustain, only "reasonable" royalties. 35 U.S.C. § 284. Yet at $98, the per-unit royalty applied by the jury

54

represents over 60% of the sticker price of the standard edition of Word 2007, for a patent covering only a small part of Word's custom XML technology, which is itself but one of thousands of functionalities of the multi-faceted Word program.

Under Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 589 (1993), a district court must make "certain" that proposed expert testimony is both "relevant" and "reliable," and employs the "intellectual rigor" that is customary in the relevant field. Kumho Tire Co. v. Carmichael, 526 U.S. 137, 152 (1999). The district court here completely abandoned its gatekeeping role, however, allowing i4i to introduce expert testimony that makes a mockery of the concept of expert knowledge and of the Georgia-Pacific framework.

A. The Results Of The Wecker Survey Should Have Been Excluded

Because surveys are compilations of out-of-court statements that are not subject to cross-examination, courts closely scrutinize them to ensure that they display basic indicia of reliability and trustworthiness so as to warrant an exception to the hearsay rule. In addition, courts also closely police a survey's methodology for "flaws" in "the manner of conducting the survey" and "the way in which participants are questioned," which "will make any reliance on that survey unreasonable." Scott Fetzer Co. v. House of Vacuums, Inc., 381 F.3d 477, 487-88 (5th Cir. 2004). For at least six reasons, Wecker's survey does not comport with this legal standard.

55

First, the 46 respondents who actually answered the survey (out of which only 19 reported relevant custom XML usage) were inadequately screened to establish whether they had adequate knowledge of the information they reported. In this case, to identify the "correct" respondent, the survey first asked whomever answered the phone at a company for "the person who handles the set-up and administration of computers." (A4245.) It then offered that person $35 to answer questions, after which it asked a few ineffective screening questions. Nothing in these questions, which ask only about the "approximate number of computers in your company that use Microsoft Word" and mere "aware[ness]" about XML (A4245-46), suggests that respondents had personal knowledge about specific custom XML usage by their co-workers. Thus, "no [serious] effort was made to screen respondents to insure that they were `qualified'" to discuss custom XML usage. Albert v. Warner-Lambert Co., 234 F. Supp. 2d 101, 106 (D. Mass. 2002), casting serious doubt on the validity of the survey. See, e.g., Bank of Tx. v. Commerce Sw., Inc., 741 F.2d 785, 789 (5th Cir. 1984).

Second, the survey's questions, which asked about the work habits of each respondent's coworkers, introduced impermissible lay opinion, as to which, as explained, there was no attempt to establish a foundation of personal knowledge. Where courts do not allow a litigant to present direct testimony about the habits of coworkers, one cannot backdoor such evidence through the use of a "survey,"

56

particularly where, unlike with direct testimony, the survey respondent rendering an opinion on the habits of his co-workers is not even subject to cross-examination in court. See FED. R. EVID. 701 (except for expert testimony, testimony based on "opinions or inferences" is limited to those which are "rationally based on the perception of the witness"). Indeed, the Fifth Circuit has excluded even direct testimony about the habits of a witness's co-workers. United States v. Carlock, 806 F.2d 535, 552 (5th Cir. 1986).

Third, the survey's questions were impossible for respondents to answer accurately and encouraged random guessing. Such requests to speculate are fatal to a survey's reliability. Scott Fetzer, 381 F.3d at 488. For example, the survey asked respondents to give their estimate of their fellow employees' use of custom XML in Word 2003 and Word 2007, and instructed respondents that "Your best estimate is fine." (A4248-58.)

The survey also asked respondents to recall the percentage of computers per year--starting in 2008 and going back five years--that had been used to open and save an XML document containing custom XML in the relevant file format. (A4252-58.) But just as this Court's IT specialists are unlikely to know, without specialized systems analysis, how many judges or law clerks access Word's Thesaurus or Track Changes features on a daily or yearly basis, the survey respondents had no means other than rank conjecture to know offhand how often

57

the custom XML functionality might have been accessed within their respective organizations. Merely because a person helps install Word on computers does not mean he or she knows how the end users actually use Word.

