`U.S. Patent No. 6,630,935
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`
`________________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`________________
`
`VOLKSWAGEN GROUP OF AMERICA, INC.,
`Petitioner,
`
`v.
`
`ADVANCED SILICON TECHNOLOGIES LLC
`Patent Owner
`________________
`
`Case IPR2016-00897
`U.S. Patent No. 6,630,935
`________________
`
`PATENT OWNER’S PRELIMINARY RESPONSE
`PURSUANT TO 37 C.F.R. § 42.107(a)
`
`
`
`TABLE OF CONTENTS
`
`Case IPR2016-00897
`U.S. Patent No. 6,630,935
`
`PAGE
`
`I.
`
`II.
`
`III.
`
`IV.
`
`Introduction & Summary of Arguments..........................................................1
`
`The ’935 Patent................................................................................................2
`
`A.
`
`B.
`
`C.
`
`D.
`
`E.
`
`The ’935 Patent Discloses An Inventive Apparatus And Method
`For The Parallel Processing Of Graphics Primitives ............................2
`
`Background of the ’935 Patent Technology..........................................2
`
`The ’935 Patent’s Improved Thread Processing ...................................5
`
`The Challenged Claims of the ’935 Patent .........................................10
`
`The ’935 Patent’s Relevant Litigation History ...................................13
`
`The Primary Asserted References .................................................................15
`
`The Correct Claim Construction of Material Disputed Terms......................17
`
`A.
`
`B.
`
`The Controlling Claim Construction Standard ...................................17
`
`The Material Claim Construction Issues Facing The Board...............18
`
`1.
`
`2.
`
`“arbitration module [that] utilizes an application specific
`prioritization scheme”...............................................................19
`
`“in an order to minimize idle time of the computation
`engine” ......................................................................................22
`
`V.
`
`Volkswagen Has Failed To Prove That the Challenged Claims Are
`Obvious In Light Of Hirata In Combination With Any Other
`Reference .......................................................................................................25
`
`A.
`
`Volkswagen Fails to Construe or Interpret Any of the Claim
`Terms, and Fails to Address the Constructions Provided at the
`ITC.......................................................................................................25
`
`ii
`
`
`
`Case IPR2016-00897
`U.S. Patent No. 6,630,935
`
`B.
`
`Volkswagen Has Failed To Prove That Hirata Discloses the
`“arbitration module [that] utilizes an application specific
`prioritization scheme” Limitation In Claims 1-8 ................................27
`
`C.
`
`Volkswagen Has Failed To Prove That Claims 9-18 Are Invalid ......30
`
`1.
`
`2.
`
`Volkswagen’s Argument That Claims 9-18 Are Obvious
`Is Legally Deficient Because It Fails to Demonstrate
`Why A POSA Would Combine The Five References
`Identified In Its Petition ............................................................31
`
`Volkswagen Has Failed To Prove That Hirata Discloses
`The Claimed Arbitration Module..............................................33
`
`VI. Conclusion .....................................................................................................37
`
`iii
`
`
`
`TABLE OF AUTHORITIES
`
`Case IPR2016-00897
`U.S. Patent No. 6,630,935
`
`Page(s)
`
`Cases
`Aventis Pharma S.A. v. Hospira, Inc.,
`675 F.3d 1324 (Fed. Cir. 2012) ..........................................................................18
`
`Becton, Dickinson and Co. v. One StockDuq Holdings, LLC,
`IPR2013-00235, Paper 30 (PTAB Sept. 25, 2014).............................................18
`
`Certain Computing or Graphics Systems, Components Thereof, and
`Vehicles Containing Same,
`Investigation No. 337-TA-984..........................................................13, 14, 22, 24
`
`Cuozzo Speed Techs., LLC v. Lee,
`195 L. Ed. 2d 423 (U.S. 2016)............................................................................26
