`
`
`
`
`
`
`
`
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`____________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`____________
`
`NETFLIX, INC., and HULU, LLC
`Petitioner,
`
`
`v.
`
`DIVX, LLC,
`Patent Owner.
`____________
`
`Case IPR2020-00648
`Patent 9,998,515
`____________
`
`
`
`PATENT OWNER’S SUR-REPLY
`TO PETITION FOR INTER PARTES REVIEW
`
`
`
`
`
`TABLE OF CONTENTS
`
`
`Case IPR2020-00648
`Patent 9,998,515
`
`
`Page
`
`I.
`
`II.
`
`Dr. Zeger’s Credible Opinions Outweigh Dr. Reader’s. ................................. 1
`
`The Reply’s Section III Is Irrelevant ............................................................... 3
`
`III.
`
`Petitioner Does Not Dispute Patent Owner’s Claim Constructions ................ 5
`
`IV. Pyle Combination ............................................................................................ 5
`
`A.
`
`B.
`
`C.
`
`Pyle Does Not Teach Generating A Top Level Index File In
`Response To A Request For Content. ................................................... 5
`
`Pyle Does Not Teach “Retrieving” And “Filtering” A “List of
`Assets” ................................................................................................... 9
`
`Pyle Does Not Teach Generating A Top Level Index File
`Describing Each Asset In The Filtered List Of Assets ....................... 13
`
`D.
`
`Limitation 1[c] ..................................................................................... 14
`
`V.
`
`Lewis Combination ........................................................................................ 16
`
`A.
`
`B.
`
`C.
`
`D.
`
`Lewis Does Not Teach The “Retrieving” Limitation ......................... 16
`
`Lewis Does Not Teach “Filtering The List Of Assets.” ..................... 20
`
`Limitations 1[c] And 1[e] .................................................................... 22
`
`Lewis Does Not Teach “Generating A Top Level Index File” ........... 24
`
`VI. Pyle And Lewis Are Not Available Prior Art ............................................... 24
`
`VII. Unconstitutionality ........................................................................................ 25
`
`VIII. Conclusion ..................................................................................................... 25
`
`
`
`- i -
`
`
`
`TABLE OF AUTHORITIES
`
`
`
`Case IPR2020-00648
`Patent 9,998,515
`
`
`Page(s)
`
`COURT DECISIONS
`
`Game & Tech. Co. v. Activision Blizzard Inc.,
`926 F.3d 1370 (Fed. Cir. 2019) .......................................................................... 23
`
`Harari v. Lee,
`656 F.3d 1331 (Fed. Cir. 2011) .......................................................................... 12
`
`In re Dembiczak,
`175 F.3d 994 (Fed. Cir. 1999) ............................................................................ 21
`
`In re Epstein,
`32 F.3d 1559 (Fed. Cir. 1994) ............................................................................ 11
`
`In re Varma,
`816 F.3d 1352 (Fed. Cir. 2016) .......................................................................... 12
`
`OddzOn Prods., Inc. v. Just Toys, Inc.,
`122 F.3d 1396 (Fed. Cir. 1997 ............................................................................ 25
`
`TiVo, Inc. v. Echostar Communs. Corp.,
`516 F.3d 1290 (Fed. Cir. 2008) .......................................................................... 12
`
`Uber Techs., Inc. v. X One, Inc.,
`957 F.3d 1334 (Fed. Cir. 2020) .......................................................................... 11
`
`STATUTES
`
`35 U.S.C. § 311(b) ..................................................................................................... 4
`
`35 U.S.C. § 316(e) ..................................................................................................... 3
`
`
`
`
`
`- ii -
`
`
`
`AGENCY DECISIONS
`
`Case IPR2020-00648
`Patent 9,998,515
`
`
`Denso Corp. v. Collision Avoidance Techs. Inc.,