Fourth, not only were the respondents not in a position to know anything reliable about the use of custom XML, but many of the 19 respondents who reported custom XML usage did not even understand the questions. Questions that are too confusing or difficult for respondents disqualify a survey as admissible evidence. See Jackson v. Nat'l Action Fin. Servs., Inc., 441 F. Supp. 2d 877, 881 (N.D. Ill. 2006). One typical example: One respondent claimed that, although only 2 computers ran Word 2007, 5% of them had opened an XML document-- even though there is no such thing as 5% of 2 computers. (A4262 (cols. Q1A, Q1B, row 40).) Still another respondent claimed that, in a typical workday in 2008, 75% of computers used Word 2003 to open an XML document, but in the entire year 2008, none did. (A4262 (col. Q2B, row 18), A4265 (col Q4C, row 18).) i4i asked the district court to turn a blind eye to these and many other nonsensical and irreconcilable answers because Wecker had "corrected" the clearly mistaken answers through "data imputation"--e.g., by converting "`3%"' to "`3 computers.'" (A4602-03.) But even if it were acceptable for a surveyor to "correct" answers (which is to say, make them up) rather than recognizing that the survey was unclear and discarding it, for many of the 19 respondents, there was no

58

conceivable way to logically reconcile the answers. If a company has only 2 computers, for example, neither "5 computers" nor "5% of computers" will solve the problem. Even more striking, Wecker stated that when a respondent had given two inconsistent answers, he always chose the larger number (A1609-10), even if there was no way of knowing whether the smaller or the larger number was actually correct.

Fifth, the survey had an incredibly low response rate--less than 5%, which is "well below the 50 percent response rate that is considered minimally acceptable." Albert, 234 F. Supp. 2d at 106 n.6. It is no secret that low response rates often skew surveys through selection bias. "If the response rate drops below 50%, the survey should be regarded with significant caution." REFERENCE MANUAL ON SCIENTIFIC EVIDENCE, FEDERAL JUDICIAL CENTER 245 (2d ed. 2000).

Sixth, a survey is useful to aggregate very large datasets through statistical sampling and other means. A report summarizing what 19 people "estimated" others might have been doing is not a survey at allit is simply a summary of inadmissible hearsay. No "state of mind exception" can save it, as that exception must be based on a witness's actual perception, not his speculation about his coworkers' habits. See generally Schering Corp. v. Pfizer Inc., 189 F.3d 218 (2d Cir. 1999) (Sotomayor, J.).

59

Given all of these errors, the district court abused its discretion in refusing to exclude the Wecker "survey."

B. Wagner's Georgia-Pacific Analysis Should Have Been Excluded

It is well-settled that "experts must use reliable methodologies rather than simplistic extrapolation and childish arithmetic with the appearance of authority by hiring a professor to mouth damages theories that make a joke of the concept of expert knowledge." Schiller & Schmidt, Inc. v. Nordisco Corp., 969 F.2d 410, 415 (7th Cir. 1992). i4i's derivation of the royalty rate of $98 per unit in this case--an astonishingly high figure that exceeds the entire price of certain editions of Word, and is based on an "analysis" that linked the royalty rate to the price of a high-end third-party XML editor that has not even been claimed to practice the patent-- vividly epitomizes such abuse of expert testimony and a district court's failure to police it. Because the district court's discretion regarding the admissibility of expert testimony is not a carte blanche justifying its complete abdication of its gatekeeping role under Daubert, its decision must be reversed.

i4i's expert purported to apply the Georgia-Pacific analysis, which seeks to estimate the price that Microsoft would have paid i4i in an arm's length bargain. That framework requires an expert to establish, among others, the value of the patent to Microsoft and its customers, identify the available alternatives to licensing the patent, and then estimate the portion of the patent's value that

60

Microsoft would be willing to pay in light of the "economic relationship between the patented method and noninfringing alternative methods." Riles v. Shell Exploration & Prod. Co., 298 F.3d 1302, 1312 (Fed. Cir. 2002). As shown below, Wagner analyzed virtually no evidence about these material factors. This alone requires reversal. Integra Lifesciences I, Ltd. v. Merck KGaA, 331 F.3d 860, 869- 72 (Fed. Cir. 2003) (vacated on other grounds) (reversing award that did "not appear to take into account numerous factors").

