IPR2016-01106
`U.S. Patent No. 6,516,442
`
`Filed on behalf of EMC Corporation
`Docket No.: 0100157.00263US1
`
`By: Peter M. Dichiara, Reg. No. 38,005
`Theodoros Konstantakopoulos, Reg. No. 74,155
`WILMER CUTLER PICKERING HALE AND DORR LLP
`peter.dichiara@wilmerhale.com
`theodoros.konstantakopoulos@wilmerhale.com
`Tel.: 617-526-6466
`Fax: 617-526-5000
`
`
`
`
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`
`___________________________________________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`____________________________________________
`
`EMC Corporation
`Petitioner
`
`v.
`
`Intellectual Ventures II, LLC
`Patent Owner
`
`Case IPR2016-01106
`U.S. Patent No. 6,516,442
`
`PETITIONER’S REPLY
`
`
`
`
`
`
`U.S. Patent No. 6,516,442
`
`Filed on behalf of EMC Corporation
`Docket No.: 0100157.00263US1
`
`By: Peter M. Dichiara, Reg. No. 38,005
`Theodoros Konstantakopoulos, Reg. No. 74,155
`WILMER CUTLER PICKERING HALE AND DORR LLP
`peter.dichiara@wilmerhale.com
`theodoros.konstantakopoulos@wilmerhale.com
`Tel.: 617-526-6466
`Fax: 617-526-5000
`
`
`
`
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`
`___________________________________________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`____________________________________________
`
`EMC Corporation
`Petitioner
`
`v.
`
`Intellectual Ventures II, LLC
`Patent Owner
`
`Case IPR2016-01106
`U.S. Patent No. 6,516,442
`
`PETITIONER’S REPLY
`
`
`
`
`
`
`
`IPR2016-01106
`U.S. Patent No. 6,516,442
`
`I.
`II.
`
`TABLE OF CONTENTS
`Introduction ...................................................................................................... 1
`Claim Construction .......................................................................................... 1
`A.
`“Switch Fabric” ..................................................................................... 1
`B.
`“Packets Containing Data” .................................................................... 9
`III. Grounds .......................................................................................................... 15
`A. Ground 2: Claims 1, 12, 24 And 34 Are Obvious Over Reschke ...... 15
`B.
`Ground 3: It Would Be Obvious To Combine Reschke And Nishtala
` ............................................................................................................. 28
`Ground 3: Claim 5 (Synchronous Reset) Is Obvious ......................... 29
`
`
`
`C.
`
`i
`
`
`
`IPR2016-01106
`U.S. Patent No. 6,516,442
`
`I.
`
`INTRODUCTION
`Patent Owner’s Response (“POR”) confirms that the challenged claims of
`
`the ʼ442 patent are invalid. Indeed, Patent Owner largely does not (and cannot)
`
`dispute that nearly all claim elements are taught by the prior art under the Board’s
`
`claim constructions. Consequently, the Patent Owner tries to avoid the prior art
`
`through new but unsupportable claim constructions. However, even under these
`
`unsupportable constructions, the claims are still unpatentable.
`
`In the few instances where the Patent Owner Response attempts to rebut the
`
`Board’s preliminary findings, cross-examination of Patent Owner’s expert, Dr.
`
`Alpert, confirmed, among other things, that (1) Reschke discloses full-duplex
`
`channels under the correct and proper understanding of the term; (2) it would be
`
`obvious to modify processor and memory interfaces to include error correction;
`
`and (3) Reschke would enjoy a benefit taught by Nishtala confirming the
`
`underlying rationale of their obviousness combination.
`
`II. CLAIM CONSTRUCTION
`A.
` “Switch Fabric”
`The Board construed “switch fabric” to mean “a data switching circuitry
`
`having a matrix or similar arrangement of interconnections.” Decision on
`
`Institution (“DI”), 12. In so doing, the Board correctly noted that a “switch fabric”
`
`is not limited to just “interconnected switches or switch circuits” but instead covers
`
`1
`
`
`
`IPR2016-01106
`U.S. Patent No. 6,516,442
`“a broad range of data switching circuitry,” rejecting Patent Owner’s construction
`
`(“interconnected switches”) as “improperly limit[ing] the meaning of the term to
`
`preferred embodiments.” Id., 10-11.
`
`Patent Owner does not dispute that Reschke has a switch fabric under the
`
`Board’s construction. POR, 44-45; Ex. 1038, 170:6-171:14 (Reschke discloses “a
`
`data router that simultaneously routes data between a plurality of data requesters
`
`and a plurality of data providers”). Instead, Patent Owner makes a new attempt at
`
`reading limitations from the preferred embodiments into the claims, now
`
`contending that “switch fabric” is “a technical term of art” and should mean “a
`
`non-blocking arrangement of switches interconnecting in two dimensions.” See
`
`POR, 17.1
`
`The Board’s construction is correct and should be maintained, and Patent
`
`Owner’s new proposal rejected, for the following reasons.
