`Trials@uspto.gov
`571-272-7822 Entered: November 28, 2017
`UNITED STATES PATENT AND TRADEMARK OFFICE
`____________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`____________
`
`INTEL CORPORATION,
`Petitioner,
`
`v.
`
`ALACRITECH, INC.,
`Patent Owner.
`____________
`
`Case IPR2017-01406
`Patent 7,673,072 B2
`____________
`
`
`
`Before STEPHEN C. SIU, DANIEL N. FISHMAN, and
`WILLIAM M. FINK, Administrative Patent Judges.
`
`FINK, Administrative Patent Judge.
`
`
`
`
`DECISION
`Institution of Inter Partes Review
`37 C.F.R. § 42.108
`
`
`
`I. INTRODUCTION
`Intel Corporation (“Petitioner”) requests inter partes review of claims
`1–21 of U.S. Patent No. 7,673,072 B2 (“the ’072 patent,” Ex. 1001) pursuant
`to 35 U.S.C. §§ 311 et seq. Paper 1 (“Pet.”). Alacritech, Inc. (“Patent
`Owner”) filed a preliminary response. Paper 9 (“Prelim. Resp.”). Institution
`
`
`
`IPR2017-01406
`Patent 7,673,072 B2
`
`of an inter partes review is authorized by statute when “the information
`presented in the petition . . . and any response . . . shows that there is a
`reasonable likelihood that the petitioner would prevail with respect to at least
`1 of the claims challenged in the petition.” 35 U.S.C. § 314(a); see
`37 C.F.R. § 42.108. Upon consideration of the Petition and Preliminary
`Response, we conclude the information presented shows there is a
`reasonable likelihood that Petitioner would prevail in establishing the
`unpatentability of at least one of claims 1–21 of the ’072 patent.
`
`A. Related Matters
`We are informed that the ’072 patent is presently related to the
`following: Alacritech, Inc. v. CenturyLink, Inc., Case No. 2:16-cv-00693-
`JRG-RSP (E.D. Tex.); Alacritech, Inc. v. Wistron Corp., Case No. 2:16-cv-
`00692-JRG-RSP (E.D. Tex.); and Alacritech, Inc. v. Dell Inc., Case No.
`2:16-cv-00695-RWS-RSP (E.D. Tex.). Pet. 3.
`
`B. The ’072 Patent (Ex. 1001)
`The ’072 patent describes a system and method “that greatly increases
`the speed of [processing network communication] and the efficiency of
`transferring data being communicated.” Ex. 1001, 5:24–27.
`
`C. Illustrative Claim
`Independent claim 1, reproduced below, is illustrative of the claimed
`subject matter:
`1. A method comprising:
`establishing, at a host computer, a transport layer
`connection, including creating a context that includes protocol
`header information for the connection;
`transferring the protocol header information to an interface
`
`2
`
`
`
`IPR2017-01406
`Patent 7,673,072 B2
`
`device;
`transferring data from the network host to the interface
`device, after transferring the protocol header information to the
`interface device;
`dividing, by the interface device, the data into segments;
`creating headers for the segments, by the interface device,
`from a
`template header containing
`the protocol header
`information; and
`prepending the headers to the segments to form transmit
`packets.
`Id. at 97:17–31.
`
`D. Asserted Grounds of Unpatentability
`Petitioner asserts that claims 1–21 are unpatentable on the following
`grounds (Pet. 14–15):
`
`Basis
`Reference(s)
`Erickson1 and Tanenbaum2 § 103
`
`Claims challenged
`1–21
`
`II. DISCUSSION
`A. Real Party in Interest
`Intel Corporation identifies itself as a real party in interest in these
`proceedings and represents that “[n]o other parties exercised or could have
`exercised control over this [P]etition; no other parties funded or directed this
`Petition.” Pet. 2. Patent Owner argues that “the Petition . . . fails to identify
`at least Dell Inc. (‘Dell’) and Cavium Inc. (‘Cavium’)” as real parties-in-
`interest and that “[t]he Board should deny institution . . . because the Petition
`
`1 U.S. Patent No. 5,768,618, issued June 16, 1998 (“Erickson,” Ex. 1005).
