• Cockburn's reasonable royalty calculation rests entirely on improper categories not recoverable as patent damages because Cockburn has
• improperly included all of Google's worldwide advertising revenue from Android phones in his royalty calculation
• improperly smuggled Oracle's lost profits in his royalty calculation, with no attempt to satisfy the stringent Panduit test
• improperly included damages for so- called "fragmentation" in his royalty base
• Cockburn also calculates an unreasonable, inflated royalty rate by refusing to value the patented technology at issue; among other errors Cockburn's report makes no effort to base his royalty rate on the value of the patented technology
• Cockburn's methodology contains numerous other legal errors, including:
• inflating his damages estimate by double counting, including Oracle's alleged future damages despite Oracle's request for an injunction
• improperly including Google's revenues from international sales in his damages calculation
• improperly setting the date of the hypothetical negotiation in 2008, when Android phones first launched, rather than when the Android code was first written
• Cockburn ignores record facts and undisputed market realities by
• ignoring the actual market value of Java, as proven by numerous licenses granted by Sun, Google's negotiations with Sun, and the value assigned to Java by Oracle
• assuming, without any factual basis, that Sun and Google would have negotiated an expensive, Java-incompatible license rather than a cheap, Java-compatible license, as evidenced by:
• Far from prohibiting fragmentation of Java, Sun promoted such fragmentation through its licensing program
• Even if Cockburn's assumptions about fragmentation are correct, it is improper to base a damages award on compatibility with nonpatented features
• Cockburn relying on fragmentation as a reason to ignore relevant evidence, including Java-compatible licenses
• offering no meaningful analysis regarding copyright damages
• Cockburn offers no meaningful analysis regarding copyright damages

In response, Oracle argued:

• Professor Cockburn’s calculations do not rely on any unrecoverable categories of damages
• Professor Cockburn correctly considers Google’s anticipated Android-related advertising revenues
• Professor Cockburn’s consideration of Oracle’s anticipated losses from the infringement is appropriate
• Professor Cockburn does not attach a dollar value to fragmentation—but his consideration of it is appropriate
• Professor Cockburn’s calculations properly account for the value of the intellectual property at issue
• Professor Cockburn’s calculations do not reflect any “legal errors”
• Professor Cockburn does not double count future damages
• Professor Cockburn does not improperly include worldwide revenues
• Professor Cockburn correctly assesses the hypothetical negotiation as of October 2008
• Professor Cockburn Considered Copyright Damages
• The Cockburn Report reflects careful consideration of the relevant evidence and market realities
• Professor Cockburn carefully considered evidence regarding the substantial value of Java
• Professor Cockburn correctly considered other licenses
• Professor Cockburn correctly considered a license for an incompatible version of Java

So now we have Google's reply to Oracle's responses, the final piece of the briefing before oral arguments which are scheduled to take place on July 21 at 2:00 p.m. What comes of all of this are what appear to be the core arguments:

• If Google has infringed, are damages to be determined solely in relation to the infringed patent(s) or to all of Java?
• If Google has infringed, is it appropriate to look to Google's advertising revenue as the base against which damages are to be calculated?
• Central to the previous question is: What is the infringing act, the production, distribution, and use of the Android software or the running of ads on Android-powered devices?

In addition to the arguments and briefing related to Prof. Cockburn's right to testify, the court is also seeking further guidance on the issue of damages. To that end, under Federal Rule of Evidence 706 - Court Appointed Experts, the court has issued an order [PDF] calling for the parties to name two qualified experts to be available to testify as to damages. This order clearly stems from the contention that has arisen with respect to Prof. Cockburn and his report. The parties have until July 26 to submit the proposed candidates.

Rule 706 provides:

The court may on its own motion or on the motion of any party enter an order to show cause why expert witnesses should not be appointed, and may request the parties to submit nominations. The court may appoint any expert witnesses agreed upon by the parties, and may appoint expert witnesses of its own selection. An expert witness shall not be appointed by the court unless the witness consents to act. A witness so appointed shall be informed of the witness' duties by the court in writing, a copy of which shall be filed with the clerk, or at a conference in which the parties shall have opportunity to participate. A witness so appointed shall advise the parties of the witness' findings, if any; the witness' deposition may be taken by any party; and the witness may be called to testify by the court or any party. The witness shall be subject to cross-examination by each party, including a party calling the witness.

Order Regarding Rule 706 Experts:

IN THE UNITED STATES DISTRICT COURT
FOR THE NORTHERN DISTRICT OF CALIFORNIA

ORACLE AMERICA, INC.,
Plaintiff,
v.
GOOGLE INC.,
Defendant.

No. C 10-03561 WHA

ORDER REGARDING RULE 706 EXPERTS

In order to be in a position to appoint a damages expert under FRE 706, counsel shall agree on two qualified candidates in this district and jointly call them to make sure they would be available and have no conflicts. After vetting the two candidates, counsel shall submit the two names and resumes to the Court by JULY 26, 2011. The Court will then tentatively pick one of the two.

Due to the large damages at issue, the Court tentatively plans to use the selected expert to testify before the jury at trial under FRE 706 and not as a confidential advisor to the judge. Similarly, both sides shall recommend two qualified economists or other experts to testify under FRE 706 concerning whether or not the claims tried to the jury, if infringed, constitute the basis for demand for Android. Each side may, by said date, also submit declarations showing cause why such appointments would (or would not) be advisable, and describing the assignments counsel would prefer be given to the experts, but the names must be given anyway.