`
`Ericcson, Inc. v. Intellectual Ventures I LLC,
`IPR2014-00921, Paper 8 (PTAB Dec. 16, 2014) ...............................................18
`
`Hill-Rom Services, Inc. v. Stryker Corporation,
`755 F.3d 1367 (Fed. Cir. 2014) ....................................................................17, 18
`
`Intellectual Ventures Mgmt, LLC, v. Xilinx, Inc.,
`IPR2012-00019, Paper 33 (PTAB February 10, 2014) ......................................17
`
`Jiawei Tech. (HK) Ltd. v. Richmond.,
`IPR2014-00938, Paper 20 (PTAB Dec. 16, 2015) .................................25, 26, 35
`
`KSR Int’l Co. v. Teleflex Inc.,
`550 U.S. 398 (2007)......................................................................................32, 33
`
`Phillips v. AWH Corp.,
`415 F.3d 1303 (Fed. Cir. 2005) (en banc) ..........................................................17
`
`Purdue Pharma L.P. v. Depomed, Inc.,
`2016 U.S. App. LEXIS 5475 (Fed. Cir. Mar. 24, 2016) ....................................33
`
`RPX Corp. v. MD Security Solutions, LLC,
`IPR2016-00285, Paper 9 (PTAB Jun. 6, 2016) ..................................................18
`
`iv
`
`
`
`Case IPR2016-00897
`U.S. Patent No. 6,630,935
`
`Space Exploration Technologies Corp. v. Blue Origin LLC,
`IPR2014-01378, Paper 6 (PTAB Mar. 3, 2015).................................................26
`
`Universal Remote Control, Inc. v. Universal Electronics, Inc.,
`IPR2013-00127, Paper 32 (PTAB June 30, 2014) .............................................17
`
`Wowza Media Sys., LLC v. Adobe Systems Inc.,
`IPR2013-00054, No. 12 (PTAB Apr. 8, 2013)...................................................17
`
`Other Authorities
`
`77 Fed. Reg. 48,756, 48,766 (Aug. 14, 2012) .........................................................17
`
`U.S. Patent No. 6,512,524..........................................................................................5
`
`U.S. Patent No. 6,630,935.................................................................................passim
`
`v
`
`
`
`Case IPR2016-00897
`U.S. Patent No. 6,630,935
`
`TABLE OF EXHIBITS
`
`Exhibit
`
`Description
`
`2001
`
`2002
`
`2003
`
`2004
`
`2005
`
`Declaration of John Hart
`
`U.S. Patent No. 6,412,524
`
`AST’s Preliminary Proposed Claim Constructions, Inv. No. 337-
`TA-984, served on April 8, 2016.
`
`Respondents Preliminary Proposed Claim Constructions, Inv. No.
`337-TA-984, served on April 8, 2016.
`
`Appendix A: Compendium of Extrinsic Evidence Regarding
`“Throughput”
`
`2006
`
`Order No. 42: Construing Terms Of The Asserted Patents
`
`vi
`
`
`
`Case IPR2016-00897
`U.S. Patent No. 6,630,935
`
`I.
`
`Introduction & Summary of Arguments
`
`For at least the following reasons, Volkswagen’s Petition fails to establish
`
`the required likelihood that Volkswagen will prove that the challenged claims of
`
`Advanced Silicon Technologies LLC’s (“AST”) U.S. Patent No. 6,630,935 (the
`
`“’935 Patent”) are unpatentable:
`
` Volkswagen has failed to provide a construction or interpretation of the key
`
`claim terms “arbitration module” and “in an order to minimize idle time of
`
`the computation engine;”
`
` The claims require an arbitration module that uses an application specific
`
`prioritization scheme, but Volkswagen the alleged arbitration module
`
`identified by Volkswagen does not use the alleged arbitration scheme
`
`identified by Volkswagen; and
`
` Volkswagen has failed to demonstrate why a person of ordinary skill in the
`
`art would combine the five references that it uses to challenge claims 9-18.
`
`Accordingly, Volkswagen has not met the threshold requirement of section
`
`314(a), as it has not demonstrated that there is reasonable likelihood that it will
`
`prevail with respect to at least 1 of the claims challenged in the petition. Therefore,
`
`the Patent Trial and Appeal Board (the “Board”) should deny in its entirety
`
`Volkswagen’s Petition requesting inter partes review of AST’s ’935 patent.
`
`1
`
`
`
`Case IPR2016-00897
`U.S. Patent No. 6,630,935
`
`II.
`
`The ’935 Patent
`
`A.