`2 Andrew S. Tanenbaum, COMPUTER NETWORKS (3rd ed. 1996)
`(“Tanenbaum,” Ex. 1006).
`
`3
`
`
`
`IPR2017-01406
`Patent 7,673,072 B2
`
`fails to identify all real parties in interest as required by 35 U.S.C.
`§ 312(a)(2) and 37 CFR § 42.8(b)(1).” Prelim. Resp. 12–13. We disagree.
`As an initial matter, as Patent Owner points out, in determining
`whether a party is a real party-in-interest, “[a] common consideration is
`whether the non-party exercised or could have exercised control over a
`party’s participation in a proceeding.” Id. at 13–14 (citing Office Patent
`Trial Practice Guide, 77 Fed. Reg. 48756, 48759–60). Patent Owner further
`argues that “Intel has agreed to defend and partially indemnify Dell,” “Intel
`is Dell’s supplier with regard to Dell’s accused products,” “Intel admitted
`that it would have to work closely with Dell in . . . litigation,” “Intel also
`admitted that it has a close relationship to Dell financially in the district
`court case,” “Intel chose . . . to passively reimburse Dell [and also] play an
`active role to assist, protect, and defend Dell,” “Dell [was] originally
`accused of infringing the ’072 patent in the district court case, not Intel,”
`“Intel’s products were not accused in the original pleading,” “Dell desires
`review of the ’072 Patent,” “Dell and Intel have repeatedly coordinated their
`invalidity theories,” “Dell and Intel also shared a technical expert, Mr. Mark
`Lanning,” and that “Intel has effective choice of invalidity theories and
`proofs.” Id. at 14–20.
`Even accepting all of these contentions, for purposes of this Decision,
`we are not persuaded on these facts that Dell exercised or could have
`exercised control over the preparation or filing of the present Petition.
`Indeed, the alleged financial relationship with Intel as the indemnitor of Dell
`suggests that, if anything, Intel would be the one to control the preparation
`and filing of and present Petition. The other assertions relating to
`coordinating theories and sharing experts are common activities between
`
`4
`
`
`
`IPR2017-01406
`Patent 7,673,072 B2
`
`cooperating co-defendants in related litigation and are not suggestive of
`control of or ability to control this petition. See Weatherford Int’l, LLC, et
`al. v. Packers Plus Energy Services, Inc., Case IPR2016-01514, slip op. 12–
`16 (PTAB Feb. 22, 2017) (Paper 23).
`Patent Owner also argues that “Cavium is also a supplier of Dell,”
`“petitions filed by Intel and Cavium also share an identical declaration from
`the same expert, Dr. Robert Horst,” and “Cavium also filed an almost
`verbatim petition.” Prelim. Resp. 19. However, for reasons similar to those
`discussed above, we are not persuaded by Patent Owner’s argument that
`Cavium should have been identified as a real party-in-interest on these facts.
`We reject Patent Owner’s argument that we deny institution on this
`basis for an additional reason. Significantly, Patent Owner does not allege
`that inclusion of either Dell or Cavium as a real party in interest would have
`barred the Petition under 35 U.S.C. § 315. Aside from a bar defense, under
`the Board’s precedential decision in Lumentum Holdings, Inc. v. Capella
`Photonics, Inc., our jurisdiction to consider a petition does not require a
`“correct” identification of all real parties in interest in a petition. Case
`IPR2015-00739, slip op. at 6 (PTAB March 4, 2016) (Paper 38)
`(precedential); see also Blue Coat Sys., Inc. v. Finjan, Inc., Case IPR2016-
`01444, slip op. 10 (PTAB July 18, 2017) (Paper 11) (“Evidence [of failure to
`identify all real parties in interest] is, at best, suggestive of an issue that is
`not jurisdictional.”). Consequently, even if Dell and Cavium should be
`named real parties-in-interest, as Patent Owner alleges, failure to identify
`them as such at the time the Petition was filed does not require us to
`terminate the proceeding. Indeed, later PTAB decisions indicate that a
`petition may be corrected after institution of trial to add a real party in
`
`5
`
`
`
`IPR2017-01406
`Patent 7,673,072 B2
`
`interest if warranted without assigning a new filing date to the petition. E.g.,
`Axon EP, Inc., et al. v. Derrick Corp., Case IPR2016-00642, slip op. at 3
`(PTAB November 21, 2016) (Paper 17).
