`IPR2014-01207, Paper No. 77
`IPR2014-01209, Paper No. 76
`November 30, 2015
`
`trials@uspto.gov
`
`
`571-272-7822
`
`
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`____________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`____________
`
`ORACLE CORP., NETAPP INC., HUAWEI TECHNOLOGIES
`CO., LTD., and DOT HILL SYSTEMS CORPORATION,
`Petitioners,
`
`v.
`
`CROSSROADS SYSTEMS, INC.,
`Patent Owner.
`____________
`
`Cases: IPR2014-01197, IPR2014-01207, IPR2014-01209
`Patents: 6,425,035 B2, 7,051,147 B2
`____________
`
`Held: October 30, 2015
`____________
`
`
`
`BEFORE: NEIL T. POWELL, KRISTINA M. KALAN, J. JOHN
`LEE, and KEVIN W. CHERRY, Administrative Patent Judges.
`
`
`
`
`
` The above-entitled matter came on for hearing on Friday, October
`30, 2015, commencing at 1:00 p.m., at the U.S. Patent and
`Trademark Office, 600 Dulany Street, 9th Floor, Hearing Room
`D, Alexandria, Virginia.
`
`
`
`Cases: IPR2014-01197, IPR2014-01207, IPR2014-01209
`Patents: 6,425,035 B2, 7,051,147 B2
`
`APPEARANCES:
`
`ON BEHALF OF THE PETITIONERS:
`
`
`
`
`
`
`
`
`
`and
`
`ON BEHALF OF THE PATENT OWNER:
`
`
`
`JARED BOBROW, ESQUIRE
`AARON Y. HUANG, ESQUIRE
`Weil, Gotshal & Manges, LLP
`201 Redwood Shores Parkway
`Redwood Shores, California 94065-1134
`
`GREG H. GARDELLA, ESQUIRE
`Oblon, McClelland, Maier & Neustadt, LLP
`1940 Duke Street
`Alexandria, Virginia 22314
`
`KEITH A. RUTHERFORD, ESQUIRE
`JAMES H. HALL, ESQUIRE
`Blank & Rome, LLP
`717 Texas Avenue
`Suite 1400
`Houston, Texas 77002
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
` 2
`
`
`
`
`
`
`
`Cases: IPR2014-01197, IPR2014-01207, IPR2014-01209
`Patents: 6,425,035 B2, 7,051,147 B2
`
`
`P R O C E E D I N G S
`- - - - -
`JUDGE POWELL: Good afternoon. This is the oral
`hearing for three related cases, IPR2014-01197 which involves
`patent 6,425,035 B2; IPR2014-01207 which involves patent
`7,051,147 B2; and IPR2014-01209 which involves the same
`patent. IPR2015-00822 has been joined with IPR2014-01197.
`In the hearing room with me I have Judge Lee and
`Judge Cherry and joining us via video from Denver, we have
`Judge Kalan. Can counsel please state your names for the record.
`MR. GARDELLA: Good afternoon, Your Honors.
`Greg Gardella on behalf of petitioners, Oracle and NetApp. I'm
`joined today by Jared Bobrow and Aaron Huang of the Weil
`Gotshal firm. Mr. Bobrow will be doing the presentation today.
`And with your permission, I'll approach with your copies of the
`demonstratives.
`JUDGE POWELL: Sure.
`MR. RUTHERFORD: Keith Rutherford again for the
`patent owner, Crossroads. Also in attendance is Steve Sprinkle,
`lead counsel. With me at counsel table, James Hall, and we have
`in the audience our CEO and COO of Crossroads. And may we
`also approach with our demonstratives?
`JUDGE POWELL: Sure. Okay. I'll start by addressing
`the joint list of objections to demonstrative exhibits. We have
`reviewed and considered them and we are going to allow both
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
` 3
`
`
`
`
`
`
`
`Cases: IPR2014-01197, IPR2014-01207, IPR2014-01209
`Patents: 6,425,035 B2, 7,051,147 B2
`
`parties to use any of their demonstrative exhibits in today's
`hearing. We will carefully consider which evidence and
`arguments have been properly presented as we prepare our final
`decisions in these cases.