`IPR2017-01715, Paper 27 (PTAB Jan. 22, 2019) .............................................. 12
`
`Ralph Lauren Corp. v. Lexos Media IP, LLC,
`IPR2018-01749, Paper 21 (PTAB Apr. 3, 2020) ........................................ 12, 14
`
`TRW Auto US LLC v. Magna Elecs. Inc.,
`IPR2014-00262, Paper 27 (PTAB June 25, 2015) ............................................. 14
`
`Unified Patents Inc., v. DivX, LLC,
`IPR2019-01379, Paper 52 (PTAB Feb. 8, 2021) ......................................... 19, 23
`
`RULES, RULEMAKING, AND OTHER AGENCY AUTHORITIES
`
`37 C.F.R. § 42.104(b)(3) ..................................................................................... 3, 11
`
`37 C.F.R. § 42.104(b)(4) ............................................................................................ 4
`
`37 C.F.R. § 42.6(a)(3) .............................................................................................. 23
`
`
`
`
`
`
`
`
`
`- iii -
`
`
`
`Case IPR2020-00648
`Patent 9,998,515
`
`
`EXHIBIT LIST
`
`2001
`
`U.S. Provisional Application 61/366,059 [Pyle Provisional]
`
`2002
`
`2003
`
`2004
`
`2005
`
`2006
`
`2007
`
`Wowza Media Systems, “Using dynamically generated SMIL file
`possible?”, http://www.wowza.com/forums/showthread.php?12458-
`Using-dynamically-generated -SMIL-file-possible [Wowza]
`
`Wowza Media Systems, “What is transcoding and how does it impact
`streaming quality?”, https://www.wowza.com/uploads/images/
`Wowza-Tech-Brief-Transcoding-and-Its-Impact.pdf [Transcoding]
`
`Verizon Media, HDS, HLS, HSS -- Adaptive HTTP Streaming
`Demystified (July 11, 2013), https://www.verizondigitalmedia.com/
`blog/hds-hls-hss-adaptive-http-streaming/ [Verizon]
`
`Christopher Mueller, Bitmovin, Microsoft Smooth Streaming (May
`21, 2015), https://bitmovin.com/microsoft-smooth-streaming-mss/
`[MSS]
`
`Alex Zambelli, Microsoft Corp., IIS Smooth Streaming
`Technical Overview (Mar. 2009),
`https://www.bogotobogo.com/VideoStreaming/Files/iis8/
`IIS_Smooth_Streaming_Technical_Overview.pdf [IIS Overview]
`
`K. Hong, Ph.D., Smooth Streaming on IIS 8 - 2020
`https://www.bogotobogo.com/VideoStreaming/
`AdaptiveLiveStreaming_SmoothStreaming_iis8.php [IIS Smooth
`Streaming]
`
`2008
`
`PAIR Transaction History for U.S. Application 10/009,593
`[Transaction History]
`
`2009 Declaration of Bridget A. Smith [served but not filed]
`
`2010 Deposition of Clifford Reader, Jan. 5, 2021 [Reader Tr.]
`
`
`
`- iv -
`
`
`
`Case IPR2020-00648
`Patent 9,998,515
`
`
`2011
`
`2012
`
`Definition of “list” from Oxford English & Spanish Dictionary,
`Thesaurus, and Spanish to English Translator,
`https://www.lexico.com/en/definition/list [Oxford]
`
`Definition of “list” from The American Heritage Dictionary of the
`English Language, https://www.ahdictionary.com/word/
`search.html?q=list [Am. Heritage]
`
`2013 Definition of “list” from Dictionary.com, https://
`www.dictionary.com/browse/list [Dictionary.com]
`
`2014 GSMA IMEI Database, https://imeidb.gsma.com/imei/index#
`[GSMA-1]
`
`2015
`
`GSMA TAC Allocation and IMEI Programming Rules for Device
`Brand Owners and Manufacturers: Training Guide v1.0, Feb. 2018
`[GSMA-2]
`
`2016 Declaration of Kenneth A. Zeger, Ph.D. [Zeger Decl.]
`
`
`
`- v -
`
`
`
`Case IPR2020-00648
`Patent 9,998,515
`Petitioner’s Reply arguments fail to plug the fatal gaps in the Petition.
`
`I.
`
`DR. ZEGER’S CREDIBLE OPINIONS OUTWEIGH DR. READER’S.
`
`The conclusory arguments by Petitioner are not supported by Dr. Reader’s
`
`opinions, which merely parrot Petitioner’s attorney arguments. POR, 13. The
`
`Board has disregarded similar unsupported arguments in other cases, including one
`
`that rejected Dr. Reader’s own testimony that was similarly unsupported. Id., 13-
`
`14. Petitioner chooses to respond to this observation not by rehabilitating Dr.
`
`Reader’s credibility, but by insinuating that Patent Owner’s expert Dr. Zeger also
`
`lacks credibility. Reply, 4. Petitioner’s argument is both misplaced and
`
`unsupported.