`
`1.
`The Intrinsic Evidence Supports The Board’s Construction
`As the Board correctly explained, “the language of the claims does not limit
`
`the switch fabric for switching or exchanging data packets to any particular
`
`circuitry.” DI, 10. Similarly, nothing in the specification limits the claimed
`
`“switch fabric” to any particular embodiment, even the preferred embodiments.
`
`
` 1
`
` All emphasis used herein is added unless otherwise noted.
`
`2
`
`
`
`IPR2016-01106
`U.S. Patent No. 6,516,442
`Id., 11. For the same reasons that the Patent Owner’s “interconnected switches”
`
`construction would have improperly limited the claimed “switch fabric” to the
`
`preferred embodiments (id.), the claims and specification confirm that a switch
`
`fabric, as recited, need not be non-blocking or two dimensional. See DI, 12 (citing
`
`Ex. 1001, 1:49-51; 4:30-34).
`
`As Dr. Clark explains, a person of skill would not read the claims to demand
`
`that the switch fabric must be a non-blocking crossbar. Claim 1 recites a “plurality
`
`of microprocessors” and “a memory device,” thus covering a system with a single
`
`memory device. Ex. 1038, 175:21-177:7. Dr. Clark explains that, in such an
`
`arrangement, the single memory device would become busy and unavailable when
`
`a first processor was communicating with it, thereby preventing other
`
`(simultaneous) transactions to it, and negating any purported benefit of a non-
`
`blocking fabric (versus a blocking species). Ex. 1028, ¶10.
`
`2.
`The Extrinsic Evidence Supports The Board’s Construction
`Dr. Clark explains that extrinsic evidence (if it were to be consulted)
`
`confirms the Board’s construction, and does not limit switch fabrics to non-
`
`blocking species. Ex. 1028, ¶11.
`
`For example, “switch fabric” is defined in one computer encyclopedia as
`
`“(1) The internal interconnect architecture used by a switching device, which
`
`redirects the data coming in on one of its ports out to another of its ports. (2) The
`
`3
`
`
`
`IPR2016-01106
`U.S. Patent No. 6,516,442
`combination of interconnected switches used throughout a campus or large
`
`geographic area, which collectively provide a routing infrastructure.” Ex. 1032,
`
`875. Ex. 1028, ¶12.
`
`Textbooks use similar descriptions and definitions and also confirm that
`
`switch fabrics have blocking and non-blocking species. Peterson states that
`
`“[m]ost switches look conceptually similar to the one shown in Figure 3.28
`
`[below]. They consist of a number of input ports and output ports and a fabric.”
`
`
`
`Ex. 1030, 212-215. Peterson’s glossary defines “switching fabric” as “[t]he
`
`component of a switch that directs packets from their inputs to the correct outputs.”
`
`Id., 739; see also Ex. 1031, 190-191; Ex. 1028, ¶13.
`
`Mir states “[a] switch fabric is an interconnected network of smaller
`
`switching units,” before describing in more detail various species of “crossbar
`
`switch fabrics,” “blocking switch fabrics,” and “nonblocking switch fabrics.” Ex.
`
`1033, 349-361; Ex. 1028, ¶14.
`
`4
`
`
`
`IPR2016-01106
`U.S. Patent No. 6,516,442
`Even Dr. Alpert’s cited textbook (Hennessy-Patterson) compares two
`
`examples of “switch topologies,” the crossbar interconnection, which is
`
`nonblocking, and the Omega interconnection, which is blocking. Ex. 1034, 583-
`
`584; Ex. 1028, ¶16.
`
`Likewise, the Ph.D. dissertation of Peter Newman, in its “Simple
`
`Classification of Switch Designs,” specifically calls out “Blocking” and “Non-
`
`Blocking” species of “Non-Buffered Switch Fabric,” illustrated below in its Figure
`
`3.1 (annotated).
`
`
`
`5
`
`
`
`IPR2016-01106
`U.S. Patent No. 6,516,442
`
`Ex. 1029, 23-24; Ex. 1028, ¶15.2
`
`3.
`
`Patent Owner’s New Construction Is Inconsistent With The
`Intrinsic Evidence
`Because the claim language does not recite “non-blocking,” Patent Owner
`
`resorts to arguing that “non-blocking” is an inherent limitation of all switch fabrics.