Instead, Wagner based his outsized royalty on a high-end third party XML editor not shown to practice the patent, and then assumed that Microsoft would be willing to pay i4i 25% of the hypothetical profits it might make if it charged $499 for Word. (See A3940-46.) This "ipse dixit" warrants reversal under Daubert. Not only is the 25% "rule-of-thumb" itself problematic for widely acknowledged reasons (see, e.g., Richard S. Toikka, Patent Licensing Under Competitive and Non-Competitive Conditions, 82 J. Pat. & Trademark Off. Soc'y 279, 292-93 (2000)), but its complete distortion in this case--through linkage to the value of a third-party's product but Microsoft's profit margin--bears no resemblance to what the rule was intended for, much less any resemblance to Georgia-Pacific.

First, Wagner erroneously used as his starting point a high-end XML editor marketed to a small subset of professionals that costs three times as much as some versions of Word, does not practice the patented process, and includes a plethora of

61

functionalities beyond the patented invention and beyond Word's additional, noninfringing capabilities, as is reflected in its far higher ($499) market price. (A1467-68; A3938, A3943-44; see also A5599-5621.)

Wagner justified his reliance on XMetaL by pointing to an e-mail from a Microsoft employee identifying XMetaL as one of Microsoft's competitors in the XML (not the custom XML) field. (A3943 & n.62; A1387-88.) He appears to have done no market research beyond that. (A1467-69.) He admitted that he did not know the average selling price of XMetaL or even whether anyone actually pays the list price for XMetaL, and he conceded that "XMetaL includes lots of functionalities that have nothing to do with opening documents containing custom XML and Word." (A1468-69.) But he nevertheless concluded that Microsoft would come to the hypothetical bargaining table valuing the '449 Patent as if it alone would enable Microsoft to market a $499 high-end, specialized product-- even if Microsoft cannot even sell the thousands of noninfringing functionalities in Word for about a third of that price. (A1389-90.) It is completely illogical to assume that Microsoft would ever agree to a benchmark based on such expensive specialized software.

To be sure, it may be proper in certain cases to benchmark a reasonable- royalty analysis to the price of a third-party product so as to identify the patent's value independent of other functionalities in the defendant's product. But to use

62

such a benchmark, the entire value of that benchmark product must be attributable to the patent. Here, however, not only does the record lack any evidence as to whether XMetaL has the capability to practice i4i's method, but XMetaL has numerous functionalities beyond opening and saving XML files, and so it cannot possibly serve as a proxy for the value of the '449 Patent to Microsoft without violating the entire market value rule.

What is more, Wagner had access to far more relevant and specific evidence of the value of custom XML to Microsoft, but chose to ignore it. He stated in his report that it "appears that custom XML support was a major factor accounting" for the $50 price difference between two versions of Office, one of which did not include custom XML. (A3962.) This $50 difference therefore shows that Microsoft and the market cannot attach more than $50 in value to custom XML, but even that overstates the value because not all uses of custom XML are alleged to infringe. 9 Even correcting for this single error and accepting all the rest of i4i's theories, the per-unit royalty would be 25% of $50, and the overall damages would have been $20 million.

9 In addition, it was not shown that the full $50 was solely attributable to the allegedly infringing functionality. For example, Office Professional contains Access, while Small Business does not. (A3938.)

63

Second, Wagner should not have applied the so-called "25% Rule," which anchored the Georgia-Pacific analysis to the factually unsupportable assumption that i4i could successfully demand a 25% cut of the profits in a hypothetical negotiation.

Word's custom XML functionality permits users not only to perform the accused process of opening and saving custom XML files but also to use Word's hundreds of word-processing tools to edit documents containing XML and/or custom XML. In addition, a large part of the functionality's value reflects the cost of writing the high-quality software necessary to perform these tasks. i4i's patent, addressed to a method, not the actual software, provides thus at best a miniscule part of the value of the custom XML functionality. In this context, where the patent represents a fraction of the inputs into the accused functionality, it blinks reality to suggest that the infringer would pay a quarter of its anticipated profits on the technology to the patent holder. In fact, even Wagner conceded that it is "problematic to use the 25-percent rule" in such a case because "it's very difficult to differentiate the contribution of the asserted patent from the contributions of all the other technologies" included in the product. (A1453.)