IT IS SO ORDERED.

Dated: July 5, 2011.

WILLIAM ALSUP
UNITED STATES DISTRICT JUDGE

Google's Reply Brief:

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

KEKER & VAN NEST LLP
ROBERT A. VAN NEST - #84065
[email]
CHRISTA M. ANDERSON - #184325
[email]
DANIEL PURCELL - #191424
[email]
[address, phone fax]

KING & SPALDING LLP
DONALD F. ZIMMER, JR. - #224323
[email]
CHERYL A. SABNIS - #224323
[email]
[address, phone, fax]

KING & SPALDING LLP
SCOTT T. WEINGAERTNER (Pro Hac Vice)
[email]
ROBERT F. PERRY
[email]
BRUCE W. BABER (Pro Hac VIce)
[email]
[address, phone, fax]

GREENBERG TRAURIG, LLP
IAN C. BALLON - #141819
[email]
HEATHER MEEKER - #172148
[email]
[address, phone, fax]

Attorneys for Defendant
GOOGLE INC.

UNITED STATES DISTRICT COURT
NORTHERN DISTRICT OF CALIFORNIA
SAN FRANCISCO DIVISION

__________________

Case No. 3:10-cv-03561 WHA

GOOGLE INC.'S REPLY BRIEF IN
SUPPORT OF DAUBERT MOTION

Date: July 21, 2011
Time: 2:00 p.m.
Courtroom: 9, 19th Floor
Judge: The Honorable William Alsup

Trial Date: October 31, 2011

_______________________

TABLE OF CONTENTS

II. ARGUMENT................................2

A. Cockburn has failed to tie his opinion to the particular facts of this
case.........................................4

1. Cockburn never values the patents and copyrights at issue,
instead equating the value of those patents and copyrights with
the value of all of "Java."......................................4

2. Cockburn ignores relevant, comparable licenses and instead
relies on inapposite licenses for unrelated technology......................7

B. Cockburn ignores the clearly established legal framework for
the calculating patent damages, including numerous legally unrecoverable
damages categories in his estimate and using an arbitrary
methodology.........................9

1. Oracle cannot recover Google's advertising revenues as a
matter of law.................................9

2. Oracle cannot recover lost profits as part of a royalty as a
matter of law..............................11

3. Oracle cannot recover "fragmentation" damages as a matter of
law.....................................12

4. Oracle boosts its damages estimate through double counting.........13

5. Oracle cannot recover Google's international revenues as a
matter of law.....................................13

6. Oracle uses the wrong date for its hypothetical royalty
negotiation..............................14

7. Cockburn's use of the Nash Bargaining Solution is arbitrary
and improper under Federal Circuit law...............................15

i

TABLE OF AUTHORITIES

Cases

Concord Boat v. Brunswick Corp., 207 F.3d 1039 (8th Cir. 2000).......3

DataQuill Ltd. v. Handspring, Inc., No. 01 C 4635,
2004 WL 1102309 (N.D. Ill. May 10, 2004)..........................14

Daubert v. Merrill Dow Pharms., Inc., 509 U.S. 579 (1993)..........3, 4

Deepsouth Packing Co. v. Laitram Corp., 406 U.S. 518 (1972)..........13

Fonar Corp. v. General Elec. Co., 107 F.3d 1543 (Fed. Cir. 1997).......10

Garcia v. Coleman, No. C-07-2279 EMC, 2009 WL 799393 (N.D. Cal. Mar. 24, 2009)...............11

Garretson v. Clark, 111 U.S. 120 (1884)..................4

Imonex Servs., Inc. v. W.H. Munzprufer Dietmar Trenner GMbH,
408 F.3d 1374 (Fed. Cir. 2005).........................4

Interactive Pictures Corp. v. Infinite Pictures, Inc., 21A F.3d 1371 (Fed. Cir. 2001)........................14

Litecubes, LLC v. Northern Light Prods., Inc., 523 F.3d 1353 (Fed. Cir. 2008)...............14

Lucent Techs., Inc. v. Gateway, Inc., 580 F.3d 1301 (Fed. Cir. 2009).......3, 4, 6, 8, 9

Metro-Goldwyn-Mayer Studios, Inc. v. Grokster, 545 U.S. 913 (2005).......11

Panduit Corp. v. Stahlin Bros. Fibre Works, Inc., 575 F.2d 1152 (6th Cir. 1978)....................4

Polar Bear Prods., Inc. v. Timex Corp., 384 F.3d 700 (9th Cir. 2004)..........11

RestQNet.com, Inc. v. Lansa, Inc., 594 F.3d 860 (Fed. Cir. 2010).......5

Riles v. Shell Exploration & Prod. Co., 298 F.3d 1302 (Fed. Cir. 2002).......5

Trans-World Mfg. Corp. v. AlNyman & Sons, Inc., 750 F.2d 1552 (Fed. Cir. 1984).................11

Uniloc USA, Inc. v. Microsoft Corp., 632 F.3d 1292 (Fed. Cir. 2011).....3, 7, 10, 15

