`
`Petition for Inter Partes Review
`
`Attorney Docket No.:
`
`47415.430
`
`Customer No.:
`
`______
`
`Real Parties in Interest: Dell Inc.,
`Hewlett-Packard Company, and NetApp,
`Inc.
`
`§§§§§§§§§§§
`
`In re patent of: Baek et al.
`
`U.S. Patent No. 6,978,346
`
`Issued: December 20, 2005
`
`Title: APPARATUS FOR
`REDUNDANT INTER-
`CONNECTION
`BETWEEN MULTIPLE
`HOSTS AND RAID
`
`Second Declaration of Dr. M. Ray Mercer
`Under 37 C.F.R. § 1.68
`
`I, Dr. M. Ray Mercer, do hereby declare:
`
`1.
`
`I am making this declaration at the request of Dell Inc., Hewlett-
`
`Packard Company, and NetApp, Inc. in the matter of the Inter Partes Review of
`
`U.S. Patent No 6,978,346 (“the ‘346 Patent”) to Baek et al. I am being
`
`compensated for my work in this matter. My compensation in no way depends
`
`upon the outcome of this proceeding.
`
`2.
`
`I provide my qualifications and professional experience in paragraphs
`
`5-11 of my previous declaration in this proceeding.
`
`–1–
`
`DHPN-1012
`
`
`
`3.
`
`(1)
`
`(2)
`
`In the preparation of this declaration, I have studied:
`
`The documents listed in paragraph 3 of my previous declaration;
`
`The Patent Owner’s Reply and all exhibits thereto, including the
`
`Declaration of Dr. Thomas Conte (hereinafter, the “Conte
`
`Declaration”);
`
`(3)
`
`Peter Chen et al., RAID: High-Performance, Reliable Secondary
`
`Storage, submitted to ACM Computing Surveys, October 29, 1993,
`
`DHPN-1011;
`
`(4)
`
`The Authoritative Dictionary of IEEE Standards Terms, 7th Ed., 2000
`
`definition of “network interface controller”
`
`(5) Microsoft Computer Dictionary, 4th Edition, definition of “network
`
`interface card”
`
`(6)
`
`Conte’s Deposition transcript
`
`4.
`
`(1)
`
`(2)
`
`In forming the opinions expressed below, I have considered:
`
`The documents listed above,
`
`The relevant legal standards, including the standard for obviousness
`
`provided in KSR International Co. v. Teleflex, Inc., 550 U.S. 398 (2007) and
`
`any additional authoritative documents as cited in the body of this
`
`–2–
`
`DHPN-1012
`
`
`
`declaration or my previous declaration in this proceeding, and
`
`(3) My knowledge and experience based upon my work in this area as
`
`described below.
`
`Interpretation of Claims
`
`RAID
`
`5.
`
`I disagree with Dr. Conte’s definition of RAID in paragraphs 37 and
`
`38 of the Conte declaration. Dr. Conte attempts to narrow the term RAID beyond
`
`what is appropriate for the broadest reasonable interpretation. Specifically, Dr.
`
`Conte asserts that a RAID would be understood to be a “single logical unit,”
`
`whereas I note that not all definitions of the term RAID require a single logical
`
`unit. I refer back to my first declaration at paragraphs 42 – 45, where I provide at
`
`least two independent definitions of RAID and note that different definitions of
`
`RAID have been propounded.
`
`6.
`
`Particularly, I note that the definition from the reference Weygant,
`
`which I cited in my first declaration, starts off by noting that RAID “is an acronym
`
`for redundant array of inexpensive disks.” It also states that, “a RAID device
`
`consists of a group of disks that can be configured in many ways, either as a single
`
`unit or in various combinations of striped and mirrored configurations.” The
`
`Weygant definition of the term RAID is consistent with the board’s definition – a
`
`–3–
`
`DHPN-1012
`
`
`
`redundant array of inexpensive disks. The Weygant definition is inconsistent with
`
`Dr. Conte’s definition because, on its face, it indicates that a group of disks do not
`
`have to be configured as a single unit. Weygant indicates that Dr. Conte’s
`
`definition of the term RAID includes an unnecessary limitation and is therefore too
`
`narrow.
`
`7.
