`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`MICROSOFT CORPORATION,
`Petitioner,
`
`v.
`
`WORLDS INC.,
`Patent Owner.
`
`Case IPR2021-00277
`Patent 8,082,501
`
`SUPPLEMENTAL DECLARATION OF MICHAEL ZYDA, D.SC.
`
`1
`
`MS 1034
`
`
`
`
`I, Michael Zyda, declare as follows:
`
`Supplemental Declaration of Michael Zyda
`re U.S. Patent No. 8,082,501
`
`1.
`
`I have been retained on behalf of Microsoft Corporation to offer tech-
`
`nical opinions relating to U.S. Patent No. 8,082,501 (the ’501 Patent), and prior art
`
`references relating to its subject matter.
`
`2.
`
`I was previously retained by Bungie Inc. to provide opinions with re-
`
`spect to the ’501 Patent in relation to IPR2015-01319. I submitted two declarations
`
`in that proceeding: a first dated June 1, 2015 (Ex. 1002) (“First Declaration”), and a
`
`second dated March 4, 2015 (Ex. 1038) (“Second Declaration”). I was also deposed
`
`in that proceeding, a transcript of which was submitted as Ex. 2016. I stand by the
`
`testimony I gave in that proceeding, and incorporate it herein. I provide the follow-
`
`ing testimony to explain additional opinions I have with respect to the ’501 Patent
`
`and related prior art.
`
`I.
`
`ADDITIONAL GROUNDS BASED ON COMBINATIONS OF FUNK-
`HOUSER AND DURWARD
`3.
`In my First Declaration, I presented and explained three grounds based
`
`on combinations including Funkhouser (Ex. 1005) as the primary reference. See Ex.
`
`1002, ¶¶ 68-157. Since I submitted that first declaration, I am aware that a district
`
`court issued an order in which it construed various terms of the ’501 Patent. Ex.
`
`1032. One of the claim terms construed in that order was “participant condition,”
`
`which is recited in claim elements [1.2], [12.4], and [14.2]. Ex. 1032, 14-18. For
`
`
`
`2
`
`
`
`
`example, independent claim 1 recites “receiving, by the client device, position in-
`
`Supplemental Declaration of Michael Zyda
`re U.S. Patent No. 8,082,501
`
`formation associated with fewer than all of the other user avatars in an interaction
`
`room of the virtual space, from a server process, wherein the client device does not
`
`receive position information of at least some avatars that fail to satisfy a participant
`
`condition imposed on avatars displayable on a client device display of the client
`
`device.” Ex. 1001, 19:27-33 (emphasis added).
`
`4.
`
`The district court construed the term “participant condition” to mean “a
`
`condition set by the client.” Ex. 1032, 18. In construing the term, the court ex-
`
`plained:
`
`“conditions” constitute additional limits and that in the ‘501 and ‘998
`patents: “(1) the client receives position information for less than all
`of the other users’ avatars, and (2) at least some, but not necessarily
`all, of the avatars for which the client does not receive position infor-
`mation are ones that failed to satisfy a ‘participant condition’ or ‘con-
`dition.’” D. 63 at 25-26. The “conditions” contemplated in the ‘501
`and ‘998 patents then must be distinct from the server conditions de-
`scribed in the specification and are properly construed to be consistent
`with the user or client conditions contemplated by the specification,
`including user ID and “other variables in addition to proximity.” ‘690,
`D. 62-2 at 10. And while the specification explicitly considers that
`there may be a wide range of variables that a client might set, nothing
`in the patent record suggests that the server will set these additional
`conditions.
`
`
`
`3
`
`
`
`
`Id. at 17-18.
`
`Supplemental Declaration of Michael Zyda
`re U.S. Patent No. 8,082,501
`
`5. My previous declarations did not explicitly address this construction.
`
`Having considered the district court’s construction, its relationship to claim elements
`
`[1.2], [12.4], and [14.2], and the three Funkhouser-based grounds I set forth in my
`
`First Declaration, it is my expert opinion that a POSITA would have found claim
`
`elements [1.2], [12.4], and [14.2]—and thus independent claims 1, 12, and 14 over-
`
`all—obvious in light of the teachings of Funkhouser, particularly when considered
`
`in light of the teachings of Durward (Ex. 1008).