`
`See Ex. 1038, 57:18-58:3 (“[I]f it blocks, then it’s not a switch fabric.”). However,
`
`the specification does not treat a “switch fabric,” a “switch matrix” or a “data
`
`switch” as inherently non-blocking, and in fact uses “non-blocking” as an adjective
`
`to describe a particular species of “data switch,” consistent with the extrinsic
`
`evidence mentioned above. Ex. 1001, Fig. 4. Likewise, the specification uses the
`
`non-defining and non-restrictive pronoun “which” when referring to “non-
`
`blocking” capabilities (i.e., “Simultaneous Switch Matrix…, which supports non-
`
`blocking data transfers” (Ex. 1001 at 3:23-25; POR, 19-20 (omitting “which” from
`
`quotation))), further confirming that “non-blocking” is just a species of switch
`
`fabric, not an inherent attribute of all switch fabrics (Ex. 1039, 59).
`
`
`
` 2
`
` This extrinsic evidence is also consistent with the construction proposed in the
`
`Petition: “a communication subsystem that provides for parallel routing of packets
`
`between multiple sources and targets.” Pet., 11.
`
`6
`
`
`
`IPR2016-01106
`U.S. Patent No. 6,516,442
`Patent Owner’s Extrinsic Evidence Does Not Support Its
`Construction
`While not disputing Dr. Alpert’s general ability to provide testimony in this
`
`4.
`
`Trial, Petitioner maintains that Dr. Alpert is in no position to provide an
`
`authoritative definition of the term based on his own experience. He could recall
`
`no basic history of switch fabrics, has no patents concerning them, has never
`
`authored an article about them, and has never before consulted on them other than
`
`in another unidentified but pending legal matter. Ex. 1038, 27:14-28:10; 31:3-8;
`
`34:18-35:1; 36:11-37:10; 39:10-40:8.
`
` Dr. Alpert contends “switch fabric” is a “technical term of art,” but
`
`identifies no extrinsic evidence actually using the term. Ex. 2020, ¶46; Ex. 1038,
`
`77:17-78:9; 77:11-15. To be clear, Dr. Alpert searched for such evidence, but
`
`could not recall whether he found any, and could not deny that he found extrinsic
`
`evidence that he did not cite because it did not match his view of the ʼ442 patent.
`
`Ex. 1038, 77:22-80:3.
`
`Instead, Dr. Alpert relies on Exhibit 2013, which never uses the term “switch
`
`fabric,” and instead refers to a particular species, a “crossbar switch matrix.”
`
`POR, 18; Ex. 2020, ¶22; Ex. 2013, Abstract; Ex. 1038, 67:2-14; 69:9-14 (“I don’t
`
`believe a switch matrix is limited to a crossbar.”); 69:20-70:5; 78:13-14 (“the term
`
`‘switch fabric’ as used in the ʼ442 patent is the same as switch matrix”). Thus, the
`
`7
`
`
`
`IPR2016-01106
`U.S. Patent No. 6,516,442
`exhibit, and Dr. Alpert’s testimony in connection with it, is of little or no probative
`
`value to the more general term “switch fabric.”
`
`
`
`Moreover, the basis of Dr. Alpert’s construction is Exhibit 2013’s statement
`
`that “nonblocking” is “usually omitted” because “crossbar switches” are “complete
`
`with respect to memory units (i.e., there is a separate bus associated with each
`
`memory, and the maximum number of transfers that can take place simultaneously
`
`is limited by the number of memory boxes….)” Ex. 2013, 107-108; POR, 18-19.
`
`From this, Dr. Alpert extrapolates that a switch fabric requires transfers to and
`
`from memory to any unused processor. Ex. 2020, ¶43. But the use of
`
`“nonblocking” in Exhibit 2013 is not only limited to a particular species of fabric,
`
`it confirms that non-blocking is with reference to memory access, not processor
`
`access. As explained below, this is significant because Exhibit 2013, to the extent
`
`it is relevant at all, confirms that Reschke is similarly non-blocking. Ex. 1028,
`
`¶23.
`
`5. Other Flaws With Patent Owner’s Construction
`Patent Owner’s construction also requires that a switch fabric must
`
`interconnect switches in “two dimensions.” POR, 18-19. Patent Owner and Dr.