`Consistent with the hearing order, each party will have
`60 minutes. Petitioners will present their case first and may
`reserve time for rebuttal. And then patent owner will respond to
`petitioner's presentation after which petitioners may use any
`remaining time to respond to patent owner's presentation.
`During your presentations you must identify each
`demonstrative exhibit clearly and specifically. You can refer to it
`by slide number or screen number or however else is appropriate.
`That's particularly important because Judge Kalan cannot see the
`projection screen here in the hearing room.
`And with that, do we have any questions before we
`start? Well, I'll yield the floor to petitioners.
`MR. BOBROW: Good afternoon, Your Honors. Jared
`Bobrow for the petitioners. With the Board's permission, I would
`like to reserve 25 minutes.
`JUDGE POWELL: Okay.
`MR. BOBROW: Thank you. Starting with slide 2 of
`our slide presentation, set forth there are the claims on which this
`trial is proceeding and the three obviousness combinations that
`are at issue. The CRD-5500 combination, the Bergsten/Hirai
`combination and the Kikuchi/Bergsten combination.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
` 4
`
`
`
`
`
`
`
`Cases: IPR2014-01197, IPR2014-01207, IPR2014-01209
`Patents: 6,425,035 B2, 7,051,147 B2
`
`
`Let me proceed first with the CRD-5500 combination, if
`I may. As the Board heard this morning, the CRD-5500 manual
`describes a storage system. That storage system includes a
`multiplicity of hosts on one side. It includes a transport medium.
`We have cables going to channels. We then have those channels,
`the I/O ports as part of the CRD-5500. And on the other side of
`that storage router, what we have are storage devices. The
`CRD-5500 manual also describes the functions of mapping and
`describes the function of access control resident in the
`CRD-5500.
`The only element missing from the CRD-5500 is an
`express statement of the use of fibre channel. That's the only
`thing we submit that is missing from the CRD-5500. The
`CRD-5500 data sheet, which is Exhibit 1004, states that the 5500
`itself was designed to include and be adapted for fibre channel
`functionality. So there's no dispute that the cards and the slots of
`the CRD-5500 were adapted for a fibre channel use.
`To provide the fibre channel element, we provide the
`Smith reference which discloses the Tachyon chip. And the
`Tachyon chip, it discloses, is used for SCSI encapsulation over
`fibre channel. The idea of the combination is including the
`Tachyon chip on the card of the CRD-5500 so that, as shown on
`slide 6 of our presentation, we have hosts, each one connected
`through the Tachyon I/O card that's in the CRD-5500 and we
`have storage on the other end.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
` 5
`
`
`
`
`
`
`
`Cases: IPR2014-01197, IPR2014-01207, IPR2014-01209
`Patents: 6,425,035 B2, 7,051,147 B2
`
`
`So that is the combination, that use. And the only
`argument, primary argument, I should say, on which the patent
`owner responds is to say that the combination is somehow
`deficient because the mapping in their view does not include a
`map of the host in some particular concrete, nearly immutable
`way, but instead only maps to channel ID.
`And let me address that in a couple of ways, if I may.
`The first way is as a matter of claim construction. Neither party
`proposed an express construction for the Board to adopt as it
`relates to this concept of mapping. But the District Court below
`certainly received from the patent owner and adopted at the
`patent owner's urging a construction of mapping that is quite
`broad in our view and certainly is -- at least meets the broadest
`reasonable interpretation standard that is applicable here. What
`the patent owner said in District Court was that a map need only
`contain, quote, a representation of devices on each side of the
`storage router. Merely a representation.
`And if we can please call up the Levy declaration, the
`patent owner's expert, at page 27 in Exhibit 2053, this is the
`patent owner's expert's declaration. It sets forth the construction
`that the patent owner urged the District Court to adopt and which
`the District Court, in fact, adopted.
`And the breadth of this construction is substantially
`broader at District Court than the construction being urged now
`by the patent owner. Mapping, according to the patent owner in
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
` 6
`
`
`
`
`
`
`
`Cases: IPR2014-01197, IPR2014-01207, IPR2014-01209
`Patents: 6,425,035 B2, 7,051,147 B2
`
`the District Court proceedings, simply is to create a path from a
`device on one side of the storage router to a device on the other
`side of the storage router, merely creating a path.