`
`In the first place, if neither expert were credible, the Board should rule
`
`against Petitioner. If Petitioner failed to present credible evidence in support of its
`
`grounds, they fail even without Patent Owner needing to present a rebuttal, for
`
`Petitioner bears the burden of proof.
`
`In any event, Petitioner’s attacks on Dr. Zeger fail even on their own terms.
`
`First, Petitioner argues Dr. Zeger’s credibility is impugned by a Board case
`
`unrelated to this one in which the Board did not entirely agree with his opinions.
`
`Id. But Petitioner does not contend Dr. Zeger’s testimony in that case (or any other
`
`case) was conclusory or incredible. Id. The Board’s resolution of a dispute
`
`between credible expert opinions in another case has no bearing on the credibility
`- 1 -
`
`
`
`
`
`Case IPR2020-00648
`Patent 9,998,515
`of Dr. Zeger’s well-supported opinions in this case. Id. And Petitioner fails to
`
`show that Dr. Reader boasts a track record before the Board as good as Dr. Zeger.
`
`Petitioner slings more mud by mischaracterizing Dr. Zeger’s testimony.
`
`Petitioner argues Dr. Zeger would not “admit” in deposition that Pyle or Lewis
`
`teach an adaptive bit rate streaming system. Reply, 4. But Dr. Zeger was asked
`
`whether Pyle or Lewis were examples of “an adaptive bit rate streaming system as
`
`that term was known in 2011.” Ex. 1010, 89:6-91:14. He explained that he hadn’t
`
`offered an opinion on how “adaptive bit rate streaming system” was known then or
`
`if that would appropriately apply to Pyle and Lewis. Id. Therefore, he explained,
`
`he “wouldn’t be able to give a quick answer.” Id. Dr. Zeger can hardly be faulted
`
`for needing time to evaluate new questions.
`
`Petitioner asserts broadly that Dr. Zeger was “unwilling to describe the state
`
`of the art in 2011.” Reply, 5. But again, these were particularized questions on
`
`issues where Dr. Zeger had offered no opinion. For example, “what did a database
`
`refer to in 2011 to a POSITA?” Ex. 1010, 26:16-28:3. Dr. Zeger answered he did
`
`not recall offering “an opinion construing what a database was to a POSITA.” Id.
`
`Nevertheless, he immediately offered “one example”. Id. Petitioner’s mud again
`
`fails to stick.
`
`Finally, Petitioner suggests Dr. Zeger’s testimony that “a list is a very well-
`
`known concept outside of the patent” contradicts his declaration testimony on the
`- 2 -
`
`
`
`
`
`Case IPR2020-00648
`Patent 9,998,515
`“retrieving” a “list of assets” limitations. Reply, 5. Petitioner fails to cite any
`
`deposition testimony in tension with Dr. Zeger’s declaration testimony regarding
`
`the concept of a list. Id.
`
`Dr. Zeger’s testimony is unimpeached and entitled to full weight at trial.
`
`II. THE REPLY’S SECTION III IS IRRELEVANT
`
`Section III of the Reply complains that supposedly (i) Patent Owner did not
`
`offer constructions for the “retrieving” and “filtering” limitations, (ii) the patent
`
`does not describe these limitations in detail, and (iii) Dr. Zeger did not identify
`
`specific passages teaching these limitations. Reply, 5-8. Petitioner’s arguments
`
`have no merit.
`
`First, Petitioner complains Patent Owner did not propose constructions for
`
`“retrieving” and “filtering.” But Petitioner has the burden to prove its grounds
`
`and explain “[h]ow the challenged claim is to be construed.” 37 C.F.R.
`
`§ 42.104(b)(3); 35 U.S.C. § 316(e). In any event, Petitioner is wrong. Patent
`
`Owner did argue that the required “retrieving” and “filtering” must occur for the
`
`same “list of assets.” POR, 6-7.
`
`Next, Petitioner argues Patent Owner seeks a narrow meaning for these
`
`terms here but a broader meaning in related infringement litigation. Reply, 5.
`
`Once again, Petitioner fails to actually identify any contradictions in Patent
`
`Owner’s positions. Id.
`
`
`
`- 3 -
`
`
`
`Case IPR2020-00648
`Patent 9,998,515
`Next, Petitioner suggests the ’515 patent does not adequately describe the
`
`
`
`claimed “retrieving” and “filtering.” Reply, 6. Petitioner points to four subsections
`
`of the specification that supposedly do not “provide any more detail for how to
`
`perform the retrieving or filtering limitations as compared to the Petition.” Id. But
`
`the Patent contains further detailed disclosures of these limitations. See e.g., Ex.
`
`1001, 11:12-12:38, 12:39-13:55, 5:60-6:2, 7:57-8:33. Petitioner does not even
`
`suggest that these limitations are unsupported or non-enabled. Rather, Petitioner
`
`seems to suggest that, because the Patent does not describe them in even more
`
`detail, Petitioner is not obligated to show how the limitations are met by its
`
`combinations. Reply, 6. Petitioner offers no law supporting this theory—
`
`unsurprisingly, as it is contrary to Board rules. 37 C.F.R. § 42.104(b)(4).
`
`Petitioner fails to explain how the Patent’s alleged lack of more specific
`
`implementation details for “retrieving” and “filtering” supports Petitioner’s case.
`
`
`
`Finally, Petitioner returns to attacking Dr. Zeger, alleging he could not
`
`“identify where the patent specifically teaches how to perform the ‘retrieving’ and
`
`‘filtering’ limitations.” Reply, 7. But Dr. Zeger was asked whether two sections in
`
`isolation “would be the sections to look at … to understand what is required for the
`
`retrieving and filtering limitations of the patent.” Ex. 1010, 154:7-11. He
`
`responded he had not “index[ed] … which sections had which material” and had
`
`offered no opinion on what parts of the patent might alone be enough to teach the
`- 4 -
`
`
`
`
`
`Case IPR2020-00648
`Patent 9,998,515
`“retrieving” and “filtering” limitations. Id., 155:15-25; 164:1-165:14; 165:15-
`
`166:10. Dr. Zeger cannot be faulted for declining to take on-the-fly positions on
`
`questions for which he had not needed to offer an opinion. And Petitioner fails to
`
`explain how Dr. Zeger’s declining to propose ready-fire-aim opinions on how
`
`these sections “explain implementing the retrieving or filtering limitations”
`
`supposedly supports Petitioner’s grounds.
`
`III. PETITIONER DOES NOT DISPUTE PATENT OWNER’S CLAIM
`CONSTRUCTIONS
`
`“[T]he list of assets” filtered in limitation 1[d] (“filtering the list of assets”)
`
`is the same “list of assets” retrieved in limitation 1[c] (“retrieving a list of assets”).
`
`POR, 19-20. The Reply does not dispute this.
`
`“[T]he filtered list of assets” in limitation 1[e] (“generating a top level index
`
`file describing each asset in the filtered list of assets.”) is the list of assets resulting
`
`from limitation 1[d]’s “filtering the list of assets.” POR, 21-22. Again, no dispute.
`
`IV. PYLE COMBINATION
`
`A.
`
`Pyle Does Not Teach Generating A Top Level Index File In
`Response To A Request For Content.
`
`Pyle, unlike the Patent, does not generate a customized top level index file in
`
`response to a request from a device. POR, 3-6. Rather, Pyle responds to requests
`
`for content by sending a pre-existing manifest currently stored on the system. Id.
`
`
`
`- 5 -
`
`
`
`Case IPR2020-00648
`Patent 9,998,515
`The Reply repeats the Petition’s assertion that Pyle generates manifest files
`
`in response to requests for content. But for the first time, it raises an untimely
`
`argument that Pyle’s Figure 4 supports this contention. Reply, 8-9.1 Wrong.
`
`Pyle Figure 4 does not teach “responding to a request for content with a
`
`selected manifest or a new manifest.” Id. Pyle Figure 4 shows two separate
`
`functions of Pyle’s composition component 210, as annotated below: (1)
`
`facilitating the creation of new manifests, which are then stored amongst other
`
`manifests, and (2) responding to requests for content by sending selected pre-
`
`existing manifests.
`
`
`1 Pyle’s Figure 4 is cited by the Petition and Dr. Reader, but neither ever presented
`
`
`
`this argument. Pet., 34-35; Ex. 1003 ¶183-184.
`
`
`
`- 6 -
`
`
`
`Case IPR2020-00648
`Patent 9,998,515
`
`Ex. 1004, Fig. 4 (excerpted) (annotated).
