Oracle America, Inc. v. Google Inc.

Filing 493

MOTION in Limine NO. 2 filed by Google Inc.. Responses due by 10/21/2011. Replies due by 10/28/2011. (Attachments: # 1 Oracle Opposition)(Kamber, Matthias) (Filed on 10/7/2011)

Download PDF
1 2 3 4 5 6 7 8 9 10 11 12 13 14 KEKER & VAN NEST LLP ROBERT A. VAN NEST - #84065 rvannest@kvn.com CHRISTA M. ANDERSON - #184325 canderson@kvn.com 633 Battery Street San Francisco, CA 94111-1809 Telephone: 415.391.5400 Facsimile: 415.397.7188 KING & SPALDING LLP DONALD F. ZIMMER, JR. - #112279 fzimmer@kslaw.com CHERYL A. SABNIS - #224323 csabnis@kslaw.com 101 Second St., Suite 2300 San Francisco, CA 94105 Tel: 415.318.1200 Fax: 415.318.1300 KING & SPALDING LLP SCOTT T. WEINGAERTNER (Pro Hac Vice) sweingaertner@kslaw.com ROBERT F. PERRY rperry@kslaw.com BRUCE W. BABER (Pro Hac Vice) 1185 Avenue of the Americas New York, NY 10036 Tel: 212.556.2100 Fax: 212.556.2222 IAN C. BALLON - #141819 ballon@gtlaw.com HEATHER MEEKER - #172148 meekerh@gtlaw.com GREENBERG TRAURIG, LLP 1900 University Avenue East Palo Alto, CA 94303 Tel: 650.328.8500 Fax: 650.328-8508 Attorneys for Defendant GOOGLE INC. 15 UNITED STATES DISTRICT COURT 16 NORTHERN DISTRICT OF CALIFORNIA 17 SAN FRANCISCO DIVISION 18 19 Case No. 3:10-cv-03561-WHA ORACLE AMERICA, INC., 20 Plaintiff, 21 v. 22 GOOGLE INC., 23 24 Defendant. GOOGLE’S MOTION IN LIMINE NO. 2 TO EXCLUDE ALLEGED PERFORMANCE BENEFITS OF ANDROID WITH THE ACCUSED FUNCTIONALITY Judge: Hon. William Alsup Date Comp. Filed: October 27, 2010 Trial Date: October 31, 2011 25 26 27 28 581324.02 GOOGLE’S MOTION IN LIMINE NO. 2 TO EXCLUDE ALLEGED PERFORMANCE BENEFITS OF ANDROID WITH THE ACCUSED FUNCTIONALITY – CASE NO. 3:10-cv-03561-WHA 1 MOTION AND RELIEF REQUESTED 2 Under Federal Rules of Evidence 401, 402, 403 and 702, defendant Google Inc. 3 (“Google”) hereby moves the Court for an order excluding in limine all evidence and argument 4 concerning the alleged performance benefits of Android with the accused functionality. 5 DISCUSSION 6 Oracle submitted four short reports by its own employees, Robert Vandette, Noel Poore, 7 Erez Landau and Seeon Birger (“the performance reports”), that collectively purport to quantify 8 the performance of Android with and without the alleged inventions claimed in five of Oracle’s 9 asserted patents. Based on these reports, Oracle experts Dr. Mitchell (on patent infringement), 10 Dr. Goldberg (on patent validity) and Dr. Cockburn (on damages) conclude that alleged 11 infringement of the asserted patents results in performance gains on Android devices, and by 12 extension, is the basis for consumer demand and Android’s success. First, the Oracle employees who submitted the performance reports conceded that they 13 14 knew little about the Oracle patents, the asserted claims or the accused functionality. Moreover, 15 the employees lack any semblance of objectivity and failed to comply with the basic tenets of 16 Daubert and Rule 702. Second, the opinions are irrelevant to any issue of consequence in the 17 trial’s initial liability phase—and thus are inadmissible under Rule 402 except in the damages 18 phase, at most. Third, the opinions should be excluded under Rule 403 even for the damages 19 phase of trial because the limited probative value of the performance reports is substantially 20 outweighed by the danger of misleading the jury into thinking that actual Android devices 21 running actual Android applications would experience a similar performance impact. 