`
`Alpert repeatedly used “horizontal” and “vertical” the dimensions to describe the
`
`’442 patent’s preferred embodiment. Id.; Ex. 2020, ¶¶44-45. However, during
`
`cross-examination Dr. Alpert testified (incredibly) that “two dimensions” in his
`
`8
`
`
`
`IPR2016-01106
`U.S. Patent No. 6,516,442
`construction does not mean “horizontal” or “vertical,” rather it just means there are
`
`“two sets of devices.” Ex. 1038, 51:8-21; 52:6-21; id., 76:5-13 (confirming that a
`
`configuration with processors and memories in parallel dimensions is a “switch
`
`fabric”). Dr. Alpert’s testimony confirms that Patent Owner’s “in two dimensions”
`
`requirement is not part of any “term of art,” is not used consistently even by Dr.
`
`Alpert, and should not be read into the claims.
`
`6.
`Conclusion
` Left unqualified, “switch fabric” may be either a blocking or non-blocking
`
`species. If a “non-blocking” switch fabric were intended, then the claim term
`
`should have been amended accordingly (which the Patent Owner had the
`
`opportunity to do through a Motion to Amend). Likewise, there is no justification
`
`for demanding two-dimensional interconnections. Just as the Board rejected Patent
`
`Owner’s prior construction (“interconnected switches”) because it was improperly
`
`limited to preferred embodiments, Patent Owner’s new construction is even
`
`narrower, and fails for similar reasons. DI, 10.
`
`B.
`“Packets Containing Data”
`The Board construed “packet” as “a basic unit of transport over the
`
`channel,” relying on the lexicography provided in the specification. DI, 12-13
`
`(citing Ex. 1001, 6:53 (“A ‘packet’ is the basic unit of transport over the
`
`Channel.”)). In so doing, the Board correctly noted that Patent Owner’s argument
`
`9
`
`
`
`IPR2016-01106
`U.S. Patent No. 6,516,442
`required a packet to include control information, but that the lexicography defining
`
`“packet” does not, nor “does [it] mention any particular embodiment” that includes
`
`control information. DI, 13 (citing Ex. 1001, 6:53-63).
`
`Patent Owner does not dispute that Reschke discloses a packet as construed
`
`by the Board. POR, 47. Instead, Patent Owner again attempts to import
`
`limitations from the preferred embodiments into the claims, this time contending
`
`that the larger phrase “packets containing data” should be construed as “a
`
`formatted transmission unit including at least data and control information.” Id.,
`
`24. Patent Owner bases its construction on dictionary definitions of “packet,” and
`
`an argument that the patent’s lexicography is not sufficiently “clear.” Id., 24-26.
`
`The Board previously rejected Patent Owner’s construction for trying to read
`
`in limitations from disclosed embodiments. DI, 15-16. It should do so again for
`
`the following reasons.
`
`1.
`The Intrinsic Evidence Supports The Board’s Construction
`The Board previously observed that the specification set off the term
`
`“packet” with quotation marks, followed by the definitional clause “a basic unit of
`
`transport over the channel.” DI, 13 (citing Ex. 1001, 6:53 and Sinorgchem Co.,
`
`Shandong v. Int’l Trade Comm’n, 511 F.3d 1132, 1136 (Fed. Cir. 2007)). Dr.
`
`Alpert uses similar forms of sentences when he proposes constructions in his own
`
`declaration. Ex. 2020, ¶43 (“nonblocking”), ¶54 (“packets containing data”), ¶65
`
`10
`
`
`
`IPR2016-01106
`U.S. Patent No. 6,516,442
`(“channel”). The definition of “packet” is also found in the portion of the
`
`specification devoted to “Channel Overview and Terminology.” Ex. 1001, 6:39.
`
`Similarly, the claims recite “packets containing data,” showing that if the
`
`patentee intended to limit packets to those including “control information,” the
`
`patentee would have done so, e.g., “containing data and control information.”
`
`2.
`The Extrinsic Evidence Supports The Board’s Construction
`Patent Owner’s “dictionary definition” from Exhibit 2007 is reproduced
`
`below in its entirety, using emphasis to denote portions of the definition which
`
`were omitted from the Patent Owner Response and Dr. Alpert’s Declaration.
`
`“packet: In general usage, a unit of information transmitted as a
`
`whole from one device to another on a network. In packet-switching
`
`networks, a packet is defined more specifically as a transmission unit
`
`of fixed maximum size that consists of binary digits representing both
`
`data and a header containing an identification number, source and
`
`destination addresses, and, sometimes, error-control data.”
`
`Ex. 2007, 5; compare with Ex. 2020, ¶63; Ex. 1038, 100:9-14.
`
`The “general usage” definition is nearly identical to the patent’s
`
`lexicography and the Board’s construction: “a basic unit of transport over the
`
`channel.” DI, 17; Ex. 1001, 6:53. Dr. Alpert had no explanation for why he
`
`11
`
`
`
`IPR2016-01106
`U.S. Patent No. 6,516,442
`omitted the “general usage” definition when trying to provide a construction of
`
`“packet” for the Board. Ex. 1038, 99:8-18; 103:1-11.
