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
`
`26
`
`27
`
`28
`
`Case 5:11-cv-01079-PSG Document 346-19 Filed 02/14/14 Page 1 of 7
`
`
`
`
`
`SARAH BARROWS (SBN 253278)
`barrowss@gtlaw.com
`William Coats (SBN 94864)
`coatsw@gtlaw.com
`Stephen Ullmer (SBN 277537)
`ullmers@gtlaw.com
`GREENBERG TRAURIG, LLP
`4 Embarcadero Center, Suite 3000
`San Francisco, CA 94111-5983
`Telephone: (415) 655-1300
`Facsimile: (415) 707-2010
`
`James J. DeCarlo (Admitted Pro Hac Vice)
`decarloj@gtlaw.com
`Michael A. Nicodema (Admitted Pro Hac Vice)
`nicodemam@gtlaw.com
`GREENBERG TRAURIG, LLP
`MetLife Building
`200 Park Avenue, 34th Floor
`New York, New York 10166
`Tel.: (212) 801-9200
`Fax: (212) 801-6400
`
`Attorneys for Defendant,
`Apple Inc.
`
`UNITED STATES DISTRICT COURT
`FOR THE NORTHERN DISTRICT OF CALIFORNIA
`SAN JOSE DIVISION
`
`EMBLAZE LTD.,
`
`
`
`
`Plaintiff;
`v.
`
`APPLE INC., a California Corporation,
`
`Defendant.
`
`
`
`POLISH DECL.
`
`CASE NO. 11-CV-01079 PSG
`
`DECLARATION OF DR. NATHANIEL
`POLISH, PH. D., IN SUPPORT OF
`DEFENDANT APPLE INC.’S MOTION FOR
`SUMMARY JUDGMENT OF NON-
`INFRINGEMENT
`
`
`
`
`
`CASE NO. 11-CV-01079 PSG
`
`

`

`Case 5:11-cv-01079-PSG Document 346-19 Filed 02/14/14 Page 2 of 7
`
`
`
`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
`
`26
`
`27
`
`28
`
`I, Nathaniel Polish, Ph. D., hereby declare:
`
`1.
`
`I have been asked to provide this declaration regarding Defendant Apple Inc.’s Motion for
`
`Summary Judgment of Non-Infringement (“Apple’s Motion”) in the above-captioned case. I have
`
`personal knowledge of the facts stated herein.
`
`2.
`
`My analysis and opinions detailed in this declaration are consistent with my December 20,
`
`2013 Expert Report regarding non-infringement. I understand that Apple has been accused of
`
`infringing U.S. Patent No. 6,389,473 (“’473 patent”). My report includes my opinion rebutting the
`
`infringement analysis provided by Dr. Vijay Madisetti, who provided an expert report on
`
`infringement on behalf of Emblaze on November 8, 2013.
`
`3.
`
`I understand that my report and exhibits were all provided to Emblaze’s counsel on
`
`December 20, 2013. I was also deposed regarding my opinions on January 30, 2014, in New York,
`
`New York. If asked, I could and would provide my report and accompanying exhibits to the Court,
`
`but for purposes of Apple’s Motion I have been asked to provide sworn testimony (consistent with
`
`my opinion on the same issues in the report) regarding certain issues addressed in Apple’s Motion.
`
`4.
`
`I reviewed the report and exhibits provided by Dr. Madisetti, as well as his deposition
`
`testimony, including test data regarding certain streams allegedly affiliated with MLB, the NFL,
`
`ESPN, ABC, CBS/PGA, and Apple (“the Accused Streams”). I have also done some of my own
`
`testing with respect to the Accused Streams.
`
`Predetermined Data Size Limitation:
`
`5.
`
`I understand that every claim that Apple has been accused of infringing includes the
`
`limitation “each slice having a predetermined data size associated therewith” and that the Court
`
`construed this claim to mean “Each slice having a data size, which may be established by setting a
`
`time duration of the slice, assigned in advance of the stream being divided.”
`
`6.
`
`Per the claim term and the Court’s construction, it is my opinion that a person of ordinary
`
`skill in the art (at any time, including as of March 24, 1998) (“ordinary artisan”) would have
`
`understood that setting a time duration may result in a data size assigned in advance but that the
`
`“may” condition will only result in a predetermined data size assigned in advance if the data rate has
`
`
`
`POLISH DECL.
`
`- 1 -
`
`
`
`CASE NO. 11-CV-01079 PSG
`
`