`We have considered Patent Owner’s argument that “[f]inding that
`Dell and Cavium are real parties in interest is . . . consistent with the express
`legislative intent concerning the need for quiet title.” Prelim. Resp. 20. For
`example, Patent Owner contends that
`If Intel were to be able to institute this IPR without adding Dell
`and Cavium as real parties in interest, Dell and Cavium would be
`able to use the same prior art relied upon here in other cases even
`if the patent is held to be valid in this IPR. This would be
`harassment through repeated litigation in violation of legislative
`intent, as Intel, Dell, and Cavium would be able to “double dip”
`and use the same invalidity theories to defend the same accused
`products twice.
`
`Id. at 21. We disagree. Notwithstanding our preliminary determination here
`that Dell and Cavium have not been shown to be real parties-in-interest, our
`decision does not preclude Patent Owner from raising a bar defense in a
`future proceeding and attempting to develop evidence sufficient to support
`such a defense.3
`
`B. Claim Construction
`In an inter partes review, we construe claim terms in an unexpired
`patent according to their broadest reasonable construction in light of the
`specification of the patent in which they appear. 37 C.F.R. § 42.100(b).
`
`
`3 Cavium has moved to join this proceeding, see Case IPR2017-01707
`(Paper 3), and Dell would appear to be time-barred, under 35 U.S.C.
`§ 315(b), regardless in any such proceeding. See Prelim. Resp. 12 n. 4
`(noting June 30, 2016 as the date of filing the complaint against Dell).
`6
`
`
`
`IPR2017-01406
`Patent 7,673,072 B2
`
`Consistent with the broadest reasonable construction, claim terms are
`presumed to have their ordinary and customary meaning as understood by a
`person of ordinary skill in the art in the context of the entire patent
`disclosure. In re Translogic Tech., Inc., 504 F.3d 1249, 1257 (Fed. Cir.
`2007).
`“context”
`Petitioner contends the term “context,” recited in independent claims
`1, 9, and 15, is indefinite, “because the specification and claims do not
`inform a [person of ordinary skill in the art] what information is necessary or
`sufficient to be a context.” Pet. 25. Nonetheless, for purposes of this
`Decision, “Petitioner submits that a ‘context’ for a connection must include
`at least data regarding a connection.” Id. Patent Owner contends that
`Petitioner’s contention that the term “context” is indefinite “should end the
`Board’s inquiry” into claims 1–21. Prelim. Resp. 10.
`We do not agree with Patent Owner that Petitioner’s contention that
`the term “context” is indefinite for purposes of a related proceeding “should
`end the Board’s inquiry.” Indeed, other than noting their positions in related
`litigation, neither party provides any analysis one way or the other in regards
`to whether the claim is indefinite. See Pet. 25; Prelim. Resp. 10–11. As to
`whether the scope of the term is discernable, we observe that claim 1 does
`not use the term “context” in isolation, but states “the context that includes
`protocol header information for the connection.” Thus, it is apparent from
`the claim itself that the term “context” at least requires protocol header
`information.
`
`7
`
`
`
`IPR2017-01406
`Patent 7,673,072 B2
`
`This is consistent with its meaning in the specification. For example,
`with regard to protocol stack 44 on host CPU 28 in Figure 2, the
`specification explains:
`The upper layer interface 42, along with the CPU 28 and any
`related controls can send or retrieve a file to or from the upper
`layer 46 or storage 35, as shown by arrow 48. A connection
`context 50 has been created, as will be explained below, the
`context summarizing various features of the connection, such as
`protocol type and source and destination addresses for each
`protocol layer.
`Ex. 1001, 7:62–8:2 (emphasis added), 30:54–58. In other words, similar to
`claim 1’s “context,” the “connection context” includes data relating to the
`connection for each relevant protocol layer.