Uniloc USA, Inc. v. Microsoft Corp., 632 F. Supp. 2d 147 (D.R.I. 2009).....14

Statutes

ii

I. INTRODUCTION

Cockburn's most fundamental error, which Oracle's brief repeats, is conflating the seven narrow patents and the copyrights at issue in this case with "Java"—a nebulous concept that includes, among other things, a trademarked marketing brand, an open-source programming language, and an operating platform for desktop computers, servers, and mobile devices. But the Oracle intellectual property at issue here is not "Java." The asserted claims of the patents at issue relate, at most, only to a small subset of the technology comprising the version of the Java Virtual Machine ("JVM"), as the JVM is used on the Java MicroEdition ("JavaME") platform, which is the fork of the overall Java operating platform designed to support mobile devices, such as smartphones. Yet both Cockburn and Oracle consistently refer to "the Java patents" and "the Java intellectual property," treating the technology at issue here as if it embodied, and had the same value, as all of "Java." Cockburn doesn't offer a shred of evidence that the technology at issue is important, much less essential, to "Java"—and, indeed, Oracle has admitted in discovery responses that it has no evidence that any of its intellectual property is "essential" to "Java." The relevant question here is the incremental value of the patents and copyrights at issue to Java and to Android, but Cockburn doesn't even ask that question, let alone answer it. His methodology is grounded only in a desire to maximize Oracle's damages.

• Cockburn uses the "Nash Bargaining Solution" to award Oracle half of Google's alleged incremental gain from Android as a whole, not from the allegedly infringing technology. This critical error is the inevitable result of Cockburn's refusal to value the technology at issue. Even if the Nash approach were the appropriate one legally—and Oracle doesn't cite any case where any court has approved the use of that methodology—Cockburn has misapplied Nash by equating the limited technology at issue here with the entire Android platform.

1

• Cockburn's methodological error allows him to put damages as high as $6.1 billion, nearly as much as the $7.4 billion that Oracle paid for all of Sun. Oracle claims the actual number is $2.6 billion, Opp'n at 2, which is equally unsupported. Cockburn's report describes his Exhibit 26 as setting forth "the full range of scenarios and the resulting payments from Google to Sun." Cockburn Report ("Report") ¶ 228. Exhibit 26, in turn, gives a damages range of $1.4 billion to $6.1 billion. Id. Ex. 26. By itself, the vastness of that range proves Cockburn unreliable. In fact, Cockburn's chart, and Oracle's brief, make clear that Cockburn's real estimate might be even higher, because Oracle also intends to lay claim to Google's revenue from non-mobile, non-Android advertising. Opp'n at 5 & n1; Report ¶ 218 & Ex. 26.
• Cockburn improperly includes Google's Android advertising revenue in his royalty base even though Oracle does not (and cannot) allege that Google's ads infringe. This legal error infects Cockburn's entire analysis. Contrary to Oracle's argument, Opp'n at 8-10, the Georgia-Pacific analysis allows sales of non-infringing products as one factor to be considered in setting an applicable royalty rate. Such products cannot be included in the royalty base, as Cockburn does, unless the patentee satisfies the entire market value rule, which Oracle does not even attempt to do.
• Cockburn improperly imports Oracle's lost profits into his royalty base. This is yet another legal error in Cockburn's methodology. Oracle argues that lost profits can be considered in a royalty analysis. Opp'n at 11. Again, this is true only with respect to the royalty rate under Georgia-Pacific, not the royalty base, as Cockburn does.
• Cockburn improperly uses alleged "fragmentation"damages to increase Oracle's damages. Oracle claims that "the dollar amount [Cockburn] currently assigns to fragmentation is zero." Opp'n at 2 (emphasis in original). This is false. Cockburn actually uses so-called fragmentation as the basis of his conclusion that Google would have paid Oracle a $900 million to $ 1.4 billion lump-sum payment.

Cockburn's methodology contains numerous other flaws. Although Federal Circuit cases require experts to tie their analysis closely to the facts of the case at hand, Cockburn ignores myriad (relatively modest) market valuations of Java and its components, relying instead on inapposite licenses with larger attached dollar amounts. He double-counts by importing damages for future conduct into his giant lump-sum calculation, despite Oracle's request to cut off future infringement via injunction. He gives Oracle royalties on purely foreign activities. He sets the date of his hypothetical negotiation at least a full year too late, then uses that late date to justify a larger award. His methodology is unreliable and excludable under Daubert and its progeny.

II. ARGUMENT

Oracle's opposition boils down to the mistaken argument that Google has no objection to Cockburn's "methodology," but has raised only "factual disputes" that go only to the weight of his testimony, not its admissibility. Oracle argues that, as long as Cockburn relies on evidence that is "sufficiently related" to the case, it meets the modest "minimum threshold" to support the

2

resulting opinion. Opp'n at 7:7-9. Oracle is wrong for two independent reasons. First, Google's fundamental point is that, due to Cockburn's foundational methodological mistakes, the facts underlying his opinion aren't related to this case in any meaningful way. Google is not quibbling over the accuracy of some aspects of Cockburn's data, or arguing that Cockburn ought to have used some other, marginally more relevant information. Cockburn's problem is that he has relied on the wrong categories of data entirely -- data regarding the value of Java as a whole, or other telecommunications technologies, rather than data tied to the intellectual property at issue. Second, Google has challenged Cockburn's entire methodology, because -- in addition to his central error of valuing "Java" rather than the specific patent claims and copyrights at issue -- his royalty analysis ignores the law, seeking recovery of several categories of damages that are off limits entirely. The mistakes Cockburn has made here are the same errors that have inspired the Federal Circuit to reverse jury verdicts in numerous recent cases.