`
`Pyle teaches that composition component 210 can “facilitate composition of
`
`new manifest 422” which is then “stored amongst other manifests 204.” Ex. 1004,
`
`10:57-60. Pyle never discloses composition component 210 responding to a
`
`request 420 by sending new manifest 422. Id., 10:57-11:10. Pyle teaches that the
`
`arrow labeled 422 in Figure 4 shows a new manifest that has been composed with
`
`the help of composition component 210 being stored amongst the other manifests.
`
`Pyle teaches that the manifest(s) that are sent in response to requests are different
`
`manifests, represented by a differently numbered feature of Figure 4. Pyle
`
`explains that other arrows, labeled 420 and 418, show the composition component
`
`receiving a request for content, and responding with a selected pre-existing
`
`manifest. Ex. 1004, 10:22-27.
`
`[C]omposition component 210 can be further configured to select a
`particular selected manifest 418 from the set of available manifests 204
`based upon data included in a request 420 for content 206. Accordingly,
`selected manifest 418 can be transmitted to a requesting device.
`
`Id. Thus, Pyle Figure 4 shows two functions of the composition component: (i)
`
`responding to request for content 420 with a selected pre-existing manifest 418;
`
`and (ii) facilitating the creating of a new manifest 422 and storing this new
`
`manifest amongst the other manifests. Id., 10:22-27, 10:57-11:10. Pyle does not
`
`
`
`- 7 -
`
`
`
`Case IPR2020-00648
`Patent 9,998,515
`teach responding to requests for content by creating new manifest files. The
`
`Reply’s assertion to the contrary simply misreads Pyle. POR, 9.
`
`Pyle teaches creating—and storing—new manifests “optimized” to known
`
`information on particular devices, including, inter alia, what has received the most
`
`requests, deliveries, and/or presentations. Ex. 1004, 10:60-11:9. But Pyle does not
`
`create new manifests tailored to respond, and then send them in response, to
`
`specific requests. Instead, Pyle follows a different process: it stores multiple
`
`manifests tailored to be optimized for various possible requests, then selects and
`
`sends an existing stored manifest in response to each request. This is unsurprising,
`
`as Pyle is designed to not generate updated manifests for each request. See Ex.
`
`1004, 5:12-17 (noting that “one of the benefits” for which Pyle is specifically
`
`designed is to “reduce the MPD [(i.e., manifest)] live update traffic” in its system).
`
`Pyle does not teach waiting to receive each request to generate a manifest, then
`
`sending a newly-generated manifest optimized to the request. It does the opposite:
`
`it stores multiple manifests together, organized in a complex hierarchical system so
`
`the best stored manifest can be selected when a request is received. Id., 8:17-9:44,
`
`10:40-56. It may create new manifests optimized to be likely to be responsive to
`
`future requests, but they are created only to be “stored amongst other manifests”
`
`already in its organized storage. Id., 10:57-11:9; see also, e.g., Ex. 2010 [Reader
`
`Tr.] 37:20-38:12. Pyle does not generate manifest files in response to requests.
`- 8 -
`
`
`
`
`
`Case IPR2020-00648
`Patent 9,998,515
`Petitioner points to nothing in Marusi that renders it obvious, in combination
`
`with Pyle, to generate on-the-fly manifests in response to requests.
`
`B.
`
`Pyle Does Not Teach “Retrieving” And “Filtering” A “List of
`Assets”
`
`The Petition fails to show Pyle-Marusi teaches a “list of assets” that is both
`
`“retriev[ed]” and “filter[ed].” POR, 6-19.
`
`
`
`Petitioner argues Patent Owner improperly focused on Pyle, not Marusi.
`
`Reply 9-11. But the Petition’s argument focused on Pyle, not Marusi, relying on
`
`Pyle for the “list of assets” allegedly “retriev[ed]” and “filter[ed].” Pet., 30-34.
`
`And so does the Reply, arguing that “Pyle’s server retrieves a list of assets,” and
`
`“Pyle’s server maintains manifests (a list of assets).” Id.