`In view of the foregoing, and for purposes this Decision, we do not
`find that the scope of the claim requires “considerable speculation,” In re
`Steele, 305 F.2d 859, 862–63 (CCPA 1962), such that it would be improper
`for us to analyze the challenged claims under 35 U.S.C. § 103(a). Patent
`Owner argues that “context” does not otherwise require construction.
`Prelim. Resp. 13. We agree. Because neither party presents arguments in
`support of an explicit construction or, in their contentions, rely on a
`particular scope for this term beyond that which the claim itself requires
`(i.e., protocol header information), we determine that no explicit
`construction is necessary at this stage of the proceeding.
`Neither party presents arguments that turn on the construction of other
`claim terms. Accordingly, at this stage of the proceeding, we determine that
`it is not necessary to provide an express interpretation of any other claim
`terms.
`
`8
`
`
`
`IPR2017-01406
`Patent 7,673,072 B2
`
`C. Asserted Unpatentability of Claims 1–21
`over Erickson and Tanenbaum
`Petitioner contends claims 1–21 are unpatentable as directed to
`
`obvious subject matter over the combination of Erickson and Tanenbaum.
`Pet. 34–76.
`
`1. Overview of Erickson
`Erickson is directed to a “method of controlling an input/output (I/O)
`
`device connected to a computer to facilitate fast I/O data transfers.” Ex.
`1005, Abstract. Figure 3 of Erickson is reproduced below:
`
`
`Figure 3 depicts data flow in accordance with Erickson’s invention. As
`shown in Figure 3, slow application 306 uses normal stream processing 308
`and pass-through driver 310 to send information to I/O device adapter 314 to
`commodity interface 322. Id. at 4:53–61. Alternatively, fast applications
`302 and 304 send information directly to I/O adapter 314 via “VIRTUAL
`HARDWARE” 316 and 318 avoiding the overhead of the streams
`processing and pass-through driver. Id. at 4:61–5:3.
`
`9
`
`
`
`IPR2017-01406
`Patent 7,673,072 B2
`
`2. Overview of Tanenbaum
`Tanenbaum describes general principles of data transmission in
`computer networks. Ex. 1006.
`
`3. Analysis
`Petitioner contends claims 1–21 are unpatentable under 35 U.S.C.
`§ 103(a) as obvious over the combination of Erickson and Tanenbaum. Pet.
`34–76. Petitioner relies on the testimony of Dr. Robert Horst to support its
`contentions regarding how the cited references describe the claim
`limitations. Id. (citing Ex. 1003). At this stage of the proceeding, Petitioner
`has accounted sufficiently for the limitations of at least 1 of the claims
`challenged in the Petition (i.e., claim 1) and provided sufficient articulated
`reasoning with rational underpinning to support the conclusion of
`obviousness.
`“A method comprising: establishing, at a host computer, a
`transport layer connection, including creating a context that
`includes protocol header information for the connection”
`Petitioner contends Erickson discloses an I/O device adapter 314 is
`connected to a host computer and that a user process running on the host
`computer establishes a “transport layer connection.” Pet. 38 (citing Ex.
`1005, 1:65–67, 6:1–9; Ex. 1003, Appendix A, 1.1). Petitioner further
`contends the user process creates a “context” by “opening a device driver
`and specifying the protocol type (e.g., UDP or TCP).” Id. at 39–40 (citing
`Ex. 1005, 6:1–9, 8:2–9). According to Petitioner, the created context
`“includes protocol header information,” as the claim requires, because it is
`described as “including ‘almost everything’ concerning a UDP datagram
`‘except the actual user data.’” Id. at 40 (citing Ex. 1005, 6:57–7:4).
`Furthermore, Petitioner contends, the context is “for the connection,”
`
`10
`
`
`
`IPR2017-01406
`Patent 7,673,072 B2
`
`because it discloses TCP scripts which are inherently connection oriented.
`Id. at 41 (citing Ex. 1003, Appendix A, 1.1).