`

`Case 5:11-cv-01079-PSG Document 346-19 Filed 02/14/14 Page 3 of 7
`
`
`
`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
`
`26
`
`27
`
`28
`
`a constant value. If the data rate of the stream varies from slice to slice and such variability is not
`
`knowable or predictable in advance, setting the time duration is not establishing a data size assigned
`
`in advance.
`
`7.
`
`The “data rate” of a stream cannot be assumed to be a constant value particularly in the case
`
`of a compressed video stream. An ordinary artisan would understand that, with respect to an
`
`uncompressed stream (e.g. a broadcast television feed), setting a time duration could result in slices
`
`having data sizes assigned in advance because an uncompressed stream generally has a constant
`
`data rate. In addition, some audio compression schemes such as the GSM 6.10 compression
`
`specifically described in the ’473 patent have a constant data rate, as explained in the patent and as
`
`is known to a skilled artisan. However, encoded (i.e. compressed) video streams generally do not
`
`have a constant data rate and thus setting the time duration does not result in slices having a data
`
`size assigned in advance.
`
`8.
`
`Essentially all video encoders use principles of human perception in deciding what to encode
`
`and with what detail. As a result, scenes with details that humans do not perceive very well are
`
`coded with fewer bits than scenes with many details that humans can perceive. Also, essentially all
`
`video encoders only encode the frame-to-frame differences. Thus, time periods where the video
`
`does not change much from frame to frame will be encoded with fewer bits than time periods with
`
`rapidly changing frames. So any given time period will vary substantially in the number of bits
`
`used relative to other time periods. Thus, in the case of most encoded video outputs, the data size
`
`cannot be assigned in advance of the stream being sliced.
`
`The testing relied upon by Dr. Madisetti to show that the accused streams are divided into
`9.
`slices having a data size assigned in advance of the stream being divided1 does not support that
`
`
`1 E.g., Data produced by Dr. Madisetti at VKM000377-VKM386 including exhibits 12-14 to the
`Declaration of Douglas Wieder in Support of Apple’s Motion for Summary Judgment of Non-
`Infringement of U.S. Patent No. 6,389,473 as to All Accused Streams and excerpts of that data
`referenced in exhibit 2 (Dr. Madisetti’s infringement expert report) to the Declaration of Stephen
`Ullmer in Support of Apple’s Motion for Summary Judgment of Non-Infringement of U.S. Patent
`No. 6,389,473 as to Specific Content Providers (“Specific Providers Ullmer Decl.”) at ¶¶ 122, 202-
`03, 214, 215.
`
`
`POLISH DECL.
`
`- 2 -
`
`
`
`CASE NO. 11-CV-01079 PSG
`
`

`

`Case 5:11-cv-01079-PSG Document 346-19 Filed 02/14/14 Page 4 of 7
`
`
`
`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
`
`26
`
`27
`
`28
`
`conclusion because the data sizes vary without a predictable pattern. Further, contrary to what Dr.
`
`Madisetti has said, my opinion is that an ordinary artisan would not consider “generally equal” to be
`
`a substitute for a claim term requiring “a data size … assigned in advance.”
`
`Upload Rate Generally Equal To Data Rate Of The Stream Limitation:
`
`10.
`
`I understand that every claim that Apple has been accused of infringing includes the
`
`limitation “uploading the sequence to a server at an upload rate generally equal to the data rate of
`
`the stream” (Claim 1 and its dependents) or “which uploads the sequence to a server at an upload
`
`rate generally equal to the data rate” (Claim 25 and its dependents) and that the Court construed
`
`these limitations to mean “[t]ransmitting the files from the transmitting computer to the server at an
`
`upload rate generally equal to the data rate of the stream.”
`
`11.
`
`In my opinion, an ordinary artisan would not have understood the “upload rate” to refer to a
`
`rate that includes the time during which no uploading occurs. None of the examples in the patent
`
`support this view, nor would an ordinary artisan refer to an upload (or “transmitting” as per the
`
`Court’s construction) that takes 3 seconds for a file containing 10 seconds of media as having taken
`
`10 seconds to upload or transmit. I described specific examples from the patent in my report and in
`
`several exhibits, including the patent’s description (e.g. ’473 patent at 11:60-61, 12:59-13:23) of an
`
`upload time variable referred to as “TSL” the time in which a files is actually being uploaded. The
`patent also places significant emphasis on maintaining a generally equal relationship between the
`
`data rate of the stream and the upload rate (in fact, this is a limitation for all of the claims), including
`
`describing adjusting the compression level to “adjust the data streaming rate to the available
`
`bandwidth.” (’473 patent at 7:45-49).
`
`12.
`
`An ordinary artisan reading these descriptions, the rest of the specification, and the claims
`
`would have understood that the “upload rate” is not an inherent quality of live streaming (or, put
`
`another way, the upload rate is not “necessarily equal” to the data rate of the stream for live
`
`streaming) but that some kind of generally equal relationship must be maintained between the two to
`
`meet the claims of the patent. Otherwise, there would be no recognition that the two variables are
`
`not aligned (i.e., that the actual transmitting occurs much faster than the data rate) and that the
`
`
`
`POLISH DECL.
`
`- 3 -
`
`
`
`CASE NO. 11-CV-01079 PSG
`
`