`A map contains a representation of devices on each side
`of the router, a representation. Not a specific identity, but a
`representation of the device. And it has a function. And the
`function of this representation is so that when a device on one
`side of the router, the storage router, wants to communicate with
`a device on the other side of the storage router, the storage router
`can connect the devices. That is the function. That is what
`makes an identification in the map sufficient. Does it perform
`this connection so that a device on one side can communicate
`with a device on the other?
`That's the District Court construction under the Phillips
`standard. Certainly this construction or something broader would
`need to be applicable here. And certainly, I think it's important
`under 35 USC § 301 for the Board to consider this proposed
`construction because in part, what Section 301, I submit, was
`designed to do was to prevent what may be going on here, which
`is the patent owner urging a broad construction in District Court
`and then trying to avoid invalidity or unpatentability at the Board
`by urging a narrower construction. And certainly a narrower
`construction is being adopted here.
`If we can then go back to the slide presentation and turn
`to slide 9, so in the first instance, Your Honors, I submit that
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
` 7
`
`
`
`
`
`
`
`Cases: IPR2014-01197, IPR2014-01207, IPR2014-01209
`Patents: 6,425,035 B2, 7,051,147 B2
`
`under that broad construction there is no question but that the
`CRD-5500 meets that definition, that construction proposed by
`the patent owner below. It provides a representation of storage
`through the channel ID, number one. And number two, that
`channel is providing a pathway, and it is a specific pathway. And
`we can see that on slide 10.
`This is the embodiment on which the petitioners in this
`proceeding rely. And you can see that what the CRD-5500
`manual is stating quite expressly is that you can assign
`redundancy groups, meaning storage, to a particular host. So
`what the 5500 manual is saying is that you can use the channel ID
`in order to map to a particular host. So there's no question but
`that this channel ID is being used in the CRD-5500 and what it
`does is it maps to a particular host.
`What we can see in Figure 1-2 on slide 10 is that each
`host here is connected through a specific cable. In this case it
`will be a SCSI cable. And each one goes to a specific channel.
`Each one has its own channel. There's no mixing of the channels.
`Each host has its own channel. And as a result of having its own
`channel, the channel ID is in a one-to-one-to-one correspondence
`from the card to the SCSI cable to the host. And in that way that
`channel ID is not only mapping the channel ID, it is mapping the
`host. It's mapping the cable as well. It's mapping this entire
`pathway from the channel through this cable and to the host. And
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
` 8
`
`
`
`
`
`
`
`Cases: IPR2014-01197, IPR2014-01207, IPR2014-01209
`Patents: 6,425,035 B2, 7,051,147 B2
`
`in that way, that is how it is mapping to the particular device and
`maps the particular device.
`JUDGE CHERRY: Counsel, I just had a question. So I
`don't know if you were in the hearing room this morning. Were
`you?
`
`MR. BOBROW: I was.
`JUDGE CHERRY: And this morning the counsel
`articulated a combination where all four hosts were on a single
`fibre channel. Are you articulating a similar combination?
`MR. BOBROW: This is a different the combination.
`Our petition did not include that arbitrated loop that was being
`discussed. Our petition, and it's set forth at pages 13, 17, and 21,
`is relying on the Figure 1-2 embodiment. And that is the
`embodiment on which Professor Chase opined and determined
`that these were indeed obvious.
`JUDGE CHERRY: And you are not offering a theory
`based on the similar arbitrated loop physical address theory that
`was being offered by the morning petitioners?
`MR. BOBROW: That particular embodiment with the
`host on that arbitrated loop, we are not.
`JUDGE CHERRY: Okay.
`MR. BOBROW: So what we had, then, is this mapping
`from channel and we have the channel ID, and what the 5500
`says is that you are mapping to a particular host. On slides 12
`and 13, Your Honors looked at that this morning. I would like to
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
` 9
`
`
`
`
`
`
`
`Cases: IPR2014-01197, IPR2014-01207, IPR2014-01209
`Patents: 6,425,035 B2, 7,051,147 B2
`
`go to slide 14 because what slide 14 shows is that the patent
`owner's expert even acknowledges that in this case where you
`have one channel and one host for that one channel, that there
`will be a sufficient identification of the host for mapping
`purposes. And that's from Dr. Levy's deposition, Exhibit 1218.