With respect to the factual basis of Cockburn's opinion, to be admissible, expert opinions must "sufficiently [tie the expert testimony on damages] to the facts of the case." Uniloc USA, Inc. v. Microsoft Corp., 632 F.3d 1292, 1315 (Fed. Cir. 2011) (citing Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 579, 591 (1993)) (alteration in original). "The bottom line ... is that one major determinant of whether an expert should be excluded is whether he has justified the application of a general theory to the facts of the case." Id. at 1316. Here, Cockburn fails to tie his damages methodology to the specific patent claims and copyrighted material at issue.

Even worse, Cockburn's model is "not grounded in the economic reality of the [relevant] market, for it ignored inconvenient evidence." Concord Boat v. Brunswick Corp., 207 F.3d 1039, 1056 (8th Cir. 2000). In Lucent Techs., Inc. v. Gateway, Inc., 580 F.3d 1301 (Fed. Cir. 2009), the Federal Circuit evaluated in detail and rejected an expert's hypothetical negotiation analysis where the expert used licenses (1) for technology different from what the patents in suit covered; and (2) with structures different from the hypothetical license. See id. at 1325-32. That is precisely what Cockburn did here, ignoring Sun's decade-long history of freely licensing Java to all comers for relatively little money, and Sun's own negotiations with Google over a Java license, in favor of litigation settlements and licenses for wireless technologies not at issue here.

3

As for Coburn's disregard of Federal Circuit limitations regarding calculation of patent damages, Daubert makes clear that, as a baseline, "the trial judge must ensure that any and all scientific testimony or evidence admitted is not only relevant, but reliable." Daubert, 509 U.S. at 589. If the law provides that damages must be calculated a certain way, or puts a given category of purported damages off limits, expert testimony that ignores those limitations is necessarily irrelevant to the jury's damage analysis. See Imonex Servs., Inc. v. W.H.Munzprufer Dietmar Trenner GMbH, 408 F.3d 1374, 1380 (Fed. Cir. 2005) (barring expert from providing "testimony on entire market value rule" that "bore no relation to that rule").

A. Cockburn has failed to tie his opinion to the particular facts of this case.

1. Cockburn never values the patents and copyrights at issue, instead equating
the value of those patents and copyrights with the value of all of "Java."

Oracle has accused the Android mobile software platform of infringing various specific claims of seven asserted patents and certain copyrighted material. The issue at trial will be whether that specified material is present in the Android software. Likewise, any calculation of royalties owed to Oracle must be tied to the technology that was allegedly infringed. Just two years ago, in Lucent, the Federal Circuit held that the entire purpose of the Georgia-Pacific royalty analysis is "to elucidate how the parties would have valued the patented feature during the hypothetical negotiation." Lucent, 580 F.3d at 1332 (emphasis added). That principle is nothing new. See, e.g., Panduit Corp. v. Stahlin Bros. Fibre Works, Inc., 575 F.2d 1152, 1159 (6th Cir. 1978) ("the relevant facts" in a royalty analysis include (1) "what plaintiffs property was"; (2) "to what extent defendant has taken it"; and (3) "its usefulness and commercial value as shown by its advantages over other things and by the extent of its use"); Garretson v. Clark, 111 U.S. 120, 121 (1884) (requiring apportionment between patented and unpatented features).

Here, rather than doing the hard work of calculating the value of the intellectual property at issue, Oracle and Cockburn have taken a self-serving short cut and equated the patents and copyrights at issue with all of "Java"—a floating term Oracle uses interchangeably to refer to an open-source programming language, a series of distinct operating platforms for various types of computing devices, and an organizational philosophy. This is an extreme overreach.

4

The best way to conceptualize the extent of Oracle's overreach is to visualize "Java" as a series of concentric circles. The outermost circle is the entire Java platform, including all its various forks-e.g., JavaSE, JaveEE, JavaME, and JavaCard. The second circle, contained within the first one, is the fork of Java at issue here-JavaME, which runs on mobile devices like smartphones. The third circle, contained within JavaME, is the particular application of the Java Virtual Machine that runs on JavaME. But the relevant level for purposes of this case (and Cockburn's analysis) is a fourth circle, containing only those aspects of the JVM covered by the patents and copyrights at issue. This intellectual property is vastly smaller in scope than "Java."

Because Cockburn persists in charging Google for the value of Java-related technology "unrelated to the claimed invention," his opinion "does not support compensation for infringement but pushes beyond the reach of the statute." RestQNet.com, Inc. v. Lansa, Inc., 594 F.3d 860, 869 (Fed. Cir. 2010). The Federal Circuit expressly rejected exactly this sort of overreaching damages claim in Riles v. Shell Exploration & Prod. Co., 298 F.3d 1302 (Fed. Cir. 2002). There, the patent covered part of the process of constructing offshore oil-drilling platforms, but the patentee demanded damages based on Shell's oil-drilling revenues. In other words, just as Oracle wants to do here, the Riles plaintiff tried to bootstrap infringement of one minor aspect of a device into a claim on downstream revenue allegedly produced by operation of the device. The Federal Circuit rejected that approach, ruling that "the entire revenue of the Spirit platform bears no relation to the value of the patented method." Id. at 1311-12.