`
`Patent Owner (PO) is incorrect to argue in its response that Hathorn
`
`does not include a RAID. For instance, Hathorn teaches a shadowing operation for
`
`data. A person of ordinary skill in the art would have understood that shadowing
`
`and mirroring are essentially the same concept. For example, Chen states: “The
`
`traditional solution, called mirroring or shadowing, uses twice as many disks as a
`
`nonredundant disk array [Bitton88].” Emphasis added. See, Peter Chen et al.,
`
`“RAID: High-Performance, Reliable Secondary Storage,” submitted to ACM
`
`Computing Surveys, October 29, 1992, Section 3.2.2 titled Mirrored (RAID Level
`
`1), page 10. Weygant indicates that mirroring is a configuration of RAID (RAID
`
`level 1), as I noted in paragraph 44 of my first declaration. Therefore, the act of
`
`“shadowing” taught in Hathorn should be understood to lie within the broadest
`
`reasonable interpretation of the term RAID.
`
`8.
`
`For the reasons above I agree with the board’s definition of the term
`
`RAID and disagree with Dr. Conte’s definition.
`
`RAID Controlling Unit
`
`–4–
`
`DHPN-1012
`
`
`
`9.
`
`I disagree with Dr. Conte’s definition of the term “RAID controlling
`
`unit” at paragraphs 39 – 41 of the Conte declaration. Dr. Conte did not address the
`
`file history or the specification in his discussion of the term RAID controlling unit,
`
`and he failed to note that there is nothing in the file history or the specification that
`
`would indicate that the term should be limited beyond the board’s definition of
`
`RAID controlling unit (“a component that controls operation of a RAID”). By
`
`contrast, I did discuss the specification in my first declaration at paragraphs 39 and
`
`40 and observed that the term “RAID controlling unit” does not appear in the
`
`specification and therefore is neither defined nor narrowed in the specification.
`
`10.
`
`Furthermore, Dr. Conte’s technical arguments regarding RAID
`
`controlling unit are not correct. For instance, I noted above that Dr. Conte states
`
`that a RAID must be a single logical unit; however that is not a required definition
`
`or feature, as evidenced by Weygant. Also, as evidenced by Chen,
`
`shadowing/mirroring is included within the broad concept of RAID storage, and
`
`the breadth of “RAID” and “RAID controlling unit” must include
`
`shadowing/mirroring in any event. Dr. Conte bases his definition of “RAID
`
`controlling unit” on his argument regarding a single logical unit, and this definition
`
`(as confirmed by the board’s construction) is too narrow.
`
`Network [Interface] Controlling Unit
`
`11.
`
`I also disagree with Dr. Conte’s definition of the terms, “network
`
`–5–
`
`DHPN-1012
`
`
`
`controlling unit” and “network interface controlling unit.” I discussed these terms
`
`at paragraphs 22 and 23 of my first declaration, where I note that the broadest
`
`reasonable interpretation of those terms includes any component allowing a device
`
`to communicate over a network such as a fiber channel network, an asynchronous
`
`transfer mode (ATM) network, or other network. The terms used in the claims are
`
`not found in the specification nor otherwise indicated that they should be
`
`narrowed. In contrast Dr. Conte cites actual marketed products called “card,”
`
`“server adapter,” “and “networking card,” but nowhere does Dr. Conte actually
`
`link those products in a way that logically narrows the terms “network controlling
`
`unit” and “network interface controlling unit.” In short, Dr. Conte fails to explain
`
`why the terms used in the claims would necessarily be limited to the specific
`
`hardware components that he shows as examples.
`
`12.
`
`In response to Dr. Conte, I cite to the IEEE definition of “network
`
`interface controller.” The Authoritative Dictionary of IEEE Standards Terms, 7th
`
`Ed., 2000 defines “network interface controller” as “a communication device
`
`which permits the connection of information processing devices to a network.” I
`
`also cite to the Microsoft Computer Dictionary, 4th Edition at its definition of
`
`“network interface card.” Microsoft defines “network interface card” as “An
`
`expansion card or other device used to provide network access to a computer or
`
`other device, such as a printer. Network interface cards mediate between the
`
`–6–
`
`DHPN-1012
`
`
`
`computer and the physical media, such as cabling, over which transmissions travel.
`
`Acronym: NIC. Also called network adapter, network card.”
`
`13.
`
`The dictionaries I cite above show that the word “card” or the word
`
`“adapter” are used to denote particular hardware, whereas the omission of those
`
`words indicates that a term is not limited to specific hardware. In the present case,
`
`“network controlling unit” and “network interface controlling unit” do not include
`
`words denoting particular hardware.