In addition, as XMetaL itself demonstrates, the '449 Patent was not necessary to developing the custom XML functionality. Many XML editors are available that no one alleges infringe the patent. When Microsoft came to the

64

hypothetical bargaining table, then, it would have weighed the cost of licensing the '449 Patent against its potential to develop the necessary technology in-house or to license it from another patent holder. See Riles, 298 F.3d at 1312. Microsoft routinely faces this choice. The record evidence demonstrates that for similar patents Microsoft has agreed to a lump-sum payment in the $1-5 million range. (A4793-95; see also A2188-97.) It cannot possibly be true that Microsoft would suddenly pay $200,000,000 for a technology that could have been obtained elsewhere or developed in-house for a much smaller sum.

For these reasons, the district court should have excluded Wagner's reasonable-royalty "analysis." Although the hypothetical-negotiation analysis can never be totally precise, it must bear some relation to the real world. "[N]othing in either Daubert or the Federal Rules of Evidence requires a district court to admit opinion evidence that is connected to existing data only by the ipse dixit of the expert." Gen. Elec. Co. v. Joiner, 522 U.S. 136, 146 (1997). That is because an expert's "use of a general methodology cannot vindicate a conclusion for which there is no underlying [factual] support." Black v. Food Lion, Inc., 171 F.3d 308, 314 (5th Cir. 1999). Because Daubert has tasked the court--not the jury--to perform the gatekeeping role, the district court's abdication from that role requires reversal. Indeed, one can hardly expect the jury to reach the proper decision based on cross-examination when the admission of improper damages testimony led to

65

"the usual magic tricks," Isaksen v. Vt. Castings Inc., 825 F.2d 1158, 1165 (7th Cir. 1987), as Wagner "dazzled the jury with an array of figures conveying a delusive impression of exactness," Olympia Equip. Leasing Co. v. W. Union Tel. Co., 797 F.2d 370, 382 (7th Cir. 1986).

C. Two Hundred Million Dollars Is Not A Reasonable Royalty

The enormous damages award of $200,000,000 is insupportable by the results-oriented analysis provided by i4i's experts, regardless of that testimony's admissibility, and cannot possibly qualify as a reasonable royalty. This Court should either remit the damages to a reasonable sum or remand for a new trial, because the damages verdict is clearly excessive. See King v. Ames, 179 F.3d 370, 377 n.8(5th Cir. 1999).

Section 284, which authorizes the award of a "reasonable royalty," prohibits a damages award that is "in view of all the evidence, ... so outrageously high ... as to be unsupportable as an estimation of a reasonable royalty." Lindemann Maschinenfabrik GmbH v. Am. Hoist & Derrick Co., Harris Press & Shear Div., 895 F.2d 1403, 1406 (Fed. Cir. 1990); see 35 U.S.C. § 284. Although expert testimony plays an important role, there is also "room for exercise of a common- sense estimation of what the evidence shows would be a `reasonable award.'" Lindemann, 895 F.2d 1406.

66

This Court has not hesitated in the past to reverse speculative or otherwise unsupportable damages awards in patent-infringement cases. See, e.g., Integra Lifesciences, 331 F.3d at 869-72 (reversing $15 million reasonable-royalty award that did "not appear to take into account numerous factors that would considerably reduce the value of a hypothetical license"); Oiness v. Walgreen Co., 88 F.3d 1025, 1028-34 (Fed. Cir. 1996) (reversing jury award of $11 million to patent holder for past and projected lost profits as "speculative").

The result should be no different here. No evidence could justify a royalty equal to more than half of the gross revenue per infringing user in exchange for a single infringing functionality, particularly for a software like Word that has thousands of other functionalities. It follows a fortiori that the deeply flawed expert testimony in this case, relying as it did on abject speculation and unsupported assumptions, is insufficient to justify the award. A royalty, after all, must be "supported by relevant evidence." Unisplay, S.A. v. Am. Elec. Sign Co., 69 F.3d 512, 518 (Fed. Cir. 1995). The award in this case is a product not of any good-faith attempt to assess the value of the patent but rather a poorly refereed trial in which the plaintiffs were permitted to submit completely unsubstantiated conclusions to the jury under the guise of "expert" testimony.