Although Oracle contends that Cockburn "did, in fact, carefully consider the value of the value of the technology at issue," nowhere in its opposition does it point to any part of Cockburn's report containing any valuation of (1) the specific Android performance and security improvements enabled by the asserted patent claims; or (2) the benefits to Android from using the asserted copyrighted material. Instead, Oracle changes the subject. First, it argues that Sun never licensed stand-alone Java-related patents, only Java itself. Of course, Cockburn disregards Sun's habit of licensing all of Java for amounts orders of magnitude less than what Oracle now demands. In any event, a reasonable patent royalty analysis presupposes a hypothetical negotiation between a willing licensor and a willing licensee regarding the technology at issue.

5

"See Lucent, 580 F.3d at 1324-25. Cockburn brushes aside this fundamental premise by assuming the contrary-that Oracle would have been unwilling to license the asserted technology.

Second, Oracle repeats Cockburn's conclusory statement that the patents and copyrights at issue are "essential" to Java. It points to no evidence supporting this claim. Oracle tries to excuse Cockburn's failure to support his claim that the technology at issue is "essential" to Java by pointing out that Oracle s technical experts have yet to submit their reports, but this is no answer. Cockburn had an unqualified obligation under Daubert and Uniloc to tether his opinion to the facts of the case, even if that meant obtaining the relevant evidence from a technical expert. Technical expert reports are due in less than a month, and Oracle's opposition makes clear that Cockburn prepared his report after "discussions with Oracle's technical expert." Opp'n at 5. Cockburn cannot get away with stating the unsupported conclusion that the patents and copyrights at issue are "essential"—and thus equally valuable as—"Java" as a whole.

In fact, Oracle's own discovery responses expose Cockburn's conclusions as baseless. In its Request for Production 22, Google asked Oracle for documents "sufficient to identify any intellectual property rights that are essential to practice each release of each Java specification." Reply Declaration of Daniel Purcell ("Purcell Reply Decl.") Ex. A at 1. In other words, Google asked Oracle to back up Cockburn's assumption. Oracle responded clearly in the negative:

Oracle is unaware of any responsive documents that can be found in its possession, custody, or control with reasonable efforts. If Oracle becomes aware of any non-privileged, responsive documents in its possession, custody, or control through reasonable efforts, it will produce them.

Id. (emphasis added). Oracle has never amended this response or produced documents showing that the patented or copyrighted material at issue is essential to Java.

Third, Oracle similarly fails to establish that the patented or copyrighted material at issue here is essential to Android. Oracle cites Android documents stating that Sun's JVM architecture was important to the Android software, but this is apples and oranges. This case is not about whether Android uses a virtual-machine architecture as a general matter. Oracle has admitted that virtual machines existed in the prior art long before Java. Oracle is claiming that Android uses certain patents and copyrights allegedly practiced by Sun's JVM. Accordingly, it

6

has to isolate and prove the value of those patents and copyrights. Indeed, Oracle has not proved that Sun's JVM even practices the patents in suit. Recent deposition testimony establishes that Oracle fails to practice at least one of those patents, Purcell Reply Decl. Ex. B (Griesemer Depo.) at 179:20-:25, 184:4-:19, and Oracle had admitted in its discovery responses that it does not mark its products with the patents, casting doubt on whether it practices any of them. Id. Ex. C at 3.

Fourth, Oracle recasts Cockburn's unsupported conclusion as creating a "factual dispute" that the Court is powerless to resolve. Opp'n at 15. But Cockburn has proffered no evidence or analysis regarding how the patented components of the JVM fit into the larger whole, what benefits (if any) those components provide to the functionality of the Android software, or how the market values those components. Cockburn has not done enough to create an issue of fact.

Fifth, Oracle quarrels with Google's claim that Cockburn calculated damages through the end of 2025, even though six of the seven patents-in-suit expire in 2018 or earlier. Oracle says Cockburn calculates royalties only through 2021. Opp'n at 15-16. That would be equally arbitrary, but Oracle is wrong. Cockburn's report offers various alternative calculations, including one that runs through 2025, Report Ex. 24, and presumably Cockburn is reserving the right to testify accordingly at trial. But again, the broader point is that Cockburn has not separated out any aspect of the patents or copyrights at issue. One of the asserted patents, the '104, expires next year. Similarly, Oracle does not dispute Google's contention that the '720 patent has no commercial value. Declaration of Scott Wemgaertner ("Weingaertner Decl.") ¶ 13 & Ex. M. Yet under Cockburn's calculation, Google would owe the same multi-billion-dollar amount whether it infringed just one claim of the '104 or '720 patents, or every asserted claim of all seven asserted patents. For all these reasons, Cockburn has failed to "justif[y] the application of a general theory to the facts of the case." Uniloc, 632 F.3d at 1316.

2. Cockburn ignores relevant, comparable licenses and instead relies on
inapposite licenses for unrelated technology.

In addition to valuing the wrong thing—"Java" as a whole rather than the much narrower technology at issue in this case—Cockburn makes the additional, fundamental error of ignoring relevant market valuations of the Java platform in favor of irrelevant valuations of distinct

7

technologies. In Lucent, the Federal Circuit explained that "the licenses relied on by the patentee in proving damages" must be "sufficiently comparable to the hypothetical license at issue in suit." Lucent, 580 F.3d at 1325. There, the appellate court reversed a $348 million lump-sum damages award, because the licenses Lucent's expert relied on were not comparable, leading the jury to award a damages amount that was "not supported by substantial evidence, but is based mainly on speculation or guesswork." Id. at 1355. Cockburn's analysis here is just as shoddy.