`
`As Dr. Clark explains, other extrinsic evidence is consistent with the
`
`“general usage” definition of packet omitted by Dr. Alpert, and the Board’s similar
`
`construction. Ex. 1028, ¶37 (citing Ex. 1037, 834 (“packet: a short fixed-length
`
`section of data that is transmitted as a unit in an electronic communications
`
`network”)).
`
`3.
`
`Patent Owner’s New Construction Is Inconsistent With The
`Intrinsic Evidence
`The Board rejected Patent Owner’s prior reliance on the specification’s
`
`statement that “the particular connotation of the term packet must be determined
`
`from the context in which it is used.” DI, 14. Patent Owner cites the same
`
`passages, and the same embodiments in its Patent Owner Response, but ignores the
`
`plain lexicography in the specification, and that “none of the challenged claims
`
`recites that a packet includes control information.” DI, 15; Ex. 1028, ¶38.
`
`4.
`
`Patent Owner’s Extrinsic Evidence Does Not Support Its
`Construction
`Patent Owner refers to portions of two dictionary definitions of “packet” as
`
`purported support for its new construction of “packets containing data.” POR, 25.
`
`In both cases, Patent Owner cites the dictionaries to support its claim that a packet
`
`must contain data and control information. Id. However, these passages would
`
`12
`
`
`
`IPR2016-01106
`U.S. Patent No. 6,516,442
`render the actual claim language “packets containing data” redundant and
`
`superfluous. Such constructions have long been disfavored. See Digital-Vending
`
`Servs. Int’l, LLC v. Univ. of Phoenix, Inc., 672 F.3d 1270, 1275 (Fed. Cir. 2012)
`
`(“If ‘registration server’ were construed to inherently contain the ‘free of content
`
`managed by the architecture’ characteristic, the additional ‘each registration server
`
`being further characterized in that it is free of content managed by the architecture’
`
`language in many of the asserted claims would be superfluous. This construction
`
`is thus contrary to the well-established rule that claims are interpreted with an eye
`
`toward giving effect to all terms in the claim.”). In Phillips v. AWH Corp., for
`
`example, the Federal Circuit confirmed that when a claim refers to “steel baffles,”
`
`this “strongly implies that the term ‘baffles’ does not inherently mean objects made
`
`of steel.” 415 F.3d 1303, 1314 (Fed. Cir. 2005). Here, the claim refers to packets
`
`containing data, which strongly implies that “packets” do not inherently contain
`
`data or control information.
`
`Patent Owner also concedes that it would be inappropriate to use a
`
`dictionary definition to transform the meaning of “packet” as used in the context of
`
`the specification into the meaning in the abstract. POR, 34-35 (citing Phillips, 415
`
`F.3d at 1315). But Patent Owner’s only explanation for why its construction does
`
`not do exactly that is that “the extrinsic and intrinsic evidence all point to the same
`
`construction.” POR, 35. As shown above, however, only through selective
`
`13
`
`
`
`IPR2016-01106
`U.S. Patent No. 6,516,442
`omission of “general usage” definitions can Dr. Alpert claim that his extrinsic
`
`evidence is consistent with Patent Owner’s construction. As such, the
`
`“admonitions from Phillips about dictionaries” are not “misplaced,” but directly on
`
`point. See POR, 35.
`
`5. Other Flaws With Patent Owner’s Construction
`Patent Owner not only ignores the specification’s lexicography and omits
`
`unhelpful extrinsic evidence regarding the term’s purported “ordinary meaning,”
`
`but also contradicts its own litigation position about the term’s “ordinary
`
`meaning.” Patent Owner has argued in district court that “the ʼ422 [sic] patent
`
`describes a packet consistent with its ordinary meaning as a basic unit of transport
`
`over the channel.” Ex. 1026, 71:3-13. Patent Owner’s shifting, inconsistent view
`
`of the “ordinary meaning” of “packet” demonstrates that its reliance on such
`
`“ordinary meaning” to overcome explicit lexicography deserves little weight.
`
`In addition, Patent Owner’s proposed “broadest reasonable interpretation” is
`
`narrower than its position in court, and thus cannot be correct. Ex. 1038, 89:10-
`
`90:11; see Facebook, Inc. v. Pragmatus AV, LLC, 582 F. App’x 864, 869 (Fed. Cir.
`
`2014) (the broadest reasonable interpretation of a claim term in the PTAB “cannot
`
`be narrower” “than the construction of a term under the Phillips standard”).
`
`14
`
`
`
`IPR2016-01106
`U.S. Patent No. 6,516,442
`
`6.
`Conclusion
`“Packet” left unqualified should mean a basic unit of transport over the
`
`channel. If the patentee wanted to limit the packets to contain control information,
`
`the claims would have recited such, as they did for data. The Patent Owner had the
`
`opportunity to further limit the claims through a Motion to Amend but declined to
`
`do so. Just as the Board rejected Patent Owner’s prior construction of “packet,”
`
`Patent Owner’s construction of “packets containing data” similarly fails. See Epos
`
`Techs. Ltd. v. Pegasus Techs. Ltd., 766 F.3d 1338, 1341 (Fed. Cir. 2014).
`
`III. GROUNDS
`A. Ground 2: Claims 1, 12, 24 And 34 Are Obvious Over Reschke
`Patent Owner argues that Reschke does not have a switch fabric and does
`
`not switch or exchange packets. POR, 42-48. In doing so, Patent Owner relies
`
`solely on its new constructions, arguing that Reschke’s data switch is “blocking,”
`
`and that its control information is not “packetized.” POR, 44-45. Patent Owner
`
`does not dispute that these limitations are satisfied under the Board’s claim
`
`construction. Thus, if the Board confirms its constructions (as it should), it should
`
`reject Patent Owner’s rebuttal arguments concerning these limitations for this
`
`reason alone.
`
`15
`
`
`
`IPR2016-01106
`U.S. Patent No. 6,516,442
`Even if the Board were to adopt Patent Owner’s new constructions (which it
`
`should not), Patent Owner’s arguments still fail on the merits, because Reschke
`
`meets these limitations.
`
`1.
`
`Reschke Discloses A Switch Fabric Even Under Patent
`Owner’s New Construction.
`Reschke’s system data switch is properly understood as non-blocking for
`
`two reasons. First, Exhibit 2013 describes switches as “non-blocking” when they
`
`are “complete with respect to memory units, (i.e., there is a separate bus
`
`associated with each memory, and the maximum number of transfers that can take
`
`place simultaneously is limited by the number of memory boxes….” Ex. 2013,
`
`107-108; Ex. 1038, 72:9-75:1. As Dr. Clark explains, the Reschke switch fabric
`
`has “a separate bus associated with each memory,” and allows such simultaneous
`
`transfers. Ex. 1028, ¶¶42-44.
`
`For example, Dr. Clark explains that Reschke qualifies as non-blocking and
`
`shows an example where two processors send data to MSU 141 and XSU 142.
`
`16
`
`
`
`IPR2016-01106
`U.S. Patent No. 6,516,442
`US. Patent No. 6,516,442
`
`IPR2016-01106
`
`m1“. sns
`
`cm '
`
`
`CH2 1 CLK
`LEI-ll
`
` p
`
`
`,_....p.p
`
`To KSU 142 thruugh
`MUX 421
`
`1n
`421-12?
`
`I
`I
`1
`|
`
`414:1
`
`F1 G. —4A
`FIG 13
`FIE
`4A
`
`Im
`“U “6“
`
`SUP
`15”
`asuVI
`
`E r";—
`
`
`
`17
`
`17
`
`
`SEE
`
`HH\.‘l0'
`
`._.II}I:p
`
`512
`
`To MSU 141
`
`132
`
`
`
`o—nl—I-1|1:‘U
`
`133
`
`
`
`
`
`IPR2016-01106
`U.S. Patent No. 6,516,442
`
`
`
`
`
`Ex. 1028, ¶¶45-47.
`
`Likewise, Dr. Alpert admits Reschke allows two processors to
`
`simultaneously communicate with two memories. Ex. 2020, ¶72 (“Reschke’s
`
`system data switch (110) … [allows] two processors to communicate with two
`
`memories at the same time”). This more than satisfies the claims, even if they
`
`18
`
`
`
`IPR2016-01106
`U.S. Patent No. 6,516,442
`required non-blocking transfers between two processors and one memory device.
`
`Ex. 1001, 36:35-42.
`
`Second, Reschke discloses non-blocking switch fabrics to the same degree
`
`the ’442 patent does. Patent Owner cites the following description from the ʼ442
`
`specification as “consistent with the characterization of a switch matrix as
`
`nonblocking”: “[The Data Switch] provides for parallel routing of transaction data
`
`between multiple initiators and multiple targets. The Data Switch is designed to let
`
`multiple, simultaneous data transfers take place to/from initiators and from/to
`
`targets (destinations of transactions).” POR, 20 (citing Ex. 1001, 4:30-37)
`
`(emphasis Patent Owner’s). Yet, Dr. Alpert admitted that Reschke has the nearly
`
`identical disclosure: “a data router that simultaneously routes data between a
`
`plurality of data requesters and a plurality of data providers.” Ex. 1038, 170:6-17.
`
`Patent Owner also cannot dispute that Reschke’s data switch satisfies its
`
`construction’s “two dimensions” requirement, because Dr. Alpert now admits that
`
`the two dimensions are (1) processors and (2) memories, and Reschke indisputably
`
`transfers data “between memory and multiple microprocessors.” POR, 43; Ex.
`
`1038, 51:15-21, 55:14-19; Ex. 1003, 2:15-19.
`
`2.
`
`Reschke Discloses Packets Containing Data Even Under
`Patent Owner’s New Construction
`Reschke satisfies Patent Owner’s proposed construction of “packets
`
`containing data” through its disclosure of error tags and DIE bits. Dr. Alpert
`
`19
`
`
`
`IPR2016-01106
`U.S. Patent No. 6,516,442
`admitted that Reschke includes an “error tag” along with the transmitted data. Ex.
`
`1038, 168:14-18; POR, 47 (“These datawords and transfer blocks can have an error
`
`section in some embodiments.” (citing Ex. 1003, Fig. 2 (216-218), Fig. 3)). When
`
`these tags are detected, “the data destination unit 234 discards the error-infected
`
`data and, if it had previously requested data, it sends a new data request (e.g., over
`
`a control bus such as 181 in Fig. 1) to the system control unit (SC) 180 requesting
`
`transmission of the data it had earlier requested.” Ex. 1003, 10:16-25. Thus, the
`
`error tags in Reschke are control information because they dictate whether
`
`received data is discarded and whether the recipient should request retransmission
`
`of data. Patent Owner’s own dictionary definition confirms that packets can
`
`contain “error-control data.” Ex. 2007, 5; Ex. 1028, ¶¶50-51.
`
`3.
`Reschke Discloses Full Duplex Channels
`Patent Owner contends that the Petition did not meet its burden of proving
`
`that Reschke has “full-duplex” channels. POR, 48-54. However, the Board
`
`previously noted that the Petitioner’s “argument and evidence with respect to
`
`Figures 4A and 4B” cited “sufficient disclosure from Reschke to show Reschke’s
`
`data buses are full-duplex.” DI, 30. As the Board explained in its Decision to
`
`Institute, Patent Owner’s preliminary arguments centered on Figure 2 and were not
`
`indicative of the more detailed depictions in Figures 4A-C. DI, 28-30; Ex. 1002,
`
`¶¶136-140. In view of this, the Board encouraged the parties to address whether
`
`20
`
`
`
`IPR2016-01106
`U.S. Patent No. 6,516,442
`the “forward” and “reverse” pathways in Reschke’s data switch as shown in Figure
`
`4 can operate at the same time to provide full-duplex transmission. DI, 30.
`
`Patent Owner ignores the Board’s invitation, and instead argues that “the
`
`question is not whether Reschke can provide full-duplex transmission. The
`
`question is whether Reschke would have operated in full-duplex mode.” POR, 53-
`
`54 (emphasis Patent Owner’s). To the contrary, the operative legal question is
`
`whether Reschke would have disclosed full duplex channels, not transmissions.
`
`As explained below, and as Dr. Alpert now agrees, a full duplex channel is
`
`capable of full duplex transmissions, thus “whether Reschke can provide full-
`
`duplex transmission” is precisely the right factual question.
`
`Here, the Petition and the original Clark Declaration confirm that the
`
`Reschke channels as depicted and explained in connection with Figure 4 are full
`
`duplex. These channels have forward and reverse portions (a fact not disputed by
`
`Dr. Alpert), and as Dr. Clark has explained, nothing in Reschke prevents
`
`communications from simultaneously occurring in both directions, i.e., thereby
`
`allowing for full duplex transmissions. Ex. 1002, ¶¶136-140; Ex. 2012, 39:20-
`
`40:17, 45:13-24; Ex. 1028, ¶¶54-58.
`
`Reschke’s Figures 4A-4C, annotated below, plainly show this capability as
`
`(1) PU 131 sending data to MSU 141 and (2) MSU 141 simultaneously sending
`
`21
`
`
`
`IPR2016-01106
`U.S. Patent No. 6,516,442
`data to PU 131 (thereby showing a full duplex channel between PU 131 and the
`
`SDS 110 and another full duplex channel between the MSU 141 and the SDS 110).
`
`
`
`22
`
`
`
`IPR2016-01106
`U.S. Patent No. 6,516,442
`
`
`
`
`
`Dr. Clark explains this capability exists for each channel, and certainly for the two
`
`processors and one memory recited in the claims. Ex. 1028, ¶¶57-60.
`
`As Dr. Clark testified, these figures show “that there are many paths from
`
`each source to each destination” and nothing in Reschke indicates that something
`
`could not simultaneously go out and come in at any one destination. See Ex. 2012,
`
`23
`
`
`
`IPR2016-01106
`U.S. Patent No. 6,516,442
`40:2-17. Accordingly, Dr. Clark confirmed that in his view, Reschke “expressly
`
`disclose[s] simultaneous data communications in both directions” “at least because
`
`in [Reschke’s] Figure 4s [i.e., 4A-C] about the inside of the switch it is quite clear
`
`that many, many conversations with many, many inter party communications can
`
`go on simultaneously.” Id., 40:25-41:8; see also Ex. 1038, 170:6-17 (Reschke
`
`discloses “a data router that simultaneously routes data between a plurality of data
`
`requesters and a plurality of data providers”).
`
`Patent Owner’s argument that Reschke does not disclose full duplex
`
`channels did not withstand cross-examination. Dr. Alpert confirmed that he was
`
`relying on a dictionary definition for “duplex transmission” and not for “duplex
`
`channel,” even though the latter was mere lines higher in his dictionary, and is
`
`clearly the more relevant definition since the term at issue is “channel.” Ex. 1038,
`
`131:19-132:9; Ex. 2007, 4. That definition confirms that a full duplex “channel”
`
`means “a communications link that allows for duplex (two-way) transmission.”
`
`Ex. 2007, 4. When confronted with the more relevant definition, Dr. Alpert
`
`conceded that a duplex channel is one that has the capability to provide full-duplex
`
`transmissions and “allows” full duplex transmission. Ex. 1038, 135:12-15 (“Q:
`
`But do you have any issue with the fact that the definition of ‘channel’ uses the
`
`word ‘allows’? A: No, as – as I’m reading that, I think it’s describing a
`
`capability.”); 133:7-10.
`
`24
`
`
`
`IPR2016-01106
`U.S. Patent No. 6,516,442
`Both the more relevant dictionary definition and Dr. Alpert’s cross-
`
`examination testimony confirm that the Board’s understanding of the term – i.e., a
`
`channel that “can operate at the same time to provide full duplex transmission”
`
`(DI, 30) – is correct. Indeed, the question is whether Reschke can provide full-
`
`duplex transmission (POR, 53-54), because a full duplex channel concerns a
`
`capability – what the channel allows to happen. The claim does not recite full
`
`duplex transmissions and instead only requires the capability to allow such
`
`transmissions.
`
`Indeed, on close inspection, Patent Owner never disputes that Reschke has
`
`the capability to provide full-duplex transmissions; instead it argues there is no
`
`evidence of actual full duplex transmissions. This is beside the point because such
`
`communications are a result of the software programs that would be running at any
`
`given instant. See, e.g., POR, 54 (arguing that there is no evidence Reschke
`
`“would have operated in full-duplex mode”). Patent Owner again focuses on
`
`Figure 2, arguing that conclusions drawn from Figure 2 “reinforce” otherwise
`
`uncorro
Accessing this document will incur an additional charge of $.
After purchase, you can access this document again without charge.
Accept $ Charge
Still Working On It
This document is taking longer than usual to download. This can happen if we need to contact the court directly to obtain the document and their servers are running slowly.
Give it another minute or two to complete, and then try the refresh button.
A few More Minutes ... Still Working
It can take up to 5 minutes for us to download a document if the court servers are running slowly.
Thank you for your continued patience.
This document could not be displayed.
We could not find this document within its docket. Please go back to the docket page and check the link. If that does not work, go back to the docket and refresh it to pull the newest information.
Your account does not support viewing this document.
You need a Paid Account to view this document. Click here to change your account type.
Your account does not support viewing this document.
Set your membership
status to view this document.
With a Docket Alarm membership, you'll
get a whole lot more, including:
- Up-to-date information for this case.
- Email alerts whenever there is an update.
- Full text search for other cases.
- Get email alerts whenever a new case matches your search.
One Moment Please
The filing “” is large (MB) and is being downloaded.
Please refresh this page in a few minutes to see if the filing has been downloaded. The filing will also be emailed to you when the download completes.
Your document is on its way!
If you do not receive the document in five minutes, contact support at support@docketalarm.com.
Sealed Document
We are unable to display this document, it may be under a court ordered seal.
If you have proper credentials to access the file, you may proceed directly to the court's system using your government issued username and password.
Access Government Site