`And what he was asked about was whether a fibre channel ID
`was sufficient to identify the host. And what his answer was, was
`that all that's required in the map is an identifier that's sufficient
`to distinguish between multiple hosts on the first transport
`medium. That is the function and that is what is sufficient for
`mapping is to be able to distinguish.
`And certainly what we have here is more than sufficient
`to distinguish. If we have a channel ID, ID 0, it is connected only
`to the first host on the left. That is the only connection that it has.
`And therefore, this ID is sufficient to distinguish the host
`connected to Channel 0 from the host that's connected to
`Channel 1.
`Now, in addition to this channel ID disclosure, in
`addition to that but as an alternative, our petition also sets forth
`another basis on which this combination renders the claims
`invalid and unpatentable. And that alternative ground is that
`Professor Chase opined and the petition sets forth at our petition
`page 18, if we can call that up, the additional theory and the
`additional reason that this combination renders unpatentable the
`claims is that Professor Chase also opined that the host identity
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
` 10
`
`
`
`
`
`
`
`Cases: IPR2014-01197, IPR2014-01207, IPR2014-01209
`Patents: 6,425,035 B2, 7,051,147 B2
`
`that can be used as part of the mapping can actually come from
`the fibre channel header.
`So recall that in this combination the Smith Tachyon
`chip is providing that fibre channel functionality. So the idea is
`that when the host is communicating, it is sending a fibre channel
`packet. The fibre channel packet is going to have host ID
`information in it and that host ID information is resident in the
`packet.
`
`JUDGE LEE: Is the fibre channel header in a map
`somewhere?
`MR. BOBROW: That is the alternative embodiment.
`And that is what Professor Chase opined in his declaration, is that
`the fibre channel packet address or the channel ID, either one
`could be used was his point. And that's what we set forth --
`JUDGE LEE: I understand the testimony that the fibre
`channel header could be used to identify a particular host, but the
`claims require that a configuration be maintained that maps
`between a device and remote storage. So where is that
`configuration, that mapping using fibre channel headers, where is
`that in the -- where is that described in the prior art?
`MR. BOBROW: In the petition where it's described is
`that the Tachyon chip would pass the host device identity to the
`CRD-5500 controller processor. And so what Professor Chase
`opined was that the host identity, either from the FCP header or
`from the channel host module, could be forwarded on.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
` 11
`
`
`
`
`
`
`
`Cases: IPR2014-01197, IPR2014-01207, IPR2014-01209
`Patents: 6,425,035 B2, 7,051,147 B2
`
`
`Judge Lee, I agree completely that what the CRD-5500
`manual says is that it is using that channel ID. That's the first
`embodiment that we are relying upon for unpatentability. We
`wanted to simply point out that Professor Chase has opined and
`our petition advances the additional theory that essentially the
`CRD could be modified once the Tachyon chip is added to the
`combination to read the FCP host ID that can be stored in
`memory and then can be passed on and used as part of the map.
`I agree that that is not how the map is structured in the
`CRD-5500 manual. Professor Chase opines that would be an
`adaptation that one of ordinary skill in the art could do and would
`be within the skill level of one of ordinary skill.
`Now, if we can go back to the slide presentation,
`please --
`JUDGE LEE: Actually, before we move on from this
`slide, just taking a look at this figure, which is Figure 1-2 of the
`CRD manual, Mr. Rutherford for Crossroads this morning, he
`argued that in the combination proffered, at least in the petitions
`we discussed this morning that the proffered combination is
`properly understood to correspond each of the channels, 0, 1, 2
`and 3 in this figure to be corresponding to the controller element
`in the claim. Is that a proper understanding of how you view this
`combination?
`MR. BOBROW: Well, no, it's not. May I explain.
`Each of these channels is certainly associated with that interface
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
` 12
`
`
`
`
`
`
`
`Cases: IPR2014-01197, IPR2014-01207, IPR2014-01209
`Patents: 6,425,035 B2, 7,051,147 B2
`
`card that we set forth earlier. There's the I/O card which would
`now include in the combination the Tachyon chip. It will have
`some sort of an identifier and it will have the channel ID.