The overriding problem with Cockburn's opinion is that he puts the range of damages in an extraordinarily broad range between $1.4 billion and $6.1 billion, Report Ex. 26, whereas Oracle paid $7.4 billion for all of Sun. Sun was much bigger than "Java," and, as explained already, the patents and copyrights at issue here are a small part of "Java." Cockburn's estimate would allow Oracle to substantially finance its entire acquisition of Sun.

[Redacted]

Similarly, Sun and Oracle repeatedly valued the entire Java platform—including in their regulatory filings—and have never suggested that platform is worth billions of dollars. In January 2010, as part of their disclosures when acquiring Sun, Oracle valued all of Sun's "software-related Core Technology" at about $68.8 million. Because Sun's Core Technology included Java and other assets, Java's actual value was somewhat less than $68.8 million. Declaration of Gregory Leonard ¶ 13. [Redacted]

8

As Oracle's opposition brief admits, Opp'n at 20-21, Cockburn ignored all this evidence. He relied on irrelevancies instead. [Redacted]

Third, Oracle defends Cockburn's use of "other mobile related IP agreements, such as the one between Nokia and Qualcomm," to conclude that Google would have agreed to a billion-dollar lump-sum payment along with a royalty share. Opp'n at 21. This is exactly what Lucent condemned—reliance on "license agreements [that] are radically different from the hypothetical agreement under consideration," 580 F.3d at 1327—in this case, licenses between unrelated third parties who are in completely different lines of business and practice distinct technologies.

Finally, Oracle doesn't address the fundamental problem with Cockburn's calculation of the value of a license to an incompatible variant of Java rather than a compatible one. Opp'n at 22-23. Cockburn never considers the possibility, not even as a potential next-best alternative, that in a willing negotiation, see Lucent, 580 F.3d at 1324-25, Google and Sun would have negotiated a compatible Java license, and Google would have developed the Android software accordingly. Oracle's citation to Panduit, Opp'n at 23, is likewise unhelpful. Even if Oracle could have charged more to Google than some other past licensee, there is no support whatsoever in the record for Cockburn placing damages in the many billions of dollars.

B. Cockburn ignores the clearly established legal framework for calculating patent
damages, including numerous legally unrecoverable damages categories in his estimate and using an arbitrary methodology.

1. Oracle cannot recover Google's advertising revenues as a matter of law.

The lion's share of Oracle's purported damages here comes from Google's revenue from advertising on Android-based phones throughout the world. Report Ex. 26. But Google's

9

advertising is not the product accused of infringing Oracle's patents or copyrights. The accused product is Android software. When confronted with that basic disconnect, Oracle blusters that "Google's contention fails as a matter of law and common sense." Opp'n at 8. But Oracle does not actually muster any authority contradicting the settled legal principle that a patent holder may include non-infringing product revenues in the base of a reasonable royalty only if it is able to satisfy the entire market rule--something Oracle does not even try to do.

Under the entire market rule, the base of a reasonable royalty can include revenues from non-infringing components or products "only where the patented feature creates the 'basis for customer demand' or 'substantially create[s] the value of the component parts.'" Uniloc, 623 F.3d at 1318; see also Fonar Corp. v. General Elec. Co., 107 F.3d 1543, 1552-53 (Fed. Cir. 1997) (same). Oracle does not dispute this basic rule or its application in this case. Opp'n at 10. But Oracle points to nothing in Cockburn's report or the factual record suggesting that the portions of the JavaME protected by the asserted patents or copyright causes consumers with Android phones to demand Google's advertising or that Google's advertising somehow creates the value of Android software. As a simple matter of logic, there is no such connection. The technology at issue enables a small part of the functionality of Android-based phones, regardless of whether the user is viewing ads hosted by Google or anyone else. Similarly, Google's ads are viewable on any operating system on any device, mobile or not, and are not uniquely enabled by the Android software. Oracle ignores this point in its opposition brief.

Instead, Oracle relies on several non sequiturs. First, it claims that Google has made a lot of money from Android-related advertising. Of course—this is why Oracle wants to claim that revenue as damages. But the fact that Google has made money on a separate product does not satisfy the entire market value rule. Second, Oracle argues that Cockburn "was entitled to consider Google's ancillary revenues under the Georgia-Pacific framework." Opp'n at 9. Georgia-Pacific factor 6 allows consideration of ancillary revenues for purposes of raising or lowering the applicable royalty rate, but that is not what Oracle is doing. Instead, it is trying to import Google's advertising revenue wholesale into the royalty base without satisfying the entire market value rule. Nothing in Georgia-Pacific or any other case allows that. Third, Oracle cites

10

four inapposite cases. Three of the four--Metro-Goldwyn-Mayer Studios, Inc. v. Grokster, 545 U.S. 913 (2005), Polar Bear Prods., Inc. v. Timex Corp., 384 F.3d 700 (9th Cir. 2004), and Garcia v. Coleman, No. C-07-227 EMC, 2009 WL 799393 (N.D. Cal. Mar. 24, 2009)--are copyright cases and cannot support including non-infringing products in the base of a reasonable patent royalty. To the extent Oracle is saying that it can recover Google's advertising revenue for copyright infringement, Cockburn offers no analysis of any "casual nexus" between the allegedly infringed copyrights and Google's advertising. See Polar Bear, 384 F.3d at 710-11. Oracle's fourth case, Trans-World Mfg. Corp. v. Al Nyman & Sons, Inc., 750 F.2d 1552 (Fed. Cir. 1984), is a patent case, but it holds only that ancillary revenues may be considered as part of a royalty analysis. Again, this is nothing different from what Georgia-Pacific held 40 years ago—ancillary revenues may affect the appropriate royalty rate. But Uniloc makes clear that including such revenues in the royalty base is justified only if the patentee satisfies the entire market value rule. Finally, Oracle snatches from context a single sentence written by Google's damages expert Gregory Leonard, Opp'n at 9, but Leonard's point was only that ancillary revenues may be relevant to the royalty rate. He did not endorse Cockburn's approach here. Google submits the full relevant chapter of Leonard's book with this brief, so that the Court may read Leonard's comments in the intended context. Purcell Reply Decl. Ex. D at 27-67.

2. Oracle cannot recover lost profits as part of a royalty as a matter of law.

Oracle's justification of Cockburn's inclusion of the projected "loss to Oracle" from the alleged infringement—i.e., Oracle's lost profits—in the base of his royalty calculation is even weaker. Opp'n at 11. Oracle's argues that a reasonable royalty analysis may "consider" a patentee's "anticipated losses" from infringement. Id. True, but only within well defined limits that Cockburn exceeds. From the time of Georgia-Pacific, it has been appropriate to consider a patentee's likely losses in setting a royalty rate, but that's not what Cockburn has done. In calculating Oracle's purported damages using the Nash Bargaining Solution, Cockburn adds half the incremental gain to Google from the entire Android platform, including all advertising on all Android smartphones—a $2.2 billion number that bears no relation to any arguable gain to Google from alleged infringement of the intellectual property at issue—to $410 million in

11

purported Oracle lost profits. Report ¶¶286-87. The result is Cockburn's base damages number of $2.6 billion, id. ¶287--although, as noted above, he contends the "full range" of damages estimates is actually between $1.4 billion and $6.1 billion. Id. Ex. 26. This is not mere "consideration" of anticipated losses as one factor in setting a royalty rate. This id direct recovery of lost profits damages. As such, it falls far outside what Georgia-Pacific, or any of the other cases Oracle cites, permits. To recover lost profits, a patentee must comply with the Panduit standard. Neither Oracle nor Cockburn addresses that standard, much less satisfies it.

3. Oracle cannot recover "fragmentation" damages as a matter of law.

Oracle tries to minimize Cockburn's reliance on "fragmentation" of the Java platform as a factor in his royalty analysis, arguing that Cockburn "considers fragmentation only when analyzing the structure of the hypothetical license." Opp'n at 12 (emphasis in original). This makes no difference. Oracle does not cite any legal authority—in the Georgia-Pacific factors or anywhere else—that permits consideration of this sort of purported harm in a royalty analysis. As Oracle conceives it, "fragmentation" is a tort-style injury, an alleged business harm to the Java platform resulting from Google's development of the Android software. Patent law provides no remedy for this sort of "injury." Cockburn doesn't trace any causal link between alleged infringement of the asserted patent claims or copyrights (as opposed to the existence of Android generally) and the supposed fragmentation of Java. Indeed, there is no nexus between infringement and fragmentation. Regardless of whether Android infringes, the mere existence of the competing Android platform in the market would increase variability among platforms supporting the Java language. It is not grounds to increase a reasonable royalty because Oracle would prefer not to have rival companies creating competing "forks" of "Java."¹

Oracle also minimizes the impact of Cockburn's consideration of fragmentation, arguing that "fragmentation adds zero dollars to [Oracle's] expected losses." Opp'n at 12. But the

12

alleged danger posed by fragmentation is Cockburn's primary basis for opining that, as part of its multi-billion dollar royalty, Google would have made a lump-sum payment to Oracle of $900 million to $1.4 billion. Report ¶¶304-12. This amount is massive in and of itself, and [Redacted] Moreover, given the time value of money, an increase in the lump-sum payment (and corresponding decrease in the royalty stream) effectively "adds dollars" to the total amount paid by the license, who would much rather pay the same amount strung out over time. Cockburn indeed has "attached a dollar value to fragmentation." Opp'n at 12. To make matters worse, his valuation depends on a factual mistake. In its motion, Google pointed out Cockburn's error in crediting the entire value of a $900 million settlement of complex litigation between Sun and Microsoft to fragmentation. Mot. at 20-21. In reality, fragmentation was only a small part of that case. Id. at 21. At the same time, Cockburn ignored the $20 million settlement in 2001 of a separate Sun-Microsoft lawsuit regarding fragmentation specifically. Id. Oracle doesn't even try to explain Cockburn's mistake in its opposition brief, because the only explanation is a desire to inflate damages.

4. Oracle boosts its damages estimate through double counting.

Oracle tries to explain Cockburn's double counting—piling a future damages calculation on top of Oracle's request for an injunction—by claiming that Cockburn "includes a future damages calculation to be used in the event that the Court and the parties opt for equitable relief in the form of an ongoing royalty." Opp'n at 17. This is Oracle's lawyers talking; Cockburn never offers this explanation in his report. Even if the explanation were right, Cockburn still has built future damages into his billion-dollar lump-sum payment, and that is still double counting.