`
`The ’935 Patent Discloses An Inventive Apparatus And Method
`For The Parallel Processing Of Graphics Primitives
`
`The ’935 patent is entitled “Geometric Engine Including a Computational
`
`Module for use in a Video Graphics Controller.” It improved “video graphics
`
`processing,” and, specifically, “the processing of geometric primitives.” ’935
`
`patent at 1:7-9 (attached to the Petition as Exh. 1001); see also Exh. 2001
`
`(Declaration of Dr. John C. Hart) at ¶ 16. The improvement involves a plurality of
`
`thread controllers, and an arbitration module that utilizes a prioritization scheme to
`
`minimize the idle time of a computation engine. Exh. 1001 at Abstract.
`
`B.
`
`Background of the ’935 Patent Technology
`
`Prior to the ’935 patent, a “conventional computing system” used a “central
`
`processing unit … to produce geometric primitives” for “the raster engine of the
`
`video graphics circuit of the video graphics circuit [to generate] pixel data based on
`
`the attributes … of the primitive.” Exh. 1001 at 1:13-38. The ’935 patent satisfied
`
`“a need … for a method and apparatus that provides parallel processing of graphics
`
`primitives … such that a hardware geometry engine may be practically
`
`implemented.” Id. at 1:45-48; see also Exh. 2001 at ¶ 16.
`
`A “geometry engine” of the ’935 patent first transforms the spatial
`
`coordinates of the vertices from their input model-oriented coordinates into screen-
`
`oriented coordinates appropriate for clipping or, for example, transforming for a
`2
`
`
`
`Case IPR2016-00897
`U.S. Patent No. 6,630,935
`
`given viewpoint. Exh. 2001 at ¶ 17. The geometry engine then clips the triangles to
`
`remove triangles outside the view or the portion of triangles only partially viewed.
`
`Id. For triangles at least partially viewed, the original vertices are then shaded with
`
`lighting and other per-vertex processes. Id. Finally a “barycentric interpolation”1
`
`determines the in-between positions and attributes of any new vertices produced by
`
`clipping when a partially-viewed triangle is subdivided into its viewable portion.
`
`Id.; Exh. 1001 at 2:60-3:10.
`
`This pipelined graphics process is depicted in the following diagram:
`
`1 Generally speaking, interpolation calculates a weighted average value for a data
`
`point based on a surrounding range of known data points. Barycentric coordinates
`
`specify the location of the point of a simplex (e.g., a triangle) to be calculated as
`
`the center of mass of the simplex.
`
`3
`
`
`
`Case IPR2016-00897
`U.S. Patent No. 6,630,935
`
`Exh. 2001 at ¶ 17.
`
`The pipelined graphics process continues by sending the transformed,
`
`clipped, shaded and interpolated triangle vertices to the raster engine, which scan-
`
`converts the triangle from its three vertices into the pixels used to fill it and also
`
`determines the pixels’ colors. Id. at ¶ 18. The raster engine then writes the resulting
`
`pixel data into a frame buffer memory, where it can be retrieved as needed for
`
`display. Id. The inventors of the ’935 patent noted that “[p]erforming all of the
`
`primitive processing operations in software consume[d] a large amount of
`
`processing bandwidth that can limit the overall processing speed of the computing
`
`system in which the graphics processing is occurring.” Exh. 1001 at
`
`1:39-42; Exh. 2001 at ¶ 18.
`
`4
`
`
`
`Case IPR2016-00897
`U.S. Patent No. 6,630,935
`
`C.
`
`The ’935 Patent’s Improved Thread Processing
`
`In order to overcome the problems associated with prior art systems that
`
`executed graphics processing software on a general purpose CPU, the ’935 patent
`
`“provides parallel processing of graphics primitives with limited memory
`
`requirements such that a hardware geometry engine may be practically
`
`implemented.” Exh. 1001 at 1:46-48; Exh. 2001 at ¶ 19. The ’935 patent discloses
`
`embodiments of the specialized graphics computation engine on the “geometric
`
`engine of a video graphics circuit” that includes vertex transformations, triangle
`
`clipping, vertex lighting and barycentric interpolation. See Exh. 2001 at ¶ 19. For
`
`example, the figure below demonstrates the clipping process, which determines the
`
`viewable portion of a triangle (in this example the quadrilateral portion to the left
`
`of the dashed line representing the right edge of the display screen). Id. In this
`
`case, clipping removes vertex V2, circled in red, and introduces two new vertices:
`
`C0 and C1, circled in blue. See Exh. 2002 (U.S. Patent No. 6,512,524, filed by a
`
`co-inventor of the ’935 patent) at Figure 4 (annotations added); Exh. 2001 at ¶ 19.