`I think where we differ is that the channel ID, as the
`CRD manual says, maps to a particular device and it describes
`that in connection with this embodiment saying you can assign
`redundancy groups to a particular host, and you do that by the
`channel ID.
`JUDGE LEE: I guess as I understand Mr. Rutherford's
`argument, if you are mapping to a host channel and a host
`channel really collapses down to be just the I/O card, the SCSI
`I/O card, that really you are mapping to the controller. And the
`controller is an element in the claim. And if it meant to claim
`mapping between the remote storage and a controller would have
`said that instead of saying mapping between the devices and the
`remote storage. So I think there is some importance to what
`exactly in your view is the controller and what is the device or the
`representation of the device in your combination?
`MR. BOBROW: So the controller, as I mentioned a
`moment ago, does correspond to the card. That would be the
`controller. And the device is the host in this situation, in this
`embodiment. And the express teaching of the CRD-5500 is that
`by mapping the channel in this configuration, you are also, also
`mapping to the host.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
` 13
`
`
`
`
`
`
`
`Cases: IPR2014-01197, IPR2014-01207, IPR2014-01209
`Patents: 6,425,035 B2, 7,051,147 B2
`
`
`A single number can represent more than one thing. It
`needn't be this is exclusively mapping this; this is only mapping
`that. We submit that what the channel ID is doing is mapping
`multiple things. You could say that it's mapping the channel but
`it is also certainly mapping the host. And it does so because
`there's a one-for-one correspondence between channel and host as
`shown in Figure 1-2. So the single number is mapping multiple
`components. Not just one.
`With that, if there are no further questions on the
`CRD-5500, I would move to the second combination, and that
`begins at slide 19. This is the Bergsten/Hirai combination.
`Now, the Bergsten patent is a patent that once again
`teaches and discloses nearly every element of the claims here.
`We have multiple hosts. We have connections through SCSI and
`through fibre channel from hosts to what amounts to a storage
`router. It says that we map the host interface ID and block
`number to a logical device. So we have this virtualized storage
`mechanism where we are mapping host IDs to virtual logical
`storage. And then thereafter we are also mapping from this
`logical device ID, we are also mapping from there to the physical
`storage devices. And there are multiple storage devices, so-called
`MSDs or mass storage devices that are described in the Bergsten
`reference. So we clearly have a disclosure of mapping. We
`clearly have the disclosure of multiple hosts and multiple storage
`and we have virtualized storage.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
` 14
`
`
`
`
`
`
`
`Cases: IPR2014-01197, IPR2014-01207, IPR2014-01209
`Patents: 6,425,035 B2, 7,051,147 B2
`
`
`And the combination is with the Hirai patent
`application. And what the Hirai patent application provides is it
`builds on the suggestion in column 15 of the Bergsten patent.
`The Bergsten patent describes how write protection, which is a
`type of access write limitation to protect files that have -- so that
`they can't be written to or overwritten, what Bergsten is saying is
`that's a good thing. You may want to, in the context of the
`Bergsten invention, provide write protection.
`The Bergsten reference doesn't show that in the map. It
`doesn't go into the specifics of it. And that's why we combined
`that reference with the Hirai patent application, because what
`Hirai has is a storage system with multiple hosts, a controller,
`mass storage devices. And what we had in Hirai are access
`controls because what Hirai says is that you can have different
`computers, we'll call those hosts, and they can have access to
`different partitions of storage and their access rights can be
`restricted in various ways. One device might be given read rights
`to one partition and won't have read rights to another partition.
`So it's clearly disclosing access controls. And the combination
`merely takes this concept, this teaching of access rights, access
`controls and essentially combines that with the Bergsten map, the
`map which is taking that logical -- from host to logical to physical
`storage. So that's the combination.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`
` 15
`
`
`
`
`
`
`
`Cases: IPR2014-01197, IPR2014-01207, IPR2014-01209
`Patents: 6,425,035 B2, 7,051,147 B2
`
`
`JUDGE KALAN: What about patent owner's assertion
`that the access rights in Hirai are directed to high-level file
`system access permissions and not NLLBP?
`MR. BOBROW: On slide 22 of the presentation, we
`begin to address that. And turning to slide 23, certainly Bergsten
`discloses NLLBP, native low-level block protocols, I think that's
`unquestioned because it discloses, for example, SCSI. What
`Professor Chase opined in his declaration submitted with our
`petition is that one of ordinary skill would have adapted that
`Bergsten map in view of Hirai's teaching of access control.