`
`Petitioner now argues that the Petition did not rely “entirely on Pyle for the
`
`‘list of assets.’” Reply, 11. Is Petitioner referring to the Petition’s suggestion that
`
`Marusi discloses “records of the database” that might constitute “a list of assets” as
`
`claimed? Pet., 34. But the Reply does not repeat that allegation, only contending
`
`vaguely that “the retrieving and filtering limitations are rendered obvious by
`
`Marusi’s database teachings” and “Marusi is an example of a reference” that
`
`“describe[s] basic implementation details.” Id., 11. The Reply does not describe
`
`which “database teachings” and “basic implementation details” might combine
`
`with Pyle to make retrieving and filtering the list of assets obvious. The Reply’s
`
`
`
`- 9 -
`
`
`
`Case IPR2020-00648
`Patent 9,998,515
`use of Marusi mirrors its approach in the Petition, which, for the “retrieving” and
`
`“filtering” a “list of assets,” relied on Pyle except for vague references to Marusi.
`
`Neither the Petition nor the Reply makes any attempt to show that Marusi teaches
`
`retrieving the records of its database before filtering them. The Reply never
`
`articulates how combining Pyle and Marusi might render it obvious to both receive
`
`and filter a list of assets.
`
`
`
`Petitioner asserts “‘filtering … the list of assets,’ is made obvious by Pyle’s
`
`“using multiple manifest files to provide a new or optimized manifest.” Reply, 12.
`
`As already shown, however, this is a blatant misread of Pyle. POR, 9. Pyle says:
`
`by employing multiple manifests 204, individual manifests 204 can be
`optimized for particular delivery formats, wire formats… and so
`forth ….
`
`Ex. 1004, 8:23-27. As explained in the POR, this does not teach using two stored
`
`manifests to create a new third manifest. Instead, because Pyle maintains, e.g., two
`
`manifests, one can be an HD manifest for HD devices and the other a standard
`
`definition manifest for standard definition devices. POR, 9 (citing Ex. 1004, 9:6-
`
`9). Dr. Reader agrees. See Ex. 2010 [Reader Tr.] 37:20-38:12.
`
`Petitioner excuses its failure to articulate its obviousness arguments by
`
`arguing that the “Board should reject any argument that the Petition does not
`
`adequately describe the claimed features because the Petition’s teachings are
`
`
`
`- 10 -
`
`
`
`Case IPR2020-00648
`Patent 9,998,515
`commensurate with the ’515 patent’s disclosure.” Reply, 13. As shown above,
`
`however, the ’515 patent contains numerous disclosures regarding “retrieving” and
`
`“filtering” a “list of assets.” Supra Section II. Petitioner fails to explain how the
`
`patent’s alleged lack of implementation details could somehow eliminate the legal
`
`requirements for Petitioner to demonstrate how a limitation is met. 37 C.F.R. §
`
`42.104(b)(3). Petitioner says “aspects of claimed features that are not described by
`
`a patent itself demonstrates [sic] that it is within the skill and knowledge of a
`
`POSITA,” Reply, 14, but the case Petitioner cites did not even involve a missing
`
`limitation: it involved motivation to combine. Uber Techs., Inc. v. X One, Inc.,
`
`957 F.3d 1334, 1339-1340 (Fed. Cir. 2020).2 In Uber the record showed only two
`
`possible methods to “transmit user locations and maps,” both of which “were
`
`undisputedly known in the prior art” and disclosed by the cited art, so the court
`
`found that the patent-at-issue’s brief disclosure on this issue suggested that a
`
`POSITA was “capable of selecting between the known methods.” Id. This case is
`
`nothing like Uber. This is not a case where references disclose a small number of
`
`methods and the question is the rationale to select one of them. Here, neither
`
`
`2 Petitioner also cites In re Epstein, 32 F.3d 1559, 1568 (Fed. Cir. 1994), but it
`
`deals with enablement, which is not at issue here.
`
`
`
`- 11 -
`
`
`
`Case IPR2020-00648
`Patent 9,998,515
`asserted reference or combination teaches the “retrieving” and “filtering”
`
`limitations at all.
`
`Next, Petitioner presents a new claim construction argument, attempting to
`
`construe “a list of assets” as “one or more list [sic] of assets.” Reply, 13-14.
`
`Petitioner offers no reason this argument could not have been raised in the Petition.
`
`Id. The Board should follow its routine practice and reject this untimely argument.
`
`Ralph Lauren Corp. v. Lexos Media IP, LLC, IPR2018-01749, Paper 21, 11
`
`(PTAB Apr. 3, 2020) (refusing to consider “new construction” “raised improperly
`
`for the first time in the Reply”); Denso Corp. v. Collision Avoidance Techs. Inc.,
`
`IPR2017-01715, Paper 27, 35 (PTAB Jan. 22, 2019) (similar). Moreover,
`
`Petitioner ignores the countless decisions that refuse to construe such language to
`
`cover plural items. The law is that “whether ‘a’ or ‘an’ is treated as singular or
`
`plural depends heavily on the context of its use” and that the plural rule Petitioner
`
`cites “does not apply when the context clearly evidences that the usage is limited to
`
`the singular.” TiVo, Inc. v. Echostar Communs. Corp., 516 F.3d 1290, 1303 (Fed.