`“transferring the protocol header information to an interface
`device”
`Petitioner contends Erickson I/O device adapter 314 is the claimed
`“interface device.” Pet. 41. Petitioner also contends that Erickson’s
`datagram template 702 in the I/O device adapter’s memory, comprises
`protocol header information that was transferred to the I/O device adapter
`from the host computer. Id. at 41–42 (citing Ex. 1005, 7:39–41, 5:37–51,
`8:2–4).
`“transferring data from the network host to the interface device,
`after transferring the protocol header information to the
`interface device”
`Petitioner contends that in Erickson, data is transferred from the host
`to the interface device, after transferring the protocol header information to
`the interface device, because template 702 resides in the adapter’s memory
`before the script is executed to store data in the I/O device adapter’s
`memory. Id. at 42–43 (quoting Ex. 1005, 7:39–48 (“FIG. 7 is a block
`diagram illustrating a UDP datagram template 702 (without a user data area)
`residing in the I/O device adapter’s memory. . . . The I/O device adapter
`stores the user data provided by the user process in the I/O device adapter’s
`memory, and then transmits the completed UDP datagram 702 over the
`media.”).
`“dividing, by the interface device, the data into segments”
`Petitioner contends Erickson in combination with Tanenbaum
`discloses this limitation. Pet. 43. Petitioner contends that, although
`Erickson’s exemplary context is UDP, it also discloses the use of TCP/IP
`and refers readers to the 1981 edition of Tanenbaum and discloses use of
`11
`
`
`
`IPR2017-01406
`Patent 7,673,072 B2
`
`TCP/IP scripts. Pet. 34–35 (citing Ex. 1005, 4:38–43); id. at 44 (citing Ex.
`1005, 5:47–51). Petitioner provides evidence that a person of ordinary skill
`in the art would have been motivated to implement “fast path” TCP/IP on
`Erickson’s I/O device adapter in view of the 1996 version of Tanenbaum
`(i.e., Ex. 1006) with a high expectation of success to take advantage of
`TCP/IP’s popularity for Internet and World Wide Web processing. Id. at
`35–38 (citing Ex. 1003 ¶¶ 137–148).
`Petitioner further contends that Tanenbaum discloses that TCP/IP data
`is exchanged in segments. Pet. 43 (citing Ex. 1006, 525 (“The sending and
`receiving TCP entities exchange data in the form of segments.”), 522 (“A
`TCP entity accepts user data streams from local processes, breaks them up
`into pieces . . . and sends each piece as a separate IP datagram.”).4 Petitioner
`contends that the “transport entity, which does the dividing, can be on a
`network interface card (network interface device).” Id. at 43–44 (citing Ex.
`1006, 480). According to Petitioner,
`A [person of ordinary skill in the art] would know how to
`modify Erickson’s UDP script (udpscript at Ex.1005, Erickson at
`7:51-64) so that, rather than filling in the UDP Length and
`Checksum shown in Erickson, the modified script would fill in
`the TCP Sequence number, Length and Checksum. This script
`involves minimal changes from the UDP script. Ex.1003, Horst
`Decl. Appendix A limitation 1.4. The “dividing, by the interface
`device, the data into segments” limitation is met by the foregoing
`obvious TCP script for Erickson. Erickson’s senduserdatagram
`(Ex.1005, Erickson at 7:23-32) user process triggers the I/O
`device adapter to send each segment, by “spanking” the GO
`register. The adapter calculates the TCP checksum, transmits the
`packet, updates the shared state in Hardware Register 504, and
`
`4 We refer to the original page numbering in Tanenbaum, not the exhibit
`page numbers added by Petitioner.
`
`12
`
`
`
`IPR2017-01406
`Patent 7,673,072 B2
`
`waits for GO to be set again for the next segment. Repeated
`invocations of the TCP script (by repeatedly spanking the GO
`register) with pointers to data for consecutive segments result in
`the interface device dividing the user data stream into TCP
`segments.
`Pet. 44–45. Petitioner contends other TCP scripts would also have been
`obvious modifications to Erickson. Id. at 45–46 (citing Ex. 1003, Appendix
`A, 1.4).
`
`Patent Owner advances various arguments as to why the proposed
`
`combination does not teach dividing the data into segments or teaches away
`from such a combination.