`
`5
`
`
`
`Case IPR2016-00897
`U.S. Patent No. 6,630,935
`
`Generally speaking, an application is simply a program. However, in the
`
`’935 patent, the inventors use the term “application” very specifically to refer to a
`
`plurality of threads that implement a parallel processing video graphics task, such
`
`as, for example, processing graphics primitives. See Exh. 2001 at ¶ 20. The ’935
`
`patent discloses that “the plurality of threads constitutes an application.” Exh.
`
`1001 at 2:51-52 (emphasis added). Dividing a video graphics task into multiple
`
`parallel threads, or “decomposition,” provides additional flexibility when
`
`executing an application program by improving “throughput,” or efficiency, as
`
`commonly measured by the number of instructions executed in a given time
`
`interval. See Exh. 2001 at ¶ 20.
`
`The execution of each thread is controlled by its “thread controller,” which
`
`sends the next operation code (instruction) of its thread to a processor for
`
`execution, but only when the computation module is ready to execute it. Id. at ¶ 21.
`6
`
`
`
`Case IPR2016-00897
`U.S. Patent No. 6,630,935
`
`This allows flexibility at the time of execution, to choose between several
`
`concurrent threads in an application program, and thus improves throughput, for
`
`example by hiding latency. Id. For example, if one thread is going to process data
`
`but the needed data is not yet available, then that thread’s controller can prevent
`
`execution of its operation codes until the data becomes available, allowing other
`
`threads in the application to be executed instead of stalling execution of the entire
`
`application while waiting for the data to become available. See id.
`
`In particular, “[t]he thread controllers . . . each only release operation codes
`
`. . . when the operation codes can be executed without any potential for delay in
`
`waiting for the results of previously issued operation codes.” Exh. 1001 at 3:34-37.
`
`For example, when an operation code is dependent on the results of a previously-
`
`issued operation code, the thread controller may not release the dependent
`
`operation code until a certain amount of time has passed corresponding to the
`
`latency associated with executing the operation code that produces the data
`
`required by the dependent operation code. Id. at 3:34-43.
`
`The ’935 patent demonstrates the use of thread processing for threads in the
`
`computation engine circuitry that are specialized for processing parallel graphics
`
`tasks. See Exh. 2001 at ¶ 22. For example, the ’935 patent discloses that “[t]he
`
`application . . . may be an application corresponding to processing geometric
`
`primitives for use in a video graphics circuit.” Id. at 2:67-3:3. Such an application
`7
`
`
`
`Case IPR2016-00897
`U.S. Patent No. 6,630,935
`
`may include threads for determining the vertices and associated attributes of
`
`primitives to be rendered, performing transform operations on the vertices,
`
`performing clipping operations on the primitives, determining lighting effects, and
`
`determining texture coordinate values.” See id.; Exh. 1001 at 2:67-3:8.
`
`All independent claims of the ’935 patent include an arbitration module. For
`
`example, claim 1 includes an arbitration module that determines the next operation
`
`code, utilizing an “application specific prioritization scheme.” Exh. 1001 at claim
`
`1. The specification demonstrates this by using the application of a geometry
`
`engine consisting of three threads for vertex transformation; one thread for
`
`clipping; multiple threads for vertex shading, or “attribute processing” (“AP”); and
`
`a thread for barycentric interpolation. See Exh. 2001 at ¶ 23. Each of these threads
`
`has its own thread controller, except that groups of AP threads can share the same
`
`thread controller. See id. Figure 4 of the ‘935 patent is reproduced below, and
`
`depicts these thread controllers as items 115, 117, 119, 121, 127, 129, 131, and
`
`133:
`
`8
`
`
`
`Case IPR2016-00897
`U.S. Patent No. 6,630,935
`
`As disclosed in the Abstract, “[t]he arbitration module … utilizes an
`
`application specific prioritization scheme to provide operation codes from the
`
`plurality of thread controllers to the computation engine such that idle time of the
`
`computation engine is minimized.” Exh. 1001 at Abstract. The “prioritization
`
`scheme orders operation codes by prioritizing certain threads over other threads
`
`such that the throughput through the computation module is maximized.” Id. at
`
`2:52-59. Hence throughput (efficiency) is improved by reducing the idle time of
`
`the computation engine. See Exh. 2001 at ¶ 24.