`
`14.
`
`I further disagree with the patent owner's statement: "The controllers
`
`provide two ports—one for transmission and one for reception” at page 16 of the
`
`Response. It is my opinion that the patent owner mischaracterizes the term “port”
`
`as used in the ‘346 patent The ‘346 patent never refers to the RX and TX parts of a
`
`NETWORK INTERFACE CONTROLLER (such as those found with respect to
`
`elements 470, 471, 480, and 481) as “ports.” Instead, the ‘346 specification refers
`
`only to (1) “hub ports” (such as 420 to 424, 430 to 434, 422,423, 432, and 433 –
`
`see ‘346:3:19, 3:45 & 47), and (2) “switch port 611,” (see ‘346:4:54 - 55). Note
`
`that in Figure 6 of the ‘346 patent, boxes labeled RX and TX are shown as part of
`
`element 611 (the identified a “port”). Thus, the patent owner mischaracterizes the
`
`TX and RX elements of the NETWORK INTERFACE CONTROLLERS in order
`
`to support their erroneous contention: “At the time of the '346 Patent, as is true
`
`today, a Network Controller may include multiple of these 'ports' . . . Accordingly,
`
`–7–
`
`DHPN-1012
`
`
`
`a network controlling unit should be interpreted as ‘a controller that supplies
`
`communication functionality when attached to a computer network and including
`
`one or more ports’ . . . It would be unduly limiting to require only one port,
`
`especially where, as here, the specification discloses multi-port network
`
`controllers.” See pages 16 and 17 of the Response.
`
`15.
`
`I believe that Dr. Conte is incorrectly limiting the terms “network
`
`controlling unit” and “network interface controlling unit” to specific hardware
`
`cards. By contrast, the board is correct to apply the plain and ordinary meaning to
`
`those terms, which is consistent with the definition I proposed in my first
`
`declaration (“a component allowing a device to communicate over a network, e.g.,
`
`Fibre Channel, ATM, or other networks”).
`
`16.
`
`It is also my opinion that the patent owner mischaracterizes the term
`
`“port” as used in the ‘346 patent for another reason. In fact, the term “port” is used
`
`to identify fibre channel components in the ‘346 patent.
`
`Connected
`
`17.
`
`I take exception to PO’s accusation that I defined “connected” to
`
`require direct connection without an intermediary. See PO response at p. 50-51.
`
`Such an accusation is misleading and without any basis. In my first declaration I
`
`did define “coupled,” but I did not define “connected,” and I certainly did not say
`
`–8–
`
`DHPN-1012
`
`
`
`that “connected” is limited to a direct connection.
`
`18.
`
`In response to the PO’s accusation, I note that the word “connected” is
`
`much more tricky to define in the ‘346 patent than is the word “coupled” because
`
`the meaning of the term “connected” depends on context. That is because
`
`“connected” is sometimes used in the claims and in the specification to indicate an
`
`indirect connection. For instance. The ‘346 patent at column 3, lines 31-38, states
`
`“network interface controllers… are connected with one another by two networks
`
`through two hubs 440, 441 . . .” (Emphasis added.) Such use of the term
`
`“connected” specifies an indirect connection between the network interface
`
`controllers via hubs. Another instance is in claim 1, which states “a plurality of
`
`connection units for connecting the first RAID controlling units and the second
`
`RAID controlling unit to the numerous host computers.” (Emphasis added.) Once
`
`again, connecting is used to refer to an indirect connection that includes some kind
`
`of intermediate component, such as connection units.
`
`19.
`
`Therefore, in response to PO’s accusation, I note that a person of
`
`ordinary skill in the art would read the term “connected” or “connecting” in the
`
`‘346 patent not to exclude the use of hubs or switches as intermediate components.
`
`This observation is based on both the specification and the claim language.
`
`20.
`
`Further, in response to PO’s accusation, I see that the ‘346 patent uses
`
`the term “directly” in a way that includes a connection through an intermediate
`
`–9–
`
`DHPN-1012
`
`
`
`unit. Claim 1 recites “where in the first RAID controlling unit and the second
`
`RAID controlling unit directly exchange information with the numerous host
`
`computers through the plurality of connecting units.” So when PO asserts that I am
`
`requiring “direct connection,” this contention is simply not accurate.
`
`Hathorn Anticipates Claims 1-3 and 5-8 of the ‘346 Patent
`
`21.
`
`In response to Dr. Conte’s declaration, it appears that there is no
`
`disagreement as to where at least one set of the RAID controlling units exist in
`
`Hathorn’s Figure 3 embodiment. For instance at paragraphs 56 - 59, Dr. Conte
`
`discusses whether the mirroring provided in Hathorn discloses a RAID. Dr. Conte
`
`provides the following annotation of Hathorn’s Figure 3:
`
`–10–
`
`DHPN-1012
`
`
`
`Illustration From
`Conte Declaration,
`p. 32
`
`22. Dr. Conte then discusses the above drawing at paragraph 57 of his
`
`declaration and he notes, “making a mirrored copy of data is the function of a
`
`–11–
`
`DHPN-1012
`
`
`
`RAID Controlling Unit… For example, if host 301 writes to the DASDs 326, the
`
`storage controller 325 would initiate a remote write through switch 305 and 315 to
`
`cause the data to me [sic] mirrored to the DASDs 336, where that remote write
`
`must go through the storage controller 335.”
`
`23.
`
`In response, I have marked up Figure 3 of Hathorn to show the
`
`different components of the claims of the ‘346 patent. This is consistent with my
`
`analysis from my first declaration, where I noted that DASDs 323, 326, 333, 336,
`
`are a RAID used for mirroring. (See my original declaration at pages 132-136,
`
`140-145, where I discuss Hathorn, especially at 9:29-51.)
`
`–12–
`
`DHPN-1012
`
`
`
`‘950 Patent,
`Fig. 3
`(annotated)
`
`DASDs
`326, 336
`are a RAID
`
`Storage
`controllers 325,
`335 are RAID
`controlling units
`to control
`DASDs 326, 336
`to save mirrored
`data as a RAID
`
`24. Dr. Conte is incorrect in asserting that only one RAID controlling unit
`
`is disclosed in Hathorn. Instead, Hathorn discloses multiple RAID controlling
`
`units, where the passage at 9:29-51 of Hathorn is illustrative. For instance, the
`
`storage controllers 325 and 335 of Hathorn are each RAID controlling units
`
`because they each control some part of the mirroring operation of a RAID. In that
`
`–13–
`
`DHPN-1012
`
`
`
`particular example, Hathorn’s storage controller 325 controls operation of a RAID
`
`by storing data to DASD 326 and initiating a remote write that causes data to be
`
`stored to DASD 336. Storage controller 335 controls operation of a RAID by
`
`cooperating with controller 325 to store the data to DASD 336. Once again, this is
`
`consistent with the operation that I referenced on pages 131 and 143 of my first
`
`declaration, where I cite Hathorn at 9:29-51, inter alia.
`
`25.
`
`In my original declaration, I noted that the ports A, B, C, and D of
`
`Hathorn satisfy the claimed “network [interface] controlling units” because the
`
`ports of Hathorn control communication of the storage controllers on the network.
`
`And I provided the following annotated version of Hathorn’s Figure 3:
`
`–14–
`
`DHPN-1012
`
`
`
`First and
`second
`network
`controlling
`units
`
`Third and
`fourth
`network
`controlling
`units
`
`26. As explained above, Hathorn’s ESCON ports are described as being
`
`dynamically set to communicate either as a channel or control unit link level
`
`facility (e.g., at ‘950:7:57-8:15), and other examples within Hathorn describe
`
`communication generally with respect to figures 4-8. Hathorn’s ports are
`
`communication devices permitting the connection of the storage controllers to the
`
`network, where the network includes among other things dynamic switches 305
`
`and 315. Therefore, Hathorn’s ports control communication by the storage
`
`controllers over the network.
`
`–15–
`
`DHPN-1012
`
`
`
`Declaration
`
`27.
`
`I declare that all statements made herein on my own knowledge are
`
`true and that all statements made on information and belief are believed to be true,
`
`and further, that these statements were made with the knowledge that willful false
`
`statements and the like so made are punishable by fine or imprisonment, or both,
`
`under Section 1001 of Title 18 of the United States Code.
`
`Executed5~{1/{Jer /~ :2---D/ Y..
`
`By: JPoD~a(z:'l~
`
`-16-
`
`![)HPN-1012J
`
`