`
`First, Patent Owner contends Petitioner’s cites to Tanenbaum are out
`of context and unrelated. Prelim. Resp. 25–26. Specifically, Patent Owner
`observes that the sentence before Petitioner’s quote from Tanenbaum (Pet.
`43 (quoting Ex. 1006, 522)) that a TCP entity accepts streams and breaks
`them up into pieces sent as separate IP datagrams reads, “[e]ach machine
`supporting TCP has a TCP transport entity, either a user process or part of
`the kernel that manages TCP streams and interfaces to the IP layer.” Prelim.
`Resp. 25 (citing Ex. 1006, 522). According to Patent Owner, relying on its
`declarant, “a [person of ordinary skill in the art] would understand that the
`‘TCP entity’ is software executing on the host, not the network interface
`card.” Id. (citing Ex. 2001 ¶¶ 52–53). As to Petitioner’s quote from
`Tanenbaum that a transport entity can be on the network interface card (Pet.
`43–44 (citing Ex. 1006, 480)), Patent Owner argues this is from a section 40
`pages before the discussion of the TCP entity, which is only ever described
`as being on the host or kernel. Prelim. Resp. 26.
`
`13
`
`
`
`IPR2017-01406
`Patent 7,673,072 B2
`
`As an initial matter, whether or not Tanenbaum discloses the “TCP
`entity” on “the host, and not the network interface card,” as Patent Owner
`contends, is beside the point, because Petitioner is relying on Erickson’s I/O
`device adapter as modified to use the fast path TCP processing disclosed in
`Tanenbaum. See In Keller, 642 F.2d 413, 425–26 (CCPA 1981) (“[O]ne
`cannot show non-obviousness by attacking references individually where, as
`here, the rejections are based on combinations of references.”). As
`discussed above, Petitioner relies on Tanenbaum’s disclosure that
`segmenting (i.e., “dividing . . . into segments”) is performed by the TCP
`entity and on its declarant’s testimony that a person of ordinary skill in the
`art would have understood Tanenbaum to suggest that the entity could have
`been on the network device. See Pet. 43–45. Relying on its declarant,
`Petitioner provides evidence and explanation as to how a person of ordinary
`skill in the art would have modified Erickson’s UDP scripts executing on the
`I/O device adapter into TCP scripts that would perform the claimed dividing.
`Id. Patent Owner does not point out deficiencies in this analysis, which we
`find sufficient for purposes of this Decision.
`Second, Patent Owner contends that, not only does Tanenbaum not
`disclose an interface device being able to divide data into segments, it
`actually teaches away from an interface having such a separate processor.
`Prelim. Resp. 23–24, 26–27. Specifically, Patent Owner relies on the
`following excerpt from Tanenbaum.
`
`A tempting way to go fast is to build fast network interfaces in
`hardware. The difficulty with this strategy is that unless the
`protocol is exceedingly simple, hardware just means a plug-in
`board with a second CPU and its own program. To avoid having
`the network coprocessor be as expensive as the main CPU, it is
`often a slower chip. The consequence of this design is that much
`
`14
`
`
`
`IPR2017-01406
`Patent 7,673,072 B2
`
`of the time the main (fast) CPU is idle waiting for the second
`(slow) CPU to do the critical work. It is a myth to think that the
`main CPU has other work to do while waiting. Furthermore,
`when two general-purpose CPUs communicate, race conditions
`can occur, so elaborate protocols are needed between the two
`processors to synchronize them correctly. Usually, the best
`approach is to make the protocols simple and have the main
`CPU do the work.
`Id. at 24 (quoting Ex. 1006, 588–589).
`A reference can be said to teach away “when a person of ordinary
`skill, upon reading the reference, would be discouraged from following the
`path set out in the reference, or would be led in a direction divergent from
`the path that was taken by the applicant.” Galderma Labs., L.P. v. Tolmar,
`Inc., 737 F.3d 731, 738 (Fed. Cir. 2013). However, “[a] reference that
`‘merely expresses a general preference for an alternative invention but does
`not criticize, discredit, or otherwise discourage investigation into’ the
`claimed invention does not teach away.” Meiresonne v. Google, Inc., 849
`F.3d 1379, 1382 (Fed. Cir. 2017) (quoting Galderma, 737 F.3d at 738).
`Here, for reasons discussed above, we agree with Petitioner that
`Erickson discloses a two-processor solution for performing UDP processing
`on the I/O device adapter and explicitly teaches that TCP scripts could also
`be written to do the same. See Ex. 1005, 5:36–51. Petitioner has provided
`testimony from its declarant, which we credit for purposes of this Decision,
`that a person of ordinary skill in the art would have motivated to implement
`Tanenbaum’s explicit teachings of fast path TCP processing using the two-
`processor solution of Erickson. See Ex. 1003 ¶¶ 137–148. Although
`Tanenbaum suggests that “offloading” using a two-processor solution may
`not be useful and introduces some complexities, Erickson provides
`motivation to move some protocol processing onto the I/O device adapter.
`
`15
`
`
`
`IPR2017-01406
`Patent 7,673,072 B2
`
`See Ex. 1005, 1:62–2:11. Patent Owner cites nothing in Tanenbaum to
`suggest that Erickson’s objectives would be undermined by using the fast
`path TCP processing disclosed in Tanenbaum or render Erickson inoperative
`for its intended purpose. See Meiresonne, 849 F.3d at 1383–84 (finding
`nothing to indicate the modification would detract from the goal of the
`primary reference). Accordingly, at this stage of the proceeding, we
`disagree that Tanenbaum teaches away from the proposed combination of
`Erickson and Tanenbaum.
`Third, Patent Owner contends the goal of Erickson is to reduce
`memory-to-memory copies and calls to the operating system, which
`Erickson accomplishes by mapping I/O device memory directly into the
`virtual address space of the user process. Prelim. Resp. 27–28 (citing Ex.
`1005, 2:63–66, 5:27–31, Fig. 2). Based on this characterization, Patent
`Owner concludes “[t]his suggests that the I/O adapter does not have
`processing power, but only provides working memory for the CPU.” Id. at
`28.
`
`We do not disagree with Patent Owner’s characterization of
`Erickson’s objective. However, it does not follow that the adapter only
`provides working memory and does not have its own processing power with
`which to perform the recited dividing the data into segments. As Petitioner
`points out, Erickson discloses the I/O device adapter executing a protocol
`script (e.g., udpscript). Pet. 33 (citing 7:51). For purposes of this Decision,
`we agree this supports Petitioner’s position that the I/O device adapter is
`capable of supporting a TCP-based script as Petitioner contends. See also
`Ex. 1005, 7:41–44 (“The user process provides the starting address and the
`length . . . and then ‘spanks’ a GO register to trigger the I/O device
`
`16
`
`
`
`IPR2017-01406
`Patent 7,673,072 B2
`
`adapter’s execution of a predetermined script.” (emphasis added)), 8:54–57
`(“The bus controller then transfers the data to the I/O device adapter and
`initiates the registers of the I/O device adapter to execute a predetermined
`script to process the data.”).
`Fourth, and finally, Patent Owner argues that all of the obvious scripts
`proposed by Petitioner require “a new ‘repetitive’ feature absent in the
`disclosure of Erickson itself,” that of repeating transfer of the segmented
`data until all of the data was sent. See Prelim. Resp. 29–30. Patent Owner
`argues that,
`all the scripts disclosed in Erickson are one-time scripts.
`Petitioner’s expert alleges that the script “senduserdatagram”
`disclosed at Erickson 7:23-32 will be “repeatedly spank[ed].”
`Ex. 1003, Appendix A, A-12. This allegation is technically
`wrong. Neither of the two while loops in the script includes
`anything in their braces. Ex. 2001, ¶ 54. The spanking statement
`“vhr_p->GO =1” will run at most once when the script is called.
`Id. Similarly, in the script disclosed at Erickson 7:50-64, there
`is no statement that could result in repeating execution, such as
`while loop, for statement, or if clause. Ex. 2001, ¶ 55.