`
`6.
`
`As I described in my First Declaration, Funkhouser describes “[s]erver-
`
`based message culling [that] is implemented using precomputed line-of-sight visi-
`
`bility information.” Ex. 1005 at 03. Funkhouser’s “RING servers can cull messages
`
`using high-level geometric algorithms and knowledge regarding a multiplicity of
`
`highly dynamic entity attributes (e.g., location, orientation, velocity, etc.) and inter-
`
`action types (e.g., visibility, sound, collision, etc.). Ex. 1005 at 03. In this regard,
`
`Funkhouser’s client devices clearly receive messages (e.g., messages including up-
`
`dated position information) associated with fewer than all of the other user avatars,
`
`and Funkhouser’s servers accomplish this by culling messages based (at least in part)
`
`on the proximity of the first user’s avatar to the other users’ avatars.
`
`7.
`
`However, the district court’s construction additionally requires that the
`
`claimed client device not receive position information of at least some avatars that
`4
`
`
`
`
`
`
`fail to satisfy “other variables in addition to proximity.” Ex. 1032, 18. Though not
`
`Supplemental Declaration of Michael Zyda
`re U.S. Patent No. 8,082,501
`
`a focus of my previous explanation of the Funkhouser grounds in my First Declara-
`
`tion, Funkhouser also describes culling messages based on variables in addition to
`
`proximity, as required by the district court’s construction. Specifically, Funkhouser
`
`describes that an “extension” to its system uses “multiresolution simulation to reduce
`
`network traffic and client behavioral simulation processing.” Ex. 1005 at 07. Under
`
`multiresolution simulation, “time critical computing algorithms can be used to de-
`
`termine an ‘optimal’ set of messages to send to each client based on network con-
`
`nection bandwidths, workstation processing capabilities, and many other real-time
`
`performance factors . . . .” Id. Thus, in addition to proximity, Funkhouser’s servers
`
`can use real-time performance factors such as network connection bandwidths and
`
`workstation processing capabilities to cull messages.
`
`8. While not explicitly described by Funkhouser, a POSITA would have
`
`found it obvious that the client device would have set at least some of these real-time
`
`performance factors. For example, the client device is in the best position to assess,
`
`set, and communicate to the server the “workstation processing capabilities” of the
`
`client. At the time of the ’501 patent, a POSITA would have known that relevant
`
`workstation processing capabilities would have included, for example, the client
`
`workstation’s clock rate, amount of available memory, network interface details,
`
`
`
`5
`
`
`
`
`and/or ping time1. Much of this information is specific to each individual client
`
`Supplemental Declaration of Michael Zyda
`re U.S. Patent No. 8,082,501
`
`workstation and would not be easily determinable without it being provided by the
`
`client workstation.
`
`9.
`
`In fact, there are only two entities in Funkhouser’s system that could
`
`have possibly provided the workstation processing capabilities and other real-time
`
`performance factors described by Funkhouser as being “used to determine an ‘opti-
`
`mal’ set of messages to send to each client”: the RING server or the client. See Ex.
`
`1005 at 07. There are no other relevant entities. See, e.g., Ex. 1005, FIG. 5 (showing
`
`a system diagram containing on client workstations and RING servers). Without
`
`receiving this information from the client, the RING servers would be forced to es-
`
`timate the “workstation processing capabilities,” which would have been inefficient
`
`
`1 Ping time is the amount of time is takes a request to travel from the client
`
`device to a server plus the time is takes the corresponding response to travel from
`
`the server to the client. While ping time can be calculated from either the client or
`
`the server, it was common at the time of the ’501 patent to have the client perform
`
`these calculations. This would avoid burdening the server and allow the client to
`
`couple the calculated ping time with other client-specific information (e.g., clock
`
`rate and available memory size) to determine a maximum packet size to communi-
`
`cate to the server.
`
`
`
`6
`
`
`
`
`and prone to error. Therefore, designing Funkhouser’s system to have the client
`
`Supplemental Declaration of Michael Zyda
`re U.S. Patent No. 8,082,501
`
`send data representing the workstation processing capabilities to the RING servers
`
`would have been an obvious and predictable design choice. Furthermore, there are
`
`numerous prior art references (including Durward, which I separately discussed in
`
`my First Declaration) that provide relevant teachings confirming that a client device
`
`would have communicated at least some variables in addition to proximity to the
`
`server that would have been used to cull update messages sent from the server to the
`
`client.