`
`Cir. 2008) (reversing broad plural construction); see In re Varma, 816 F.3d 1352,
`
`1362 (Fed. Cir. 2016) (rejecting plural construction because “context matters”).
`
`Here, much as in Harari v. Lee, 656 F.3d 1331, 1341 (Fed. Cir. 2011) (rejecting
`
`plural construction “in light of the claim and specification”), “[t]he plain language
`
`of the claim clearly indicates that only a single [list] is used” to list multiple assets.
`- 12 -
`
`
`
`
`
`Case IPR2020-00648
`Patent 9,998,515
`See Ex. 1001, 2:23-34, 3:1-5, 4:21-23; 12:60-67 (“a list of assets that satisfies
`
`criteria,” “a list of assets to which a…device is granted access”), 17:59-66.
`
`Finally, even if this argument were not untimely and unsupported, Petitioner does
`
`not show its combination both “retriev[es]” and “filter[s]” even one “list of assets.”
`
`
`
`Finally, Petitioner argues Patent Owner is wrong to say Pyle does not teach
`
`retrieving multiple manifests. According to the Reply, Pyle teaches “retrieving all
`
`of the manifests…so that the server could choose among manifests and the
`
`representations contained in manifests.” Reply, 14 (citing Ex. 1003 ¶175). But
`
`neither the Reply nor Dr. Reader cites anything in Pyle describing such a cattle-call
`
`of “all” manifests. Petitioner alternatively argues it was “obvious for Pyle’s server
`
`to retrieve a list of assets…because [Pyle] uses the multiple manifests to choose an
`
`optimal manifest.” Id. But neither Petitioner nor Dr. Reader shows any reason that
`
`choosing the one optimal manifest would call for retrieving the multiple manifests.
`
`C.
`
`Pyle Does Not Teach Generating A Top Level Index File
`Describing Each Asset In The Filtered List Of Assets
`
`As Patent Owner has explained, the Petition nowhere shows Pyle teaches
`
`“generating a top level index file,” or one “describing each asset in the filtered list
`
`of assets.” POR, 19-21. In response, the Reply (at 15) points to Pyle allegedly
`
`“creating a new manifest file in response to the request for content that is
`
`transmitted to the requester,” citing Pet., 34-35. But this argument is not in the
`
`
`
`- 13 -
`
`
`
`Case IPR2020-00648
`Patent 9,998,515
`Petition, which merely argues that “the manifest file sent … is generated because
`
`the manifest file must be placed in a memory for transmission to a client device.”
`
`Pet., 34-35. This untimely new argument should be ignored. Ralph Lauren, Paper
`
`21, 11; TRW Auto US LLC v. Magna Elecs. Inc., IPR2014-00262, Paper 27, 12-13
`
`(PTAB June 25, 2015). Even if this argument were timely it is wrong: Pyle does
`
`not teach generating new manifest files in response to requests for content. POR,
`
`3-6.
`
`Next, the Reply argues Pyle teaches “generating” because a “manifest file is
`
`placed in memory for transmission.” Reply, 15. As already explained, Petitioner
`
`fails to support its conclusory argument that “generating” a file is met be storing a
`
`file in memory. POR, 20. Petitioner has no response.
`
`Furthermore, Petitioner fails to even identify a “filtered list of assets.” POR,
`
`21. The Reply’s only response is that supposedly “[t]he petition explains that the
`
`new manifest file or the selected manifest file is the result of the earlier retrieving
`
`and filtering limitations.” Reply, 15 (citing Pet., 35). This was not argued in the
`
`Petition. Pet., 35. It is thus both baseless and untimely.
`
`D. Limitation 1[c]
`
`The Petition nowhere proves Pyle-Marusi teaches using a product identifier
`
`to identify “a device software version indicating a version number for an adaptive
`
`streaming software component implemented on the playback device”; Pyle alone
`- 14 -
`
`
`
`
`
`Case IPR2020-00648
`Patent 9,998,515
`does not disclose or render obvious this limitation. POR, 23-24. Petitioner does
`
`not even respond to this point, and it is now undisputed.