`
`Petition[er]’s expert fails to provide any facts or analysis
`as to why a new repetitive feature should be added to the script
`or I/O adapter, especially when such feature is inconsistent will
`the mechanisms and embodiments disclosed in Erickson itself.
`Prelim. Resp. 30.
`We disagree with Patent Owner’s argument because it is directed at
`Erickson individually and not the proposed combination. See Keller, 642
`F.2d at 425–26. Moreover, contrary to Patent Owner’s argument,
`Petitioner’s declarant does provide facts and analysis as to why a person of
`ordinary skill in the art would have been motivated to modify Erickson’s
`scripts to add the repetitive feature. For example, Dr. Horst cites
`Tanenbaum’s teaching of breaking TCP data streams into pieces not to
`
`17
`
`
`
`IPR2017-01406
`Patent 7,673,072 B2
`
`exceed 64K (or even 1500 bytes) and sending as separate datagram and
`other evidence of TCP segmentation implemented as repeated invocations of
`a loop. See Ex. 1003, Appendix A, A-12–A-14 n. 9. Although Patent
`Owner’s declarant disputes this analysis, that testimony is based solely on
`Erickson’s disclosure and does not address the script as modified in view of
`Tanenbaum proposed by Petitioner and Dr. Horst. See Ex. 2003 ¶¶ 54–55.
`Accordingly, we credit Dr. Horst’s testimony for purposes of this Decision.
`“creating headers for the segments, by the interface device, from
`a template header containing the protocol header information”
`Petitioner contends Erickson in combination with Tanenbaum
`discloses this limitation. Pet. 47. Specifically, Petitioner relies on
`Erickson’s description of a UDP script running on the I/O device adapter
`“creating UDP/IP/MAC headers for datagrams by filling in protocol
`template header 702 shown in Fig. 7,” and its disclosure of other protocol
`scripts including TCP/IP. Id. (citing Ex. 1005, 5:47–51). Similar to its
`contentions discussed above, Petitioner contends it would have been obvious
`to a person of ordinary skill in the art to create a TCP version of the script
`described in Erickson that exchanges data in the form of segments based on
`Tanenbaum’s TCP prototype header. Id. at 47–48. Figure 6-50 of
`Tanenbaum is reproduced below:
`
`18
`
`
`
`IPR2017-01406
`Patent 7,673,072 B2
`
`
`
`Figure 6-50 of Tanenbaum depicts a TCP and IP header with
`shaded areas indicating no changes during
`one-way data transmission.
`Ex. 1006, 566. Petitioner contends that in the proposed combination, “the
`modified script would fill in the TCP Sequence number, Length and
`Checksum.” Pet. 48; Ex. 1003 ¶ 40, Appendix A, 1.4.
`“and prepending the headers to the segments to form transmit
`packets.”
`Petitioner contends Erickson in combination with Tanenbaum
`discloses this limitation. Pet. 49–50. In particular, Petitioner relies on
`Erickson’s description of UDP datagram template 702 without user data
`residing in the I/O device adapter’s memory. Id. (citing Ex. 1005, 7:39–47).
`Petitioner contends a person of ordinary skill in the art would have
`recognized that one of two obvious ways to concatenate the user data and the
`header (such as the described UDP header or TCP header in Tanenbaum)
`would have been to prepend the header to the segmented user data in the
`system of Erickson as modified by Tanenbaum. Id. at 50; Ex. 1003 ¶¶ 36–
`37, Appendix A, 1.6. For purposes of this Decision, we find that Petitioner’s
`contentions are supported sufficiently.
`For the foregoing reasons, we are persuaded that Petitioner has
`demonstrated a reasonable likelihood of prevailing on its proposed grounds
`
`19
`
`
`
`IPR2017-01406
`Patent 7,673,072 B2
`
`of obviousness for claim 1. At this stage of the proceeding, aside from the
`foregoing arguments, Patent Owner does not dispute Petitioner’s analysis
`with respect to claims 2–21.
`
`III. CONCLUSION
`For the foregoing reasons, we determine that the information
`presented establishes a reasonable likelihood that Petitioner would prevail in
`showing that at least one of claims 1–21 of the ’0