67

Accordingly, this Court should remit the damages to the $5,000,000 that Microsoft generally will pay to license a patent. In the alternative, this Court should remand for a new trial on damages.

D. The District Court's Award Of $40 Million In Enhanced Damages Is Unsupportable.

Based on the jury's finding of willful infringement, Microsoft's size and profitability, and the court's extraordinary conclusion that, in advocating for his client, Microsoft's trial counsel had committed "trial misconduct" (A47), the district court decided to enhance i4i's already outrageous $200,000,000 damage award by an additional $40,000,000. Microsoft was entitled to judgment as a matter of law on the issue of willfulness. But even if that were not the case, the Court's enhancement analysis is premised on errors of law and accordingly must be vacated.

1. The District Court's Willfulness Analysis Cannot Be Reconciled With Seagate

A finding of willful infringement requires a showing "by clear and convincing evidence that the infringer acted despite an objectively high likelihood that its actions constituted infringement of a valid patent," and that "this objectively-defined risk ... was either known or so obvious that it should have been known to the accused infringer." In re Seagate Techs., LLC, 497 F.3d 1360, 1371 (Fed. Cir. 2007).

68

Microsoft, as discussed above, did not have "knowledge" of the patent; therefore, the district court's conclusion that Microsoft's conduct satisfied Seagate's subjective prong is insupportable.

But the district court's "objective" analysis is even more glaringly erroneous. An objective observer in Microsoft's position would not have found a high likelihood that the '449 Patent was valid and infringed by Word's custom XML functionality. This is amply demonstrated by the facts that: (1) one objective observer, the district court, invalidated 13 claims of the patent as indefinite (A86); (2) another objective observer, the PTO, has now provisionally invalidated the Asserted Claims (A4466); (3) the jury's finding of infringement rests on a claim construction that is sharply at odds with the '449 Patent's prosecution history; (4) i4i itself took at least three years--from "late 2002" (when i4i first obtained beta versions of Word 2003) until "late 2005"to determine that it had a viable claim for infringement, and another year-and-a-half for it to actually assert infringement (A61-62); 10 and (5) i4i's only "evidence" to rebut the documentary and third-party evidence of an on-sale bar is the inventor's

10 See Finisar Corp. v. DirecTV Group, Inc., 523 F.3d 1323, 1339 (Fed. Cir. 2008) (patentee's "successful ... laches defense" that it "did not perceive a problem" for several years precluded finding of willful infringement).

69

conclusory trial testimony, which could not have been known to anyone until trial took place.

Although the district court accused Microsoft of advocating "an overly broad reading of Seagate" (A20), it is the court's reading of Seagate--not Microsoft's--that was in error.

The district court's willfulness decision is predicated on three significant errors of law, each of which is an independently sufficient basis for vacating the district court's $40,000,000 award of enhanced damages.

First, the court found that Microsoft's invalidity and noninfringement defenses were "irrelevant to the objective prong of the Seagate analysis"--indeed, "irrelevant and inappropriate"--because they "would [not] have been apparent and considered by a reasonable person in Microsoft's position prior to its infringing activity." (A21-22.) But this construction of Seagate's objective prong cannot be reconciled with this Court's post-Seagate precedents, which uniformly look to the strength of an infringer's defenses--regardless of when those defenses were developed--as part of Seagate's objective analysis. "Under [Seagate's] objective standard, both legitimate defenses to infringement claims and credible invalidity arguments demonstrate the lack of an objectively high likelihood that a party took actions constituting infringement of a valid patent." Black & Decker, Inc. v. Robert Bosch Tool Corp., 260 F. App'x 284, 291 (Fed. Cir. 2008). And this Court

70

has consistently applied that rule to reject willfulness where, such as here, legitimate defenses of claim construction or invalidity are presented. See, e.g., Cohesive Techs., Inc. v. Waters Corp., 543 F.3d 1351, 1374 (Fed. Cir. 2008) (willfullness appropriately rejected where claims of patent-in-suit are "susceptible to a reasonable construction under which [the defendant's] products did not infringe").

Moreover, the court's crabbed construction of Seagate conflicts with its own expansive inferences of knowledge of a patent. If a jury is entitled to infer a defendant's knowledge of a patent's contents, fairness dictates that the jury must also infer knowledge of available defenses to an allegation of infringement. The Court cannot permit the jury to infer knowledge of potential theories of infringement but demand that it be blinded to their defenses.