`
`Figure 7 of the ’935 patent demonstrates thread prioritization for a geometry
`
`engine application, simplified to four threads: two transformation threads, one
`
`9
`
`
`
`Case IPR2016-00897
`U.S. Patent No. 6,630,935
`
`clipping thread and one “post-clip” thread that would represent attribute processing
`
`and interpolation. See Hart Decl. at ¶ 25. For example, the application specific
`
`prioritization scheme can prioritize threads “back-to-front,” meaning that threads
`
`toward the end of the pipeline are prioritized higher than threads toward the
`
`beginning of the pipeline. See id.; Exh. 1001 at 31:24-32:43. In this example, “the
`
`application specific prioritization scheme may prioritize operations in a back-to-
`
`front manner that ensures that processing that is nearing completion is prioritized
`
`over processing that is just beginning.” Exh. 1001 at 3:63-66; Exh. 2001 at ¶ 25.
`
`The inventors of the ’935 patent explained that “prioritizing the final steps to
`
`produce results passed to downstream circuitry may help to ensure that the
`
`resources in the pipeline of the computation engine . . . are efficiently utilized and
`
`a regular production rate of results can be maintained.” Id. at 3:63-4:3. Such a
`
`prioritization scheme, and other prioritization schemes disclosed in the ’935 patent,
`
`causes the display to be refreshed with the most up-to-date information available,
`
`such as the current visuals needed for rendering navigation graphics. See Exh.
`
`2001 at ¶ 26.
`
`D.
`
`The Challenged Claims of the ’935 Patent
`
`The Petition challenges all claims of the ’935 patent (i.e., claims 1-18). This
`
`includes independent claim 1 and its dependent claims 2-8, all of which require an
`
`10
`
`
`
`Case IPR2016-00897
`U.S. Patent No. 6,630,935
`
`arbitration module that utilizes an application specific prioritization scheme.
`
`Independent claim 1 is set forth in its entirety below:
`
`1. A computation module comprises:
`memory;
`a computation engine operable to perform an operation
`based on an operation code and to provide a corresponding result
`to the memory as indicated by the operation code;
`a plurality of thread controllers, wherein each of the
`plurality of thread controllers manages at least one corresponding
`thread of a plurality of threads, wherein the plurality of threads
`constitutes an application, and wherein each of the plurality of
`threads includes at least one operation code; and
`an arbitration module operably coupled to the plurality of
`thread controllers, wherein the arbitration module utilizes an
`application specific prioritization scheme to provide operation
`codes from the plurality of thread controllers to the computation
`engine in an order to minimize idle time of the computation engine.