`
`10. Durward (Ex. 1008) describes a three-dimensional virtual world similar
`
`to Funkhouser’s. Specifically, Durward’s system includes “a central database for
`
`defining one or more three-dimensional virtual spaces,” with the server communi-
`
`cating data to users “so that the user’s computer may display a portion of a selected
`
`virtual space on the user’s head mounted display.” Ex. 1008, 1:52-59. To “reduce[]
`
`bandwidth requirements,” Durward’s system culls messages in a similar manner as
`
`Funkhouser’s system. Id. at 4:12-29. “After initial position, motion, control and
`
`sound data is communicated to the users, only changes in the position, motion, con-
`
`trol and sound data is communicated thereafter.” Id. at 4:23-26. Like in Funkhouser,
`
`“visual relevant spaces determine which state changes are communicated to (or per-
`
`ceivable by) the users.” Id. at 4:54-56. The “user’s visual relevant space may be
`
`defined by the field of view of the virtual being and areas in close proximity to it (as
`7
`
`
`
`
`
`
`with virtual being 184), in which case the visual relevant space may move about the
`
`Supplemental Declaration of Michael Zyda
`re U.S. Patent No. 8,082,501
`
`virtual space as the perspective or position of the virtual being changes.” Id. at 5:13-
`
`18.
`
`11. Durward further teaches that “[e]ach visual relevant space may be fur-
`
`ther subdivided into one or more visual priority spaces.” Id. at 5:23-24. These visual
`
`priority spaces within each visual relevant space are akin to the multiresolution sim-
`
`ulation taught by Funkhouser. Specifically, Durward’s “visual priority spaces may
`
`be used to determine the update frequency of elements located within them.” Id. at
`
`5:27-29. In other words, different objects in a visual relevant space can be rendered
`
`at a different resolution (e.g., different update frequency) depending on whether they
`
`are located in a visual priority space. For example, “data for updating objects or
`
`sounds in one priority space may be communicated thirty times per second, whereas
`
`data for updating objects or sounds in another priority space may be communicated
`
`once per second. Id. at 6:65-7:2. “This reduces the amount of data that must be
`
`communicated to each user while maintaining realism of important elements.” Id.
`
`at 5:35-37, 6:62-65.
`
`12. Durward describes that “[c]ontrol data received from the users may be
`
`used . . . to specify visual and sound relevant spaces and their corresponding priority
`
`spaces, etc.” Id. at 6:27-34. In other words, in implementing its multiresolution
`
`
`
`8
`
`
`
`
`message culling architecture, Durward describes that its system relies upon control
`
`Supplemental Declaration of Michael Zyda
`re U.S. Patent No. 8,082,501
`
`data set by the client device (i.e., “from the user”) to the server.
`
`13. Based on the teachings of Durward, a POSITA would have found it
`
`obvious and predictable to implement Funkhouser’s multiresolution simulation ex-
`
`tension so that at least some of the “real-time performance factors” used by Funk-
`
`houser’s server to cull messages are set by the user and his/her client device. For
`
`example, as noted above, it would have been obvious to a POSITA to implement
`
`Funkhouser’s system such that the client device would set the “workstation pro-
`
`cessing capabilities.” A POSITA would have been motivated to implement Funk-
`
`houser’s system in this way, because, e.g., it would have permitted the user the abil-
`
`ity to control the amount of update messages it receives from the server according
`
`to the capabilities of the user’s system. As the user upgrades his/her client work-
`
`station, these changes could have be communicated to Funkhouser’s server by the
`
`client workstation in order to provide increased update frequency—and thus per-
`
`ceived resolution. And, as I described above, only the client and the RING servers
`
`could have provided the workstation processing capabilities and other real-time per-
`
`formance factors, so implementing Funkhouser’s system such that the clients pro-
`
`vide their workstation processing capabilities to the RING servers would have been
`
`an obvious design choice.
`
`
`
`9
`
`
`
`Supplemental Declaration of Michael Zyda
`re U.S. Patent No. 8,082,501
`
`14. This obvious implementation of Funkhouser’s system would not alter
`
`
`
`the combinations of Funkhouser with Sitrick, Wexelblat, or Funkhouser ’93 that I
`
`described in my First Declaration. As I explained in my First Declaration, a POSITA
`
`would have relied upon Sitrick for its teachings related to “tools to allow a user to
`
`update the appearance of her avatar, and further discloses updating the geometry of
`
`avatars.” Ex. 1002, ¶¶ 75-77. A client providing one or more of the real-time per-
`
`formance factors to the RING servers, as I describe above, would not have impacted
`
`the combination of Funkhouser and Sitrick, as these real-time performance factors
`
`are unrelated to avatar customization.
`
`15. As I explained in my First Declaration, a POSITA would have relied
`
`upon Wexelblat for its teachings related to “an avatar be[ing] allowed to teleport
`
`between virtual rooms in the virtual world.” Ex. 1002, ¶¶ 134-141. A client provid-
`
`ing one or more of the real-time performance factors to the RING servers, as I de-
`
`scribe above, would not have impacted the combination of Funkhouser, Sitrick, and
`
`Wexelblat, as these real-time performance factors are unrelated to avatar teleporta-
`
`tion.
`
`16. As I explained in my First Declaration, a POSITA would have relied
`
`upon Funkhouser ’93 for its teachings related to “the client determining which ob-
`
`jects, including other user avatars, to display based not only on the orientation of the
`
`client avatar but also on a maximum number of avatars to be displayed based on the
`10
`
`
`
`
`
`
`performance capabilities of the client computer and desired frame rate of the dis-
`
`Supplemental Declaration of Michael Zyda
`re U.S. Patent No. 8,082,501
`
`played environment.” Ex. 1002, ¶¶ 142-157. A client providing one or more of the
`
`real-time performance factors to the RING servers, as I describe above, is related to
`
`the combination based on Funkhouser ’93 but is complimentary. Whereas the com-
`
`bination I described in my First Declaration with respect to Funkhouser ’93 relates
`
`to the client determining a maximum number of avatars to be displayed, the combi-
`
`nation I describe here with respect to Durward related to the server using workstation
`
`processing capabilities to limit the messages sent to the client. As I described in my
`
`First Declaration, it would have been obvious for both Funkhouser’s RING servers
`
`to use the workstation processing capabilities to limit the number of messages sent
`
`to the client and Funkhouser’s clients to further limit which objects, including other
`
`user avatars, to display. Ex. 1002, ¶¶ 147-152.
`
`17. Thus, in combinations with Durward, the Funkhouser grounds I set
`
`forth in my First Declaration with respect to the claims of the ’501 patent would
`
`remain the same except with respect to claim elements [1.2], [12.4], and [14.2]. With
`
`respect to claim element [1.2], [12.4], and [14.2], in addition to culling messages
`
`based on the position of the first user’s avatar, Funkhouser’s server would also cull
`
`messages based on workstation processing capabilities set by the user, as described
`
`above. In this updated combination that includes Durward, the first user’s client
`
`device would “not receive position information of at least some avatars that fail to
`11
`
`
`
`
`
`
`satisfy a participant condition [i.e., a condition based on client-set workstation pro-
`
`Supplemental Declaration of Michael Zyda
`re U.S. Patent No. 8,082,501
`
`cessing capabilities] imposed on avatars displayable on a client device display of the
`
`client device,” as recited in claims claim elements [1.2], [12.4], and [14.2].
`
`II. CONCLUSION
`I hereby declare that all statements made herein of my own knowledge
`
`18.
`
`are true and that all statements made on information and belief are believed to be
`
`true; and further that these statements were made with the knowledge that willful
`
`false statements and the like so made are punishable by fine or imprisonment, or
`
`both, under Section 1001 of Title 18 of the United States Code and that such willful
`
`false statements may jeopardize the validity of the application or any patents issued
`
`thereon.
`
`
`
`
`
`Date:
`
`
`
`
`
`
`
`
`
`
`
`
`
`Respectfully submitted,
`
`
`
`
`Michael Zyda
`
`
`
`
`
`12
`
`