5. Oracle cannot recover Google's international revenues as a matter of law.

Oracle wrongly defends Cockburn's decision to base his royalty calculation on Google's international advertising revenue. Opp'n at 17. Oracle does not dispute (because it cannot) the basic rule that a patentee generally may recover damages for domestic acts of infringement, but not foreign ones. See Deepsouth Packing Co. v. Laitram Corp., 406 U.S. 518, 525 n.7 (1972); Cardiac Pacemakers, Inc. v. St. Jude Med., Inc., 576 F.3d 1348, 1358, 1362-65 (Fed. Cir. 2009) (en banc) (rejecting "damages for overseas sales").

13

Oracle mistakenly relies on several cases that actually support Google's position. Uniloc USA, Inc. v. Microsoft Corp., 632 F. Supp. 2d 147, 155-57 (D.R.I. 2009), held a patentee could recover proceeds of foreign sales where the accused system is both assembled and used in the United States. Here, even assuming the Android software was "assembled" in the United States, that software plainly is not "used" in the United States with respect to overseas phones. Neither are Google ads (which are not a proper basis of damages anyway, as already discussed) served on overseas phones "used" in the United States. DataQuill Ltd. v. Handspring, Inc., No. 01 C 4635, 2004 WL 1102309, at *5 (N.D. 111. May 10, 2004), involved profits from domestic offers of sale, not foreign ones. Likewise, Litecubes, LLC v. Northern Light Prods., Inc., 523 F.3d 1353, 1371 (Fed. Cir. 2008), a copyright case, did not involve foreign sales; "it [was] undisputed that [the infringer] sold the products directly to customers in the United States." The case does not hold that the Copyright Act reaches foreign sales. Finally, neither an induced nor a contributory infringement theory can permit Oracle to reach Google's international advertising revenue. Of all the provisions of 35 U.S.C. § 271, only section 271(f) is even arguably relevant to international activities, barring the supply, in the United States, of components later assembled into a patented invention overseas. See 35 U.S.C. § 271(f). Google's advertising is not alleged to be an infringing product, so the location of its assembly is irrelevant.

6. Oracle uses the wrong date for its hypothetical royalty negotiation.

Oracle's defense of Cockburn's decision to set the date of the hypothetical negotiation in November 2008 (when Android was fully developed) is unavailing. A hypothetical negotiation must take place at "the time infringement began." Interactive Pictures Corp. v. Infinite Pictures, Inc., 214 F.3d 1371, 1385 (Fed. Cir. 2001). Here, Oracle's own litigation position is that infringement began when Google placed the Android SDK on a server no later than November 12, 2007. Weingaertner Decl. ¶19 & Ex. S. It is irrelevant that the SDK "was not available on an open-source basis at this point." Opp'n 19. Further, Oracle's after-the-fact claim that "the precise start date here is of little consequence" is not credible. Id. Cockburn relied heavily on Google's limited window of time to get the Android software to market. If he were right, a full year's head start in negotiating would have significantly altered the range of possible outcomes.

14

7. Cockburn's use of the Nash Bargaining Solution is arbitrary and improper
under Federal Circuit law.

Several times in its brief, Oracle claims that Google "does not challenge" Cockburn's use of the Nash Bargaining Solution, "nor could it." Opp'n at 4,7. Google expressly did challenge Cockburn's approach in its opening brief. Mot. at 15. Oracle does not cite any case approving the use of the Nash methodology to set a reasonable royalty. Moreover, the Federal Circuit has rejected similar approaches. For years, patent damages experts employed the so-called "25 percent rule" in setting baseline royalty rates. But just this year, the Federal Circuit condemned that rule as "fundamentally flawed" for three reasons: the 25 percent rule (1) "fails to account for the unique relationship between the patent and the accused product"; (2) "fails to account for the unique relationship between the parties"; and (3) "is essentially arbitrary and does not fit within the model of the hypothetical negotiation." Uniloc, 632 F.3d at 1313-15. Cockburn's Nash analysis here suffers from all of these invalidating flaws. He fails to value the importance of the relevant technology to Android; he ignores Sun's licenses and negotiations with Google; and his decision to award Oracle 50% of Google's purported gain from infringement is arbitrary.

Indeed, Cockburn does not even get Nash right. He purports to calculate the incremental gain to Google from infringement, half of which he awards to Oracle (in addition to Oracle's alleged lost profits). But because he never values the patents and copyrighted material at issue, he never calculates the incremental value of infringement to Google at all. Instead, he calculates the incremental value of all of Android to Google and gives half of that to Oracle, even though he has no evidence that any of the technology at issue here is essential (or even important) to Android—and Oracle has admitted the technology is not essential even to Java. Even if the Nash analysis were the legally correct one, Cockburn fails to apply that methodology to the facts here.

III. CONCLUSION

For all the above reasons, this Court should exclude Cockburn's opinions and testimony.

Dated: July 5, 2011

KEKER & VAN NEST LLP
By: /s/ Daniel Purcell
DANIEL PURCELL
Attorneys for Defendant
GOOGLE INC.

_________________

¹ Although the Court need not and cannot resolve this issue now, there is a factual dispute about Sun's historical acceptance of fragmentation of Java. The various forks of Java -- JavaSE, JavaEE, JavaME, and JavaCard -- are themselves fragments, with applications written for one fork unable to run on others. Even more significantly, Sun repeatedly designed personalized variants of Java for customers, and applications designed to run on these variants would not run on Sun-designed versions of Java.

15