`
`The combination relies on adding Marusi’s Type Allocation Code (TAC) to
`
`Pyle’s request, but adding Marusi does not fill the gap. POR, 24-27. This TAC
`
`does not include a software version number. POR, 25. The Reply has no
`
`response and thus concedes this flaw, but responds by citing instead to a different
`
`part of Marusi. Reply, 16-17 (citing Marusi ¶6). But as the POPR has explained
`
`(incorporated here by reference), Marusi’s ¶6 teaching is a version number for
`
`Marusi’s “last software upgrade of the actual phone,” not “for an adaptive
`
`streaming component” as claimed. POPR, 31. The Reply ignores this fact.
`
`Next, Petitioner mischaracterizes Dr. Zeger’s testimony and argues that he
`
`was either ignorant or lying. Reply, 17. Dr. Zeger was asked whether “a POSITA
`
`would not have known that the software version of a media player is a relevant
`
`concern in choosing media for playback.” Ex. 1010, 234:14-17. Dr. Zeger
`
`responded that neither the Petition nor Dr. Reader’s declaration “explained that in
`
`any detail other than possibly just stating it in a conclusory manner” and that he
`
`was “not really aware of why that would be true in [Petitioner/Dr. Reader’s]
`
`theory.” Id., 234:23-235:1. Dr. Zeger also explained that he “didn’t specifically
`
`consider that question because it’s different than the question that arises in these
`
`IPRs.” Id., 236:6-10. Petitioner’s baseless criticisms should be disregarded.
`- 15 -
`
`
`
`
`
`Case IPR2020-00648
`Patent 9,998,515
`Finally, Petitioner cites a new reference, Thoen, submitted with the Reply.
`
`Reply, 17. Thoen is not part of the asserted combination, was produced after the
`
`close of discovery, and should be discounted as untimely. Nor does Petitioner
`
`show that Thoen means the POSITA would interpret Pyle-Marusi’s absolute
`
`silence to inherently teach changing content based on hardware version. So this
`
`new argument is unsupported as well.
`
`---------
`
`
`
`The Pyle combination does not render any of the challenged claims obvious.
`
`V. LEWIS COMBINATION
`
`A. Lewis Does Not Teach The “Retrieving” Limitation
`
`Petitioner fails to show that “retrieving … a list of assets … wherein each
`
`asset is a different stream associated with the piece of content,” and Petitioner and
`
`Dr. Reader are inconsistent in their identification of the “assets” the “list of assets”
`
`allegedly “retriev[ed].” POR, 27-37. Petitioner’s reply is to again fail to identify
`
`what these “assets” supposedly are in Lewis.
`
`At deposition Dr. Reader contradicted his declaration, testifying that
`
`processed video segments 315a-c are not “assets” in Lewis’ alleged “list of assets”
`
`“retriev[ed].” Id., 29; See Ex. 2010, 106:1-12. Dr. Reader flips back and forth
`
`like a logic gate on, for example, whether segments 315a-c are “assets” in Lewis’s
`
`alleged “list of assets” “retriev[ed]”:
`
`
`
`- 16 -
`
`
`
`Case IPR2020-00648
`Patent 9,998,515
`Q. And your contention is that a list of assets is a list of processed video
`
`segments 315(a) to 315(c); is that correct?
`
`A. No, that's not correct.
`
`Q. What do you contend is the list of assets in the Ground II Combination?
`
`A. So the list of assets includes the stored video segments 375 and can or
`
`may also include processed video segments 315.
`
`Id., 112:6-14.
`
`Petitioner has no response to the fact that its expert’s testimony is
`
`inconsistent.
`
`Even the Reply does not offer clarification of what “assets” are in Lewis’s
`
`alleged “list of assets.” Reply, 18-19. Instead, Petitioner only argues that “a
`
`POSITA would have understood that stored video files 375 … are assets,” and that
`
`processed video segments 315a-c are also “assets,” without explaining which of
`
`these are allegedly “assets” in the “list.” Id. In fact, the Reply only adds to the
`
`confusion. It now argues that Lewis’ “entries 258a through 258f” in manifest file
`
`257 also comprise a “list of media assets.” Id., 20. Therefore, both Petitioner and
`
`Dr. Reader continue to avoid clearly identifying what “assets” are in Lewis’s
`
`alle