`Professor Chase's opinion isn't tied to whether these are the file
`system level or the native low-level block level. It's not
`dependent upon that because the teaching of access controls is
`essentially a teaching of a logical, functional operation. And that
`table is going to teach one of ordinary skill in the art using
`Bergsten, which is already using native low-level block protocol,
`to simply add that functionality of checking to see whether the
`host actually has the right to go from host ID to logical storage to
`physical storage.
`So the form of the access control, whether it's high-level
`or low-level, is not the point of the combination. The
`combination is taking the teaching of Hirai and combining it with
`the native low-level block protocols of Bergsten.
`Now, essentially the same argument as I understood the
`patent owner's response is being made to the Bergsten/Hirai
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
` 16
`
`
`
`
`
`
`
`Cases: IPR2014-01197, IPR2014-01207, IPR2014-01209
`Patents: 6,425,035 B2, 7,051,147 B2
`
`combination and CRD where they say essentially the host ID
`that's in Bergsten is a host ID and it's not the host itself. But
`essentially we know from Figure 7 of Bergsten that Bergsten is
`specifically using a host interface identifier that identifies the
`particular host. And we know that the storage controller receives
`a host ID and a host memory block and then the controller
`determines whether the host ID and block number map exactly to
`the logical device. And that's at slide 26 of the presentation.
`So what this is saying in combination with Figure 7 is
`that Bergsten provides a host ID. That host ID is coming from
`the host. It's received by the controller and a determination is
`made at the logical level in that context. So that's what slide 26
`and Exhibit 1007 in Bergsten is expressly providing.
`JUDGE CHERRY: One question I had. It seems to me
`that in reviewing Dr. Chase's original testimony he didn't really
`recognize that there was or didn't feel there was a difference
`between Hirai -- the patent owner is articulating that Hirai is
`directed to high-level versus low-level. Did he provide any
`testimony about why a person of ordinary skill would look to
`high-level controls for these low-level operations?
`MR. BOBROW: Your Honor, I believe that what his
`declaration provides isn't a distinction between high-level or
`low-level as it pertained to the access rights because I believe that
`Professor Chase understood these access rights to be essentially
`an operation, a logical operation. And whether they are
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
` 17
`
`
`
`
`
`
`
`Cases: IPR2014-01197, IPR2014-01207, IPR2014-01209
`Patents: 6,425,035 B2, 7,051,147 B2
`
`implemented in native low-level or at a higher level wasn't the
`point of Hirai.
`Now, it is important for the invention that they be
`enabled at the native low-level block level because that's what the
`claim provides, but we know that Bergsten, that is the protocol
`that it is using. It is communicating using native low-level block
`protocols. So Professor Chase is saying this is a broad, high-level
`concept. Whether you implement it native low-level or higher
`level doesn't matter for purposes of whether the combination
`would work or whether one of ordinary skill would have reason
`to do the combination.
`JUDGE LEE: Is there any testimony from any witness
`indicating whether there are any technical difficulties in adapting
`access controls for high-level protocols to native low-level
`protocols or whether that is something that's routine, easily done
`by someone of ordinary skill? Is there specific testimony on that?
`MR. BOBROW: Well, my recollection, Your Honor,
`and I don't have a paragraph cite, is that Professor Chase provided
`that in his declaration, that this was within routine skill. If you
`give me a moment, I might be able to call that up. I apologize,
`Your Honor, I don't have that at my fingertips.
`Well, this is a discussion, Your Honor, at paragraphs
`245 to 247 of Professor Chase's declaration where he talks about
`partitioning of access rights. What I'm searching for is a specific
`statement that this is within the skill level of ordinary skill, which
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
` 18
`
`
`
`
`
`
`
`Cases: IPR2014-01197, IPR2014-01207, IPR2014-01209
`Patents: 6,425,035 B2, 7,051,147 B2
`
`I think was Your Honor's question. It may take longer than I
`have time in order to find that specific passage, but I think 245 to
`247 would have that.
`JUDGE LEE: You can verify that and return with it on
`your rebuttal, if you would like.
`MR. BOBROW: