Oracle America, Inc. v. Google Inc.

Filing 497

Declaration of DANIEL PURCELL in Support of #496 MOTION in Limine No. 5, #494 MOTION in Limine No. 3, #492 MOTION in Limine No. 1, #493 MOTION in Limine NO. 2, #495 MOTION in Limine No. 4 filed byGoogle Inc.. (Attachments: #1 Exhibit 1, #2 Exhibit 2, #3 Exhibit 3, #4 Exhibit 4, #5 Exhibit 5, #6 Exhibit 6, #7 Exhibit 7, #8 Exhibit 8, #9 Exhibit 9, #10 Exhibit 10, #11 Exhibit 11, #12 Exhibit 12, #13 Exhibit 13, #14 Exhibit 14, #15 Exhibit 15, #16 Exhibit 16, #17 Exhibit 17, #18 Exhibit 18, #19 Exhibit 19, #20 Exhibit 20, #21 Exhibit 21, #22 Exhibit 22, #23 Exhibit 23, #24 Exhibit 24, #25 Exhibit 25, #26 Exhibit 26, #27 Exhibit 27, #28 Exhibit 28, #29 Exhibit 29, #30 Exhibit 30, #31 Exhibit 31, #32 Exhibit 32, #33 Exhibit 33, #34 Exhibit 34, #35 Exhibit 35, #36 Exhibit 36, #37 Exhibit 37, #38 Exhibit 38, #39 Exhibit 39, #40 Exhibit 40)(Related document(s) #496 , #494 , #492 , #493 , #495 ) (Kamber, Matthias) (Filed on 10/7/2011)

Download PDF
EXHIBIT 4 HIGHLY CONFIDENTIAL-- ATTORNEYS' EYES ONLY Page 1 UNITED STATES DISTRICT COURT NORTHERN DISTRICT OF CALIFORNIA SAN FRANCISCO DIVISION ------------------------ORACLE AMERICA, INC., ) Plaintiff, ) vs. ) No. CV 10-03561 WHA GOOGLE, INC., ) VOLUME I Defendant. ) ------------------------- HIGHLY CONFIDENTIAL - ATTORNEYS' EYES ONLY Videotaped Deposition of ROBERT VANDETTE, taken at 42 Chauncy Street, Boston, Massachusetts, commencing at 10:02 a.m., Wednesday, September 7, 2011, before Jill Shepherd, RPR, MA-CSR No. 148608, NH-CSR No. 128, CA-CSR No. 13275, CLR, and Notary Public. PAGES 1 - 124 Veritext National Deposition & Litigation Services 866 299-5127 HIGHLY CONFIDENTIAL-- ATTORNEYS' EYES ONLY 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 case, and I will generally refer to it as 12:05:43 the "'104 patent." 12:05:47 A. Okay. 12:05:50 Q. Have you seen this patent before? 12:05:50 A. Yes. I have seen this patent. 12:05:52 Q. When? 12:05:54 A. When I was asked to assist in the 12:05:56 measurements that are documented in my 12:06:03 report. 12:06:08 Q. And you indicated, was that April/May time? 12:06:08 A. Yeah, spring. 12:06:17 Q. Did you review the '104 patent? 12:06:19 A. I did read it. 12:06:22 Q. Do you believe you understand it? 12:06:23 MS. AGRAWAL: Objection. Form. 12:06:34 A. I'm not a patent attorney, so I relied on 12:06:34 Dr. Mitchell to interpret it. 12:06:45 Q. Have you reviewed the asserted claims of the 12:06:47 '104 patent? 12:06:54 MS. AGRAWAL: Objection. Form. 12:06:54 A. As I have already mentioned, I did see an 12:06:57 early draft of the asserted claims around 12:07:01 the same April/May time frame. 12:07:05 Q. Do you know which claims are currently being 12:07:09 asserted from this patent? 12:07:12 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Q. Do you understand that there are 12:09:10 different -- many different claims in the 12:09:22 back of this patent that are being asserted 12:09:23 in this case? 12:09:25 MS. AGRAWAL: Objection. Form. 12:09:26 A. I don't have that level of detail. I was -- 12:09:28 as I mentioned, I came in April/May. The 12:09:35 discussions and the claims were developed 12:09:38 prior to that. 12:09:41 Q. So your only knowledge of what Android 12:09:46 functionality should be disabled for your 12:09:49 tests is based on what you were told by 12:09:54 Peter Kessler and Professor Mitchell; is 12:09:57 that correct? 12:10:01 MS. AGRAWAL: Objection. Form. 12:10:01 A. We came to the conclusion of what to disable 12:10:05 during these discussions with them. I'm not 12:10:20 sure if I -12:10:25 Q. You personally did not have knowledge 12:10:27 necessary to determine what functionality 12:10:29 should be disabled? 12:10:32 MS. AGRAWAL: Objection. Form. 12:10:33 A. I am not a patent attorney, so I did not 12:10:35 have the skill to interpret patent law. 12:10:42 Q. Is Professor Mitchell or Peter Kessler a 12:10:46 Page 74 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 MS. AGRAWAL: Objection. Form. 12:07:14 A. I have kind of a high level understanding of 12:07:19 what the claims are asserting. 12:07:47 Q. In paragraph 26 of your report, on 12:07:50 page eight, you say, and I quote, "I ran 12:08:04 experiments to disable the Android 12:08:09 functionality that Oracle accuses of 12:08:11 infringing the '104 patent"; is that 12:08:15 correct? 12:08:18 A. Which page are you on? 26? 12:08:18 Q. Paragraph 126. 12:08:20 A. Oh, 126. 12:08:22 Q. 26. 12:08:24 A. What page are you on? 12:08:25 Q. Page eight. 12:08:28 A. Page eight. That's what I missed. Yes. 12:08:29 Q. How do you know what functionality was 12:08:36 accused? 12:08:38 MS. AGRAWAL: Objection. Form. 12:08:40 A. I had conversations with Dr. Mitchell where 12:08:41 he explained the infringing areas, and also 12:08:49 had discussions with Mr. Peter Kessler, and 12:08:56 through these discussions, we came up with 12:09:03 modifications to the Android sources that 12:09:06 could disable this functionality. 12:09:08 Page 76 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 patent attorney? 12:10:49 MS. AGRAWAL: Objection. Form. 12:10:49 A. My assumption is that they worked with 12:10:51 patent attorneys outside of the discussions 12:10:58 with me. 12:11:02 Q. Do you understand that different claims in 12:11:04 the '104 patent might accuse slightly 12:11:10 different -- might be asserted against 12:11:14 slightly different functionality and they 12:11:17 are not identical? 12:11:19 MS. AGRAWAL: Objection. Form. 12:11:22 A. You have to be more specific. I don't -- 12:11:22 it's too general a question. 12:11:34 Q. Were you asked to disable functionality 12:11:42 corresponding to any particular claim of the 12:11:45 '104 patent? 12:11:50 MS. AGRAWAL: Objection. Form. 12:11:51 A. It was explained to me that I needed to 12:11:53 disable the SideTable class resolution of 12:12:09 methods, classes, and fields, and that the 12:12:16 quickening of byte codes also needed to be 12:12:22 disabled -12:12:24 Q. You don't -12:12:29 A. -- but I don't know exactly which claim in 12:12:30 the patent corresponds to that 12:12:32 Page 75 Page 77 Pages 74 to 77 Veritext National Deposition & Litigation Services 866 299-5127 HIGHLY CONFIDENTIAL-- ATTORNEYS' EYES ONLY 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 functionality. 12:12:37 Q. In paragraph 40 to 41, on page ten of your 12:12:38 report, is there a significant difference in 12:13:04 performance between the original Froyo 12:13:13 Android code and your modified code with 12:13:15 side tables, but no quickening instructions? 12:13:18 MS. AGRAWAL: Objection. Form. 12:13:21 A. So as you can see from the chart below, 12:13:22 there's not much of an impact. There is 12:13:31 some impact. 12:13:35 Q. Would you say that impact is significant? 12:13:38 MS. AGRAWAL: Objection. Form. 12:13:41 A. If you look at the spreadsheet with the data 12:13:43 and look at the standard deviation, it is 12:13:50 statistically significant. 12:13:54 Q. In your professional opinion, regarding the 12:13:57 performance of the Dalvick Virtual Machine, 12:14:02 do you believe this change in performance is 12:14:10 significant? 12:14:12 MS. AGRAWAL: Objection. Form. 12:14:13 A. As I stated at the beginning of my report, 12:14:14 at Oracle and other companies, we spend a 12:14:24 lot of engineering effort to get even small 12:14:26 incremental performance improvements. So, 12:14:30 yes, I do believe it's significant in that 12:14:36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 A. I can't quantify that. 12:15:57 Q. Are there less than 100 improvements? 12:15:58 MS. AGRAWAL: Same objection. 12:16:02 A. There's actually no way -- I have not been 12:16:04 involved in every Java VM and library 12:16:16 improvement that has gone on. There's no 12:16:19 way for me to even come up with a number. 12:16:22 Q. How many people have been involved in this 12:16:25 task? 12:16:30 MS. AGRAWAL: Objection. Form. 12:16:30 A. Any number I give you would just be a guess. 12:16:31 Q. That's fine. 12:16:57 MS. AGRAWAL: Objection. Form. 12:16:58 A. And so the fact that I have to tell the 12:17:00 truth here, I'm not going to guess. 12:17:07 Q. On page four of your report in paragraph 14, 12:17:08 you say, "The engineering team at Oracle 12:17:22 continues to invest heavily in R&D targeted 12:17:25 at increasing performance. A large portion 12:17:28 of our virtual machine engineering team is 12:17:30 typically dedicated to performance 12:17:34 improvement projects." 12:17:34 And then you go on to give some 12:17:35 examples; is that correct? 12:17:38 A. Yes. 12:17:39 Page 78 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 regard. 12:14:38 Q. How much time has Oracle and Sun before that 12:14:40 spent making incremental improvements to 12:14:58 Java Virtual Machine? 12:15:03 MS. AGRAWAL: Objection. Form. 12:15:05 A. There's no way to quantify that. It's been 12:15:07 going on since the beginning of Java. 12:15:12 Q. How many incremental improvements are there? 12:15:14 MS. AGRAWAL: Objection. Form. 12:15:17 A. We could produce, you know, a chart of 12:15:20 SPECjbb improvements since the beginning of 12:15:23 time, but every release has increased. 12:15:25 Q. So there's been an ongoing chain of 12:15:30 improvements? 12:15:37 A. Right. 12:15:37 Q. Going back over ten to 15 years -12:15:37 MS. AGRAWAL: Objection. 12:15:40 Q. -- at least? 12:15:41 MS. AGRAWAL: Form. 12:15:42 A. There's -- yes. There's been ongoing 12:15:44 performance improvements since Java first 12:15:47 came out. 12:15:50 Q. How many improvements would there be? Ten? 12:15:50 50? 100? 12:15:53 MS. AGRAWAL: Objection. Form. 12:15:55 Page 80 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Q. And you said that this has been going on 12:17:39 since the beginning of Java development? 12:17:43 A. Correct. 12:17:45 MS. AGRAWAL: Objection. Form. 12:17:45 Q. Java was released in somewhere in '94, 95; 12:17:46 is that correct? 12:17:52 MS. AGRAWAL: Objection. Form. 12:17:52 A. I would have to refer back to some files to 12:18:01 know the exact date of Java. It was 12:18:04 probably the early '90s, but the exact date 12:18:07 I don't know. This paragraph is talking 12:18:10 only about the virtual machine performance 12:18:11 improvements. There are other teams working 12:18:13 in other areas of Java. 12:18:15 Q. Okay. 12:18:17 Would it be wrong to make the 12:18:26 assumption that there are probably thousands 12:18:28 of incremental improvements that have been 12:18:30 going on since the mid '90s with respect to 12:18:33 the Java Virtual Machine? 12:18:36 MS. AGRAWAL: Objection. Form. 12:18:38 A. You are asking me to guess, and I can't 12:18:40 guess. 12:18:44 Q. You don't know? 12:18:51 MS. AGRAWAL: Objection. Form. 12:18:52 Page 79 Page 81 Pages 78 to 81 Veritext National Deposition & Litigation Services 866 299-5127 HIGHLY CONFIDENTIAL-- ATTORNEYS' EYES ONLY 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Q. I'm not asking you for a specific number. 12:18:58 Give me a general ballpark figure. 12:19:01 MS. AGRAWAL: Objection. Form. 12:19:03 A. You are asking me to describe changes or 12:19:04 incremental performance improvements in 12:19:11 groups that I really wasn't involved in. 12:19:14 Q. Turning to page eight, paragraph 28 of your 12:19:47 report, you discuss the modifications that 12:19:50 you made to conduct your experiments; is 12:20:00 that correct? 12:20:03 A. Yes. 12:20:03 Q. And you created these modifications based on 12:20:03 what you were told by Professor Mitchell and 12:20:14 Peter Kessler; is that correct? 12:20:18 MS. AGRAWAL: Objection. Form. 12:20:19 A. Yes. We discussed the functionality, and 12:20:20 Peter I both looked through the sources to 12:20:27 try to find out how to properly disable this 12:20:31 functionality, and we came to a consensus. 12:20:33 Q. It appears that you attempted two out of 12:20:42 three possible scenarios here? 12:20:46 A. That's correct. 12:20:49 MS. AGRAWAL: Sorry, objection. 12:20:49 Form. 12:20:51 Q. The first scenario in paragraph 28 is 12:20:53 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 A. It would impact the results potentially, 12:22:15 since I'd be adding additional functionality 12:22:18 to Dalvick that it doesn't currently have. 12:22:20 Q. It would, however, be technically possible 12:22:24 for someone to do so? 12:22:26 MS. AGRAWAL: Objection. Form. 12:22:34 A. Let's see. It may be technically possible 12:22:34 to build a system that does quickening 12:22:57 without side tables, but it would involve 12:23:01 adding additional overhead that Dalvick 12:23:03 doesn't currently have. 12:23:08 Q. In paragraph 36, you state that you did not 12:23:10 try running the trace compiler; is that 12:23:19 correct? 12:23:23 A. Oh, paragraph -- sorry. That is correct, 12:23:23 for the same reason that we didn't do the 12:23:35 quickening alone. 12:23:37 Q. What is the trace compiler? 12:23:38 MS. AGRAWAL: Objection. Form. 12:23:42 A. That is Dalvick's implementation of a JIT. 12:23:43 Q. Are you saying that for your performance 12:23:51 benchmark regarding the '104 patent you had 12:24:01 to disable the JIT? 12:24:04 MS. AGRAWAL: Form. 12:24:06 A. That's correct. 12:24:07 Page 82 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 building side tables, but not quickening 12:20:58 instructions, and the second scenario is not 12:20:59 building side tables or quickening 12:21:03 instructions; is that correct? 12:21:06 A. That's correct. 12:21:08 Q. Is there a third possibility of not building 12:21:10 side tables, but building quickening 12:21:16 instructions? 12:21:20 MS. AGRAWAL: Objection. Form. 12:21:22 A. The quickening was dependent upon the side 12:21:22 table for its implementation in order to 12:21:27 avoid, you know, any possible error in the 12:21:30 results. We did not want to substantially 12:21:37 modify Dalvick in order to try to attempt 12:21:39 that. We wanted to restrict our changes to 12:21:43 just simple commenting out of code that 12:21:47 would provide the before and after. 12:21:48 Q. So if it got too complicated, you did not 12:21:51 attempt it? 12:21:55 MS. AGRAWAL: Objection. Form. 12:21:55 A. It's not an issue of complication. It's an 12:21:56 issue of possibly altering the Dalvick to 12:22:00 the point where I wouldn't be measuring what 12:22:10 I wanted to measure. 12:22:12 Q. It would be -12:22:13 Page 84 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Q. Do you understand that the JIT is not part 12:24:08 of the accused functionality of the '104 12:24:13 patent? 12:24:18 MS. AGRAWAL: Objection. Form. 12:24:18 A. I'm not certain that it isn't somehow 12:24:19 involved in some of the claims, but we 12:24:40 focused on turning off the functionality in 12:24:46 a mode that was possible. 12:24:49 Q. If, in fact, JIT is not part of the accused 12:24:54 functionality, then wouldn't disabling it 12:25:08 affect the performance of this benchmark? 12:25:16 MS. AGRAWAL: Objection. Form. 12:25:20 A. I do believe that the numbers would be 12:25:21 slightly different; however, the overhead of 12:25:28 having to re-resolve all of the classes, 12:25:34 fields, and methods is a fixed overhead that 12:25:38 the JIT could not compensate for. So I 12:25:42 believe the performance reduction or 12:25:47 degradation would still be substantial. 12:25:52 Q. Despite fixed overhead, you are referring to 12:26:03 other aspects of the benchmarking programs 12:26:08 might execute faster if the JIT was enabled; 12:26:10 is that correct? 12:26:14 MS. AGRAWAL: Objection. Form. 12:26:15 A. They would be severely diminished by the 12:26:16 Page 83 Page 85 Pages 82 to 85 Veritext National Deposition & Litigation Services 866 299-5127 HIGHLY CONFIDENTIAL-- ATTORNEYS' EYES ONLY 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 I declare under penalty of perjury under the laws that the foregoing is true and correct. Executed on _________________ , 20___, at _____________, ___________________________. 1 2 3 4 5 6 7 8 9 10 INDEX WITNESS ROBERT VANDETTE Examination by Mr. Francis PAGE EXHIBITS NO. DESCRIPTION PAGE Exhibit 460 Defendant's Notice of 8 Rule 30(B)(1) Deposition of Robert ("Bob") G. Vandette Exhibit 461 Summary and Report of 10 Robert ("Bob") Vandette 11 _____________________________ ROBERT VANDETTE 12 Exhibit 462 Presentation (A JIT Compiler for Dalvik VM) 41 13 Exhibit 463 Dhrystone White Paper 15 16 Exhibit 464 OAGOOGLE0014151624 OAGOOGLE0014151630 Exhibit 465 OAGOOGLE0011198304 OAGOOGLE0011198305 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 63 66 17 Exhibit 466 US Patent No. RE 38,104 73 Exhibit 467 US Patent No. 6,910,205 91 18 19 20 21 Exhibit 468 OAGOOGLE0014151618 OAGOOGLE0014151623 Exhibit 469 OAGOOGLE0019108076 OAGOOGLE0019108077 113 Exhibit 470 OAGOOGLE0023688735 OAG00GLE0023688745 Exhibit 471 OAGOOGLE0024644910 115 114 22 23 24 25 Page 122 1 2 3 4 5 6 58 14 118 Page 124 CERTIFICATE COMMONWEALTH OF MASSACHUSETTS MIDDLESEX, SS. I, Jill Shepherd, Notary Public, in and for the Commonwealth of Massachusetts, do hereby certify that: ROBERT VANDETTE, the witness whose deposition taken on September 7, 2011 is hereinbefore set forth, was satisfactorily identified by means of driver's license, and was duly sworn by me, and that the foregoing transcript is a true and accurate record of the testimony given by such witness and such testimony is a true and accurate transcription of my stenotype notes to the best of my knowledge, skill, and ability. I further certify that I am not related to any of the parties in this matter by blood or marriage and that I am in no way interested in the outcome of this matter. IN WITNESS WHEREOF, I have hereunto set my hand and notarial seal this 8th day of September, 2011. _________________________ Jill Shepherd, RPR Notary Public My Commission expires: April 18, 2014 Page 123 Pages 122 to 124 Veritext National Deposition & Litigation Services 866 299-5127

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?