`
`The Petition also challenges independent claim 9 and its dependent claims
`
`10-18. Dependent claim 9 has significantly more claim elements than independent
`
`claim 1, but also requires an arbitration module that selects an operation code from
`
`pending operation codes based upon a prioritization scheme. Independent claim 9
`
`is set forth in its entirety below:
`
`11
`
`
`
`Case IPR2016-00897
`U.S. Patent No. 6,630,935
`
`9. A geometric engine for use in a video graphics controller,
`the geometric engine comprises:
`a vector engine operably coupled to perform an operation
`upon data as indicated by an operation code, wherein the
`operation code includes at least one source address, a destination
`address, and an operation type that indicates the operation to be
`performed;
`a transform controller operably coupled to manage a
`transform thread, wherein the transform thread includes a set of
`transform operation codes, wherein when the transform operation
`codes are executed, the vertex data corresponding to an initial set
`of vertices of a primitive are rotated from an initial coordinate
`system into a different coordinate system to produce rotated vertex
`data for the primitive;
`a clip controller operably coupled to manage a clip thread,
`wherein the clip thread includes a set of clip operation codes,
`wherein when the clip operation codes are executed, a
`determination is made as to whether at least a portion of the
`primitive lies within a clip volume based on the rotated vertex
`data, wherein when at least a portion of the primitive lies within
`the clip volume and at least a portion of the rotated vertex data
`corresponding to at least one vertex of the primitive lies outside of
`the clip volume, new vertex data is calculated for the at least one
`vertex;
`a plurality of attribute controllers, wherein each attribute
`controller of the plurality of attribute controllers is operably
`
`12
`
`
`
`Case IPR2016-00897
`U.S. Patent No. 6,630,935
`
`coupled to manage a corresponding attribute thread, wherein each
`of the corresponding attribute threads includes a set of attribute
`operation codes, wherein when executed, each set of attribute
`operation codes determines an attribute for the vertex data
`corresponding to the initial set of vertices;
`a barycentric controller operably coupled to manage a
`barycentric thread, wherein the barycentric thread includes a set
`of barycentric operation codes, wherein, when executed, the
`barycentric operation codes combine the attributes produced by
`the plurality of attribute controllers with the new vertex data to
`produce adjusted attributes for the primitive; and
`an arbitration module operably coupled to provide the
`operation code to the vector engine, wherein the arbitration
`module selects the operation code from pending operation codes
`issued for execution by the transform controller, the clip
`controller, the plurality of attribute controllers, and the
`barycentric controller, wherein the arbitration module selects the
`operation code from the pending operation codes based on a
`prioritization scheme.
`The ’935 Patent’s Relevant Litigation History
`
`E.
`
`AST is currently asserting the ’935 patent against Volkswagen at the
`
`International Trade Commission in Certain Computing or Graphics Systems,
`
`Components Thereof, and Vehicles Containing Same, Investigation No. 337-TA-
`
`984 (the “ITC Action”). On April 8, 2016, AST and the Respondents exchanged
`
`their preliminary proposed claim constructions for the patents at issue in the ITC
`13
`
`
`
`Case IPR2016-00897
`U.S. Patent No. 6,630,935
`
`Action, including the ’935 patent. See Exh. 2003 (AST’s Preliminary Proposed
`
`Claim Constructions) at 8-11; Exh. 2004 (Respondents’ Preliminary Proposed
`
`Claim Constructions) at 11-17. Although the Petition was filed on April 15,
`
`2015—after the parties exchanged claim constructions in the ITC Action,
`
`Volkswagen did not disclose, identify, or address either party’s proposed claim
`
`constructions in its Petition.
`
`As discussed in more detail below, some of Volkswagen’s claim
`
`construction positions in the Petition conflict with the claim constructions it
`
`proposed in the ITC Action. For example, in the ITC Action Volkswagen
`
`contended that “arbitration module” and “in an order to minimize idle time of the
`
`computation engine” were indefinite. But, because indefinite claims cannot be
`
`anticipated or obvious, and indefiniteness cannot be determined in an inter partes
`
`review, Volkswagen’s indefiniteness allegations are fatal to its Petition here.
`
`Volkswagen Petition failed to disclose its indefiniteness position to the Board, but
`
`still offered no meaning or construction for those terms. Compare Exh. 2004 at 15,
`
`17 with Petition at 15-16.
`
`Moreover, on July 15, 2016, the ITC rejected Volkswagen’s position that
`
`these claim terms are indefinite, and instead construed the terms consistent with
`
`AST’s positions. See Exh. 2006 (Order No. 42: Construing Terms Of The Asserted
`
`14
`
`
`
`Case IPR2016-00897
`U.S. Patent No. 6,630,935
`
`Patents). Volkswagen’s Petition does not take into account AST, and now the
`
`ITC’s claim constructions, which rejects Volkswagen’s position.