`

`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
`
`26
`
`27
`
`28
`
`Case 5:11-cv-01079-PSG Document 346-19 Filed 02/14/14 Page 5 of 7
`
`
`
`compression rate should be adjusted in view of the teachings of the ’473 patent to take advantage of
`
`the additional bandwidth.
`Dr. Madisetti’s Testimony
`Transmitting Computer:
`I understand that every claim that Apple has been accused of infringing requires a
`
`that Encoder-Plus-Akamai Architecture
`
`Is “A”
`
`13.
`
`“transmitting computer” (e.g. in Claim 1: “A method for real-time broadcasting from a transmitting
`
`computer to one or more client computers over a network” and in Claim 25: “Apparatus . . .
`
`comprising: a transmitting computer”).
`
`I also reviewed Dr. Madisetti’s testimony where he was asked questions about the Akamai
`
`14.
`RTMP Architecture during his deposition. (E.g., Ex. 8 to the Specific Providers Ullmer Decl. at
`
`368:21-377:11). In answering those questions, he identified that the claimed transmitting computer
`
`could be either a combination of the Encoder, Entrypoint, RtmpCore, and Archiver, or that same
`
`combination plus the NetStorage, as he described in his deposition. (See, e.g., id.; Ex. 9 to the
`
`Specific Providers Ullmer Decl. at 5 (depicting Akamai’s RTMP Architecture)).
`
`15.
`
`In my opinion, an ordinary artisan would not have understood the transmitting computer
`
`described in the ’473 patent to include the kind of specialized encoder and server architecture
`
`required to implement the Akamai RTMP Architecture. As I opined in my report in considerable
`
`detail, the patent describes expressly its goals of eliminating the need for dedicated broadcast
`
`encoders and hardware, using networks with common, existing server network infrastructure such as
`
`those offered by ISPs, and using conventional personal computers to provide streaming multimedia.
`
`(’473 patent at 1:50-67, 6:37-43, 6:62-63). These goals, especially in view of the other disclosures
`
`and claims in the patent, would indicate to an ordinary artisan that the transmitting computer itself
`
`cannot include the type of specialized server infrastructure that make up the Akamai RTMP
`
`Architecture.
`
`Testing data associated with HLS streams:
`
`16.
`
`As I discussed in my report and accompanying exhibits, I performed substantial testing of
`
`HLS streams and produced the results in my report. These tests included testing of some of the
`
`Accused Streams identified in Dr. Madisetti’s report. For these tests, I measured the latency
`
`
`
`POLISH DECL.
`
`- 4 -
`
`
`
`CASE NO. 11-CV-01079 PSG
`
`

`

`Case 5:11-cv-01079-PSG Document 346-19 Filed 02/14/14 Page 6 of 7
`
`
`
`between the time at which a scene appears on a live TV feed and the time at which the same scene
`
`appeared on certain Apple devices including an iPad and an iPhone. The latency I observed varied
`
`from a low of about 25 seconds to a high that exceeded 200 seconds of delay. Many of my results,
`
`including results for both the accused ESPN and ABC news streams, featured delays of 35-40
`
`seconds when tested.
`Analysis of the Software Appendix Submitted with the application that became the
`’473 patent
`17. My analysis and opinions detailed in this declaration are also consistent with my November
`
`9, 2013 Expert Report regarding the software appendix contained in the file history the ’473 patent.
`
`My report includes my opinion that software files identified in the specification of the ’473 patent
`
`that were purportedly produced with the application that became the ’473 patent were irretrievably
`
`corrupted.
`
`18.
`
`I understand that my report and exhibits were all provided to Emblaze’s counsel on
`
`November 9, 2013. If asked, I could and would provide my report and accompanying exhibits to
`
`the Court, but for purposes of Apple’s Motion I have been asked to provide sworn testimony
`
`(consistent with my opinion on the same issues in the report) regarding certain issues addressed in
`
`Apple’s Motion.
`
`19.
`
`The ’473 patent references nine software files in a portion of the specification that is labeled
`
`Software Appendix. (’473 patent at 13:53-14:16). These files are also purportedly disclosed in the
`
`appendix. The appendix consists of a printout that an ordinary artisan would recognize as a hex
`
`dump, where the first set of columns expresses an address in base 16, a second set of columns
`
`consists of 16 pairs of hexadecimal numbers, and a final set of columns contains a text
`
`representation of the data in the second set of columns. Generally speaking, an ordinary artisan
`
`would know how to utilize a hex dump to run the program(s) contained in a hex dump or to create
`
`pseudo code by disassembling the files in the hex dump using a code disassembler.
`
`20. Many of the pages of the Software Appendix are missing the first set of columns and the first
`
`character of the second set of columns. An ordinary artisan could not retrieve the files with this
`
`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
`
`26
`
`27
`
`28
`
`
`
`POLISH DECL.
`
`- 5 -
`
`
`
`CASE NO. 11-CV-01079 PSG
`
`

`

`Case 5:11-cv-01079-PSG Document 346-19 Filed 02/14/14 Page 7 of 7
`
`
`
`amount of missing information due to the truncated columns—these files are not computer-readable
`
`and they could not be used to make computer-readable files or pseudo code.
`
`I declare under penalty of perjury the foregoing is true and correct.
`
`DATED: February 14, 2014
`
`GREENBERG TRAURIG, LLP
`
`By:
`
`
`
`Nathaniel Polish, Ph. D.
`Expert for Defendant
`Apple Inc.
`
`
`
`
`
`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
`
`26
`
`27
`
`28
`
`POLISH DECL.
`
`- 6 -
`
`
`
`CASE NO. 11-CV-01079 PSG
`
`