Second, the district court did not independently analyze the strength of Microsoft's defenses, instead allowing the fact that they "were rejected wholesale" by the jury and "the size of the verdict" to dictate the conclusion that the defenses were not objectively reasonable. Of course, if the jury's conclusion alone were sufficient, there would be no need for a willfulness analysis at all; the court could skip straight to enhancement. But this Court has recognized that "[t]he mere fact that the jury ultimately found [infringement] does not diminish the difficulty of their task, which must be viewed objectively." Depuy Spine, Inc. v. Medtronic

71

Sofamor Danek, Inc., 567 F.3d 1314, 1337 (Fed. Cir. 2009). Thus, in Black & Decker, this Court held that to avoid a finding of willfulness, an accused infringer's defenses do not need to be successful; they need only be "legitimate" or "credible." 260 F. App'x. at 291.

Third, the court chastised Microsoft for advocating an application of Seagate that would allow an accused infringer with knowledge of a patent number to remain ignorant of its contents. (A20.) But this is the central command of Seagate itself; there is no longer an "affirmative duty of due care." 497 F.3d at 1371.

2. The District Court's Enhancement Analysis Fails On Its Own Terms

At least three additional errors infect the district court's enhancement analysis. Any one of the errors justifies vacatur of its $40,000,000 award of enhanced damages.

First, the district court's enhancement is predicated principally on Microsoft's "lack of due care with regard to avoiding infringement" and its failure to conduct "any investigation regarding the patent or i4i's products." (A42, A44; see also A45.) But as discussed above, Seagate abrogated the duty to investigate previously placed upon those with knowledge of a patent. See Seagate, 497 F.3d at 1371 ("we abandon the affirmative duty of due care").

Second, though the district court recognized that "[e]nhanced damages are a punitive measure to penalize a willful infringer for his or her increased culpability"

72

(A42-43), the court weighed in favor of enhancement the facts that it "is undisputedly, `the worldwide leader in software ... for personal and business computing'" and that "Microsoft's 2008 revenue was $60.42 billion." (A44.) But Microsoft's success as a business "say[s] nothing about [its] culpability [as] the infringer." Jurgens v. CBK, Ltd., 80 F.3d 1566, 1570 (Fed. Cir. 1996). The defendant's financial condition is relevant, if at all, only to "the extent of enhancement," Read Corp. v. Portec, Inc., 970 F.2d 816, 826 (Fed. Cir. 1992), not to "whether to enhance damages." (A43.) Since all litigants are equal in our civil justice system, the Due Process Clause prohibits punishing a defendant based on its wealth. Honda Motor Co. v. Oberg, 512 U.S. 415 (1994).

Third, the district court erred in holding that "Microsoft's counsel's litigation conduct" "favor[ed] enhancement." This Court has been absolutely clear that "attorney ... misconduct during litigation" is "not sufficient for an increased damages award under section 284" because, like a defendant's financial condition, it "say[s] nothing about the culpability of the infringer." Read Corp., 970 F.2d at 826.

Moreover, the conduct cited by the district court was not at all inappropriate. Indeed, when the assertedly "improper arguments" were made at trial, i4i found no cause to object. Perhaps that is because Microsoft's argument that i4i and its litigation investors had brought the litigation to achieve liquidity was obviously

73

and inescapably true in light of i4i's claim for $200,000,000 in damages and the belated request for injunctive relief. Microsoft was obligated to build a record at trial showing that i4i's technology had no proven value in the marketplace and that i4i L.P. has no business other than litigation. The district court might not have liked the implications of those facts, but careful examination of the record will show that, contrary to what the out-of-context statements cited by the district court may suggest, Microsoft never argued that simply asserting a patent in these circumstances was improper. Microsoft's argument, which is undeniably appropriate, was that it was improper for i4i to bring a baseless lawsuit simply in an attempt to achieve "liquidity," particularly where the inventors had studied the accused software and not found infringement. (A625-27; A712-13; A722; A2452- 58; 2503-06; A802; A1703-09.) And to the extent that i4i also found Microsoft's allusions to bankers and bailouts unobjectionable, perhaps it is because, if such an objection were sustained, ordinary application of the goose-gander rule would likewise prohibit i4i from referring to Microsoft as "by far the biggest software company in the world" and informing the jury of the irrelevant fact that "$200 million represents this tiny fraction of Microsoft's profits." (A697, A710.) 11

11 In fact, i4i's own expert violated a motion in limine by referring to Microsoft's "monopoly," with only perfunctory admonishment by the court upon Microsoft's objection. (A1451-52, A1517-19.)