`
`III. The Primary Asserted References
`
`Volkswagen’s Petition asserts the following Grounds:
`
`Ground
`
`References
`
`Hirata and Kimura
`
`Hirata, Kimura, and
`Hennessy
`
`Hirata, Kimura, and
`Watkins
`
`Kimura, Hirata, and
`Rentschler
`
`1
`
`2
`
`3
`
`4
`
`5
`
`Basis
`
`§ 103
`
`§ 103
`
`Challenged Claims
`
`1, 2, and 7
`
`3
`
`§ 103
`
`4 and 5
`
`§ 103
`
`6 and 8
`
`Hirata, Kimura, Hennessy,
`Watkins, and Rentschler
`
`§ 103
`
`9-18
`
`Each of the Petition’s Grounds 1-4 (against claims 1-8) rely upon Hirata in
`
`combination with one or two other references. Ground 5 (against claims 9-18),
`
`relies on a purported combination of five different references.
`
`Hirata proposes a multithreaded processor architecture that “greatly
`
`improves machine throughput.” Hirata at 136 (attached as Exh. 1003 to the
`
`Petition). The proposed architecture can be used “as the base processor in a parallel
`
`machine which could run a graphics system.” Id.
`
`15
`
`
`
`Case IPR2016-00897
`U.S. Patent No. 6,630,935
`
`The focus of the Hirata paper is on a technique called parallel multithreading
`
`where an independent instruction from one thread can be issued when an
`
`instruction from another thread cannot be issued due to a control or data
`
`dependency. Id. This technique arose from “hardware resource optimization where
`
`several processors are united and reorganized so that functional units could be fully
`
`used.” Id. at 136-17.
`
`Hirata discloses that a compiler uses one algorithm to schedule code for
`
`computer graphics programs and another algorithm to schedule code for numerical
`
`computation programs. Id. at 141. The compiler uses a simple algorithm for
`
`computer graphics programs, while it employs “more judicious code schedule
`
`techniques” for the more-predictable numerical computation programs. Id. Hirata
`
`describes both of these code scheduling techniques under a subheading entitled
`
`“Static Code Scheduling” because the scheduling is not dynamic, as the compiler
`
`schedules the instructions prior to execution. Id.
`
`Hirata also discloses an instruction schedule unit that, during execution of
`
`applications, provides code for execution according to the static priority set by the
`
`compiler. The scheduling unit cannot change the priority of the code and does not
`
`use the scheduling algorithms. Rather, the instruction schedule unit blindly follows
`
`the order and priorities of the instructions as they were previously and statically set
`
`by the compiler. Id.
`
`16
`
`
`
`Case IPR2016-00897
`U.S. Patent No. 6,630,935
`
`IV. The Correct Claim Construction of Material Disputed Terms
`
`The Controlling Claim Construction Standard
`A.
`The broadest reasonable interpretation standard applies in an inter partes
`
`review of a patent that, like the ’935 patent, will not expire prior to the Final
`
`Written Decision. 37 C.F.R. §100(b); Office Patent Trial Practice Guide, 77 Fed.
`
`Reg. 48,756, 48,766 (Aug. 14, 2012). Applying the broadest reasonable
`
`interpretation, claim terms are given their ordinary and customary meaning, as
`
`would be understood by a person of ordinary skill in the art at the time of the
`
`invention, in light of the language of the claims, the specification, and the
`
`prosecution history of the record. E.g., Wowza Media Sys., LLC v. Adobe Systems
`
`Inc., IPR2013-00054, No. 12 at 5 (PTAB Apr. 8, 2013); Intellectual Ventures
`
`Mgmt, LLC, v. Xilinx, Inc., IPR2012-00019, Paper 33 at 9 (PTAB February 10,
`
`2014); see also Phillips v. AWH Corp., 415 F.3d 1303, 1313-1317 (Fed. Cir. 2005)
`
`(en banc); Hill-Rom Services, Inc. v. Stryker Corporation, 755 F.3d 1367, 1371
`
`(Fed. Cir. 2014).