22 1. The Performance Reports Fail to Comply with Daubert and Rule 702 23 The performance reports attempt to demonstrate the benefits of implementing five of the 24 asserted patents in the Android operating system. (The ‘447 and ‘476 patents are not addressed 25 in the reports). To accomplish this, Oracle ran various “performance benchmark” tests against 26 unmodified Android code to generate various performance statistics (e.g., how fast applications 27 are executed or how much storage space they require in memory). Oracle then modified the 28 Android code to try disabling the accused functionality, and re-ran the same tests in order to 1 581324.02 GOOGLE’S MOTION IN LIMINE NO. 2 TO EXCLUDE ALLEGED PERFORMANCE BENEFITS OF ANDROID WITH THE ACCUSED FUNCTIONALITY – CASE NO. 3:10-cv-03561-WHA 1 compare the two sets of results and point to any drop in performance as indicative of the benefits 2 provided by the patented techniques. The idea may seem simple in theory, but it is a complicated 3 task in practice, and Oracle’s employees took countless shortcuts, disregarded approved 4 methodologies, and ran particular tests that were assured (and evidently designed) to reach the 5 conclusions sought by Oracle. Most critically, they made subpar modifications to the Android 6 code that were crude, hasty and in many instances crippled substantial non-accused 7 functionalities as well. The performance reports are so fraught with errors, assumptions and an 8 overall lack of detail that, had the reports been submitted for peer review to any reputable 9 technical journal or academic institution, they would be immediately rejected. See Ellis v. 10 Costco Wholesale Corp., No. 07-15838, 2011 U.S. App. LEXIS 19060 at *25-26 (9th Cir. Sept. 11 16, 2011) (“Under Daubert, the trial court must act as a ‘gatekeeper’ to exclude junk science that 12 does not meet Federal Rule of Evidence 702’s reliability standards by making a preliminary 13 determination that the expert’s testimony is reliable.” (citing Kumho Tire Co. v. Carmichael, 526 14 U.S. 137, 145 (1999).) 15 The essential prerequisite to any expert testimony is that it rest “on a reliable foundation.” 16 Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 579, 584-587 (1993). As codified in Rule 702: 17 If scientific, technical, or other specialized knowledge will assist the trier of fact to understand the evidence or to determine a fact in issue, a witness qualified as an expert by knowledge, skill, experience, training, or education, may testify thereto in the form of an opinion or otherwise, if (1) the testimony is based upon sufficient facts or data, (2) the testimony is the product of reliable principles and methods, and (3) the witness has applied the principles and methods reliably to the facts of the case. 18 19 20 21 22 A. It is Unclear What Was Done, Why, and By Whom 23 Dr. Mitchell has no firsthand knowledge about any code modifications, test conditions, or 24 the actual tests. He participated in some phone calls, after which he believed the employees used 25 “modifications that [he] suggested” and that “the actual experiments that were carried out 26 reflected [his] suggestions and requests on how to do it.” Mitchell Dep. Tr. (attached as Ex. 2 to 27 the Declaration of Daniel Purcell in Support of Google Inc.’s Motions in Limine (“Purcell 28 Decl.”) filed herewith) at 81:16 – 82:6; see also Purcell Decl. Ex. 3 (Mitchell Patent Report) at ¶ 2 581324.02 GOOGLE’S MOTION IN LIMINE NO. 2 TO EXCLUDE ALLEGED PERFORMANCE BENEFITS OF ANDROID WITH THE ACCUSED FUNCTIONALITY – CASE NO. 3:10-cv-03561-WHA 1 196. Yet, the Oracle engineers attribute everything they did to Dr. Mitchell. Robert Vandette 2 submitted a report on the ‘104 and ‘205 patents and admitted that he “relied on Dr. Mitchell to 3 interpret [the patent].” Purcell Decl. Ex. 4 (Vandette Dep. Tr.) at 74:16-17. When asked if he 4 had the “knowledge necessary to determine what functionality should be disabled,” he responded 5 that he was “not a patent attorney” and did not “know exactly which claim in the [‘104] patent 6 corresponds to [the] functionality” he disabled. Id. at 76:23-24; 77:24-78:1. Likewise, Noel 7 Poore submitted a report regarding the ‘520 and ‘702 patents and testified that he “looked 8 through the patent[s]” but had no idea which claims were asserted, and was “not necessarily 9 familiar enough with the detail of th[e] patent[s]” to know how they related to the accused 10 functionality. Purcell Decl. Ex. 5 (Poore Dep. Tr.) at 33-35; 74-75. Erez Landau revealed that 11 he reviewed only the Abstract of the ‘720 patent and “part of the claims.” Purcell Decl. Ex. 6 12 (Landau Dep. Tr.) at 21-22. 13 In addition, Vandette, Poore and Landau all indicate in their August 9, 2011 reports that 14 they downloaded the Android code, modified it, compiled it and tested it. Purcell Decl. Ex. 7 15 (Vandette report) at ¶ 27-39; Purcell Decl. Ex. 8 (Poore report) at ¶ 12-14; Purcell Decl. Ex. 9 16 (Landau report) at ¶ 25-31. Now, in a report Oracle submitted on September 12, 2011 (after the 17 close of expert discovery), a fourth employee named Seeon Birger claims that he was the one 18 who downloaded the Android code, modified it, compiled it, and provided it to the other Oracle 19 employees. See Purcell Decl. Ex. 10 (Birger report) at ¶ 11-21. Landau admitted as much 20 during his deposition (see Purcell Decl. Ex. 6 (Landau Dep. Tr.) at 58-63), while Vandette and 21 Poore appear to have concealed the fact that someone else performed the work disclosed in their 22 reports. See, e.g., Purcell Decl. Ex. 5, (Poore Dep. Tr.) at 25:9-11 (“Q. So you did all the work 23 in this report by yourself, correct? A. That’s correct”). In short, Oracle failed to provide the 24 foundational report underlying the other performance reports until weeks later, without 25 explanation, and prejudicing Google’s discovery into the benchmarking analyses. Even more 26 appalling, Oracle refuses to present Mr. Birger for deposition. See Purcell Decl. Ex. 11 27 (September 21, 2011 email from M. Francis to M. Peters.) For these reasons, all of the 28 benchmarking analyses are legitimately in question and should be excluded for Oracle’s failure 3 581324.02 GOOGLE’S MOTION IN LIMINE NO. 2 TO EXCLUDE ALLEGED PERFORMANCE BENEFITS OF ANDROID WITH THE ACCUSED FUNCTIONALITY – CASE NO. 3:10-cv-03561-WHA 1 to comply with its expert-discovery obligations. B. Oracle Made Improper Modifications to the Android Code 2 Oracle’s modifications of Android code improperly disabled or corrupted functionality in 3 4 Android that is not accused of infringement. For example, Vandette disabled the JIT compiler 5 for his ‘104 patent tests, even though the JIT compiler was never accused in connection with that 6 patent. Purcell Decl. Ex. 4 (Vandette Dep. Tr.) at 84:18-25. With respect to the ‘205 patent, 7 which is asserted against a single feature of the JIT compiler, Vandette again disabled the whole 8 JIT compiler, even though the ‘205 patent itself recognizes the existence of JIT compilers in the 9 prior art. For tests relating to the ‘702 patent, Poore modified Android code so existing 10 duplicates would remain in the Android “.dex” files—but his report openly admits that the 11 modified code created new duplicates that did not previously exist. Purcell Decl. Ex. 8 (Poore 12 Report) at ¶ 23. Landau admitted that when disabling the copy-on-write functionality allegedly 13 claimed by the ‘720 patent, his test program loaded 1,264 files into memory, but readily agreed 14 that “if copy-on-write is disabled, it would just be a waste of memory to load these [files] …” 15 and it would make no sense to have Android run that way. Purcell Decl. Ex. 6 (Landau Dep. Tr.) 16 at 86:1-87:24. As a whole, the performance reports fail to account for changes beyond the 17 accused functionality responsible for differences in performance between the original and 18 modified code. See Claar v. Burlington N.R.R., 29 F.3d 499 (9th Cir. 1994) (testimony excluded 19 where the expert failed to consider other obvious causes for the plaintiff’s condition). Oracle’s 20 experts offer no recognized methodology guiding their modifications of the Android source 21 code. Under Daubert, the Court needs to exclude expert testimony unless the “methodology 22 underlying the testimony is scientifically valid.” 509 U.S. at 592-93. C. Oracle’s Reports Draw Conclusions Not Supported by Its Tests 23 Even assuming that Oracle’s performance reports quantify the difference in running 24 25 Android code with and without the accused functionality, Oracle uses the test results to make 26 unfounded conclusions that those results are indicative of the patents’ impact on the real-world 27 performance of Android, as follows: 28  Vandette and Poore did not test the performance of any actual Android applications; 4 581324.02 GOOGLE’S MOTION IN LIMINE NO. 2 TO EXCLUDE ALLEGED PERFORMANCE BENEFITS OF ANDROID WITH THE ACCUSED FUNCTIONALITY – CASE NO. 3:10-cv-03561-WHA 1  Vandette and Poore did not test the performance of any actual Android devices; 2  Landau tested 17 Android applications at one time on the Nexus One device released nearly 3 two years ago, and—as mentioned above—he modified the code to operate in a manner that 4 would never be implemented by Google and made no practical sense. See Purcell Decl. Ex. 5 12 (Landau Dep. Ex. 494) (Android’s website specifically states that the Nexus One should 6 not be used for testing); 7  simply crippled the code in a manner that would never actually be implemented; 8 9 None of the employees attempted to implement reasonable non-infringing substitutes—they  None of the tests measure overall system performance on an Android device and they fail to 10 recognize that the Dalvik VM is used only 1/3 of the time by Android—the other 2/3 of time 11 Android is executing native machine code. See Purcell Decl. Ex. 13 (Vandette Dep. Ex. 462, 12 May 2010 Google IO presentation) at slide 5; 13  source code and not any particular code release that has been used by actual devices; 14 15 Oracle’s employees downloaded and tested a development version of Android’s “Froyo”  There is nothing in the employee reports that indicates whether they followed the specific 16 instructions on Android’s website for downloading, compiling and installing Android on 17 devices. (For example, Landau testified that Mr. Birger used version 6 of the JDK software 18 to compile Android “Froyo” code, whereas the Android website specifically requires that 19 version 5 of the JDK be used. See Purcell Decl. Ex. 6 (Landau Dep. Tr.) at 61:17-21; Purcell 20 Decl. Ex. 14 (Landau Dep. Ex. 491) Taking these factors into account, “there is simply too great an analytical gap between the 21 22 data and the opinion proffered.” General Elec. Co. v. Joiner, 522 U.S. 136, 146 (1997). All of 23 Oracle’s expert conclusions based on these reports amount to just “subjective belief or 24 unsupported speculation” and should be excluded. Daubert, 509 U.S. at 588-90 (“[I]f an opinion 25 is fundamentally unsupported, then it offers no expert assistance to the jury.”). 26 27 2. The Performance Reports Should Be Excluded From the Liability Phase of the Case under Rule 402 28 Evidence regarding the purported performance impact of the asserted patents is 5 581324.02 GOOGLE’S MOTION IN LIMINE NO. 2 TO EXCLUDE ALLEGED PERFORMANCE BENEFITS OF ANDROID WITH THE ACCUSED FUNCTIONALITY – CASE NO. 3:10-cv-03561-WHA 1 potentially probative of only damages theories, and all damages issues have been bifurcated. 2 Thus, the purported performance impact of the asserted patents has no bearing on alleged 3 infringement or invalidity and has no relevance to the liability phase of the trial. 4 As stated in Google’s motion regarding “commercial success,” claims about the 5 performance with and without the asserted patents pose at least a “modest likelihood” of unfair 6 prejudice and a more than a “small risk” of misleading the jury because it tends to create an 7 inflated sense of the patent’s importance to the success of Android. See United States v. 8 Gonzalez-Flores, 418 F.3d 1093, 1098 (9th Cir. 2005) (applying Rule 403 and holding that 9 “[w]here the evidence is of very slight (if any) probative value, it’s an abuse of discretion to 10 admit it if there’s even a modest likelihood of unfair prejudice or a small risk of misleading the 11 jury”) (citation omitted); see also United States v. Espinoza-Baza, 647 F.3d 1182, 1190 (9th Cir. 12 2011) (“evidence presenting even a ‘modest likelihood of unfair prejudice’ is ‘high enough to 13 outweigh the . . . probative value’ of marginally relevant evidence” under Rule 403) (citation 14 omitted). 15 16 3. The Performance Reports Should Be Excluded From the Damages Phase of the Case under Rule 403 17 As explained above, Oracle’s performance reports fail to analyze the performance impact 18 of the accused infringing features on real-world Android devices and applications. Furthermore, 19 the results are unreliable because they do not account for realistic non-infringing alternatives or 20 the detrimental impact of Oracle’s sloppy code modifications to non-infringing Android 21 functionality. In view of these limitations, Oracle’s reports offer little probative value, which 22 value is substantially outweighed by its capacity to be used in prejudicial ways. For example, 23 Oracle may contend that consumers’ Android products would suffer a noticeable performance 24 impact without the accused functionality. But this premise is unsupported by the performance 25 reports, which focus on micro-benchmarks of the virtual machine, not overall system 26 performance. Even if Oracle presents testimony within the scope of the reports, it threatens to 27 mislead the jury to conclude that actual Android performance will mimic Oracle’s test results, 28 given all the flaws discussed above. Either way, the Court should exclude evidence regarding 6 581324.02 GOOGLE’S MOTION IN LIMINE NO. 2 TO EXCLUDE ALLEGED PERFORMANCE BENEFITS OF ANDROID WITH THE ACCUSED FUNCTIONALITY – CASE NO. 3:10-cv-03561-WHA 1 the performance reports. See Rule 403 (“evidence may be excluded if its probative value is 2 substantially outweighed by the danger of unfair prejudice, confusion of the issues, or misleading 3 the jury…”); Daubert, 509 U.S. at 595 (“Expert evidence can be both powerful and quite 4 misleading because of the difficulty in evaluating it.” (citation omitted)). 5 As a final note, in order to safeguard the effectiveness of its ruling and thus avoid the 6 danger of prejudice, the Court also should bar counsel for both parties from making any remarks 7 in opening argument about the alleged performance benefits of the asserted patents in Android. Respectfully submitted, 8 9 Dated: September 24, 2011 KEKER & VAN NEST LLP 10 11 12 By: s/ Robert A. Van Nest ROBERT A. VAN NEST Attorneys for Defendant GOOGLE INC. 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 7 581324.02 GOOGLE’S MOTION IN LIMINE NO. 2 TO EXCLUDE ALLEGED PERFORMANCE BENEFITS OF ANDROID WITH THE ACCUSED FUNCTIONALITY – CASE NO. 3:10-cv-03561-WHA

Disclaimer: Justia Dockets & Filings provides public litigation records from the federal appellate and district courts. These filings and docket sheets should not be considered findings of fact or liability, nor do they necessarily reflect the view of Justia.


Why Is My Information Online?