74

Affirming the district court's award of $40,000,000 will force alleged infringers to choose between rebutting the patentee's case and risking an award of enhanced damages. V. The Award Of Injunctive Relief In This Case Conflicts With eBay "[I]njunctive relief is an extraordinary remedy." Silverstein v. Penguin Putnam, Inc., 368 F.3d 77, 83-84 (2d Cir. 2004) (vacating a permanent injunction). Thus, an injunction "is `to be used sparingly, and only in a clear and plain case.'" Rizzo v. Goode, 423 U.S. 362, 378 (1976). As demonstrated below and discussed more fully in Microsoft's motion to stay, i4i has failed to make the requisite showing on each of the four eBay factors. See eBay Inc. v. MercExchange, L.L.C., 547 U.S. 388, 391 (2006).

Irreparable Harm. The district court premised its irreparable-harm finding on i4i's conclusory statements regarding generic harm that might arise from infringement by an (alleged) direct competitor, such as loss of reputation, loss of market share, and loss of customers. (A52-54.) Such generic assertions of harm are legally inadequate. Indeed, if such bare-bones allegations of competitive harm were all that the law required, every successful infringement action involving competitors would result in an automatic injunction--a result expressly rejected by the Supreme Court in eBay.

75

Rather, to satisfy its burden under eBay, i4i had to provide concrete, non- conclusory evidence of the harm that would occur absent an injunction. i4i utterly failed to meet this burden--it did not even offer any evidence about the parties' respective market shares, let alone attempt to prove the market share it has lost (or projects to lose) as a result of Microsoft's infringement. Indeed, the only evidence in the record shows that while i4i consistently lost millions of dollars before Microsoft introduced Word 2003, i4i's revenues actually improved after Word 2003 was released. (A907-12.) And although i4i identified two customers that it lost years ago, they were not lost as a result of the infringement. Rather, one customer (FDA) was lost to a third competitor (Grumman), and the other (PTO) was lost because of quality problems with i4i's product. (A1675-76, A1718-20.) And i4i did not even attempt to prove that customers prefer Word because of its ability to perform the accused method, as opposed to "a desire for other [noninfringing] features of the [accused] system." Sundance, Inc. v. DeMonte Fabricating Ltd., No. 02-73543, 2007 WL 37742, at *2 (E.D. Mich. Jan. 4, 2007). Because i4i failed to connect its purported harm to Microsoft's infringing conduct, the district court had no basis to conclude that i4i will be harmed absent an injunction against that conduct. Id.

Nor did i4i present a shred of evidence on its alleged loss of brand recognition; rather, it simply asked the court to presume such a loss from the mere

76

fact that Microsoft has entered the XML-editor market. (A4626.) But such "conjectural" assertions fail as a matter of law. See Roper Corp. v. Litton Sys., Inc., 757 F.2d 1266, 1273 (Fed. Cir. 1985).

In any event, whatever losses i4i might have proved, they all occurred in the past. Indeed, i4i adduced no evidence that it will lose future sales or revenues as a result of infringement, and it is undisputed that i4i's current product is an add-on or plug-in to Word and requires Word to operate (A885-88). Thus, i4i is not currently in competition with Microsoft, but in a complementary vertical relation to it. (See also A1027-28 (K. Thomas, i4i's "Product Strategist," admitting no competition); A1479-80; see also A6037-40.) The district court acknowledged the absence of future harm in light of the fact that i4i's product and Word are currently complementary, rather than competing, but deemed it irrelevant, focusing instead on harm that i4i allegedly suffered several years ago. (A53-54 (holding that the fact that i4i's product does not currently compete with Word because it is merely an add-on to Word "does not negate the injury incurred" previously or show that i4i "has not suffered an irreparable injury").) It is axiomatic, however, that injunctive relief is not available to remedy an injury that occurred in the past. See United States v. Or. State Med. Soc., 343 U.S. 326, 333 (1952) ("The sole function of an action for injunction is to forestall future violations. It is [] unrelated to

77

punishment or reparations for those past."); Dombrowski v. Pfister, 380 U.S. 479, 485 (1965) ("injunctive relief looks to the future").

Inadequacy of Monetary Relief. It is well-established that, in all but the rarest of circumstances, economic injuries are remediable with monetary or legal damages, including an ongoing royalty. See Sampson v. Murray, 415 U.S. 61, 90 (1974) ("Mere injuries, however substantial, in terms of money ... are not enough" to warrant injunctive relief); Paice LLC v. Toyota Motor Corp., 504 F.3d 1293 (Fed. Cir. 2007) (ongoing royalty). But instead of requiring i4i to show why money damages are incalculable, the district court inexplicably faulted Microsoft for not presenting evidence on "alternative methods" for compensating i4i for "loss of customers, market share, and brand recognition." (A55.) It was (and remains) i4i's burden to prove, however, the insufficiency of legal relief; its failure to carry that burden alone warrants vacatur of the injunction. See, e.g., Praxair, Inc. v. ATMI, Inc., 479 F. Supp. 2d 440, 444 (D. Del. 2007) (patentee must explain "why it may have `difficulties calculating damages going forward'" and why money damages cannot "compensate for `lost market share'").

Balance of Equities. Given that i4i does not even have a stand-alone product that could compete with Word on the market, while Microsoft would lose its significant investments in creating the custom XML functionality of Word and would have to invest significant resources in redesigning it, the balance of the

78

equities plainly favors Microsoft. Moreover, where, as here, a patent allegedly covers only one of a multitude of functionalities of an accused product, an injunction may be inequitable. See eBay, 547 U.S. at 396-97 (Kennedy, J., concurring).

Public Interest. In conflict with well-established precedent, the district court focused its public-interest inquiry exclusively on public health and welfare considerations. But far from being restricted to those rarely implicated concerns, the public interest is a "supple principle," R.R. Comm'n v. Pullman Co., 312 U.S. 496, 500 (1941), for which courts of equity should "go much farther both to give and withhold relief than they are accustomed to go when only private interests are involved." Virginian Ry. Co. v. Sys. Fed'n, 300 U.S. 515, 552 (1937).

The district court overlooked the significant public interest that disfavors injunctions based on patents whose validity is in significant doubt, such as those as to which the PTO has granted reexamination. Cardinal Chem. Co. v. Morton Int'l, 508 U.S. 83, 100 (1993) (recognizing the strong public interest in "resolving questions of patent validity"). Indeed, while the public interest favors enforcement of valid patents, the public is just as harmed by the enforcement of invalid patents. Pope Mfg. Co. v. Gormully, 144 U.S. 224, 234 (1892); Lear, Inc. v. Adkins, 395 U.S. 653, 670 (1969). Injunctive relief is especially inappropriate--and the public- interest effects are especially pernicious--where, as here, the patent challenger was

79

required to prove invalidity by clear and convincing evidence despite relying on references not before the PTO. The public interest thus favors a final--and better- tested--determination of validity before Microsoft and its distributors are forced to incur additional costs in redesigning Word (and removing the current copies of Word from the market). The district court's unduly narrow conception of the public interest--including its failure to consider the effect of the reexamination on the public interest--is an abuse of discretion, and thus requires vacatur. Cf. Piper Aircraft Co. v. Reyno, 454 U.S. 235, 257 (1981) (a district court abuses its discretion when it fails to consider "all relevant public ... interest factors").

CONCLUSION

The injunction should be vacated. The judgment should be reversed and the case remanded with instructions to enter judgment of noninfringement and invalidity. In the alternative, the judgment should be vacated and the case remanded for a new trial on all issues as to which Microsoft is not entitled to judgment.

80

ADDENDA

  View Printable Version


Groklaw © Copyright 2003-2013 Pamela Jones.
All trademarks and copyrights on this page are owned by their respective owners.
Comments are owned by the individual posters.

PJ's articles are licensed under a Creative Commons License. ( Details )