`
`Under this test, “[t]here is a “heavy presumption’ that a claim term carries its
`
`ordinary and customary meaning.” See Intellectual Ventures, IPR2012-00019,
`
`Paper 33 at 9; Wowza, IPR2013-00054, No. 12 at 6; Universal Remote Control,
`
`Inc. v. Universal Electronics, Inc., IPR2013-00127, Paper 32 at 6 (PTAB June 30,
`
`2014). This heavy presumption is overcome in specific and limited circumstances:
`
`17
`
`
`
`Case IPR2016-00897
`U.S. Patent No. 6,630,935
`
`a claim term may be construed contrary to its ordinary meaning only where there is
`
`clear and unambiguous evidence that the patentee, as lexicographer, provided a
`
`special definition for the claim term, or the patentee otherwise disavowed the full
`
`scope of the claim term either in the specification or during prosecution. Ericcson,
`
`Inc. v. Intellectual Ventures I LLC, IPR2014-00921, Paper 8, at 8 (PTAB Dec. 16,
`
`2014); Becton, Dickinson and Co. v. One StockDuq Holdings, LLC, IPR2013-
`
`00235, Paper 30 at 6 (PTAB Sept. 25, 2014); see also Aventis Pharma S.A. v.
`
`Hospira, Inc., 675 F.3d 1324, 1330 (Fed. Cir. 2012); Hill-Rom, 755 F.3d at 1371.
`
`B.
`
`The Material Claim Construction Issues Facing The Board
`
`The Board should only construe terms to the extent such construction is
`
`necessary to resolve a controversy material to the Petition. See RPX Corp. v. MD
`
`Security Solutions, LLC, IPR2016-00285, Paper 9 at 6 (PTAB Jun. 6, 2016) (citing
`
`Vivid Techs., Inc. v. Am. Sci. & Eng’g, Inc., 200 F.3d 795, 803 (Fed. Cir. 1999)).
`
`Based on the Petition and this Preliminary Response, the only material claim
`
`construction issues currently facing the Board concern the terms “arbitration
`
`module [that] utilizes an application specific prioritization scheme” and “in an
`
`order to minimize idle time of the computation engine.”
`
`As discussed above, claim terms are construed as they would be understood
`
`by a person of ordinary skill in the art. AST proposes that a person of ordinary skill
`
`in the art would have a degree in electrical engineering, computer engineering,
`
`18
`
`
`
`Case IPR2016-00897
`U.S. Patent No. 6,630,935
`
`computer science, or a related field, and at least three years’ experience working in
`
`computer graphics hardware, computer architecture, or related fields, or an
`
`equivalent combination of graduate education and/or work experience. See Exh.
`
`2001 at ¶ 14. In contrast, Volkswagen’s Petition states that a person of ordinary
`
`skill in the art would have the same or similar degree but with only two to four
`
`years’ of industry experience. See Petition at 16. The ITC held that a person of
`
`ordinary skill in the art would have “a degree in electrical engineering, computer
`
`engineering, computer science, or a related field, and at least three to five years’
`
`experience working in computer graphics hardware, computer architecture, or
`
`related fields, or an equivalent combination of graduate education and/or work
`
`experience.” See Exh. 2006 at ¶ 13-14. For purposes of this Preliminary
`
`Responses, the competing definitions of a person of ordinary skill in the art are
`
`identical.
`
`1.
`
`“arbitration module [that] utilizes an application specific
`prioritization scheme”
`
`Consistent with its position in the ITC Action, AST proposes to construe an
`
`“arbitration module [that] utilizes an application specific prioritization scheme”
`
`using its plain and ordinary meaning. The plain and ordinary meaning of this term
`
`is an “arbitration module that orders operation codes dependent on categories of
`
`threads included in an application, i.e., that prioritizes the operation codes based on
`
`19
`
`
`
`Case IPR2016-00897
`U.S. Patent No. 6,630,935
`
`the type/category of the application/threads.” See Exh. 2001 at ¶ 28.This
`
`understanding is based on how one of ordinary skill in the art would understand
`
`this term in light of the intrinsic record. See id. As discussed below, the term
`
`“arbitration module [that] utilizes an application specific prioritization scheme” has
`
`a plain and ordinary meaning to a person of ordinary skill in the art upon review of
`
`the specification. Volkswagen offers no analysis, argument, or expert testimony to
`
`rebut this.
`
`Claim 1 requires:
`
`an arbitration module operably coupled to the plurality of thread
`
`controllers, wherein the arbitration module utilizes an application specific
`
`prioritization scheme to provide operation codes from the plurality of thread
`
`controllers to the computation engine in an order to minimize idle time of the
`
`computation engine.
`
`(Emphasis added).
`
`First, read properly in light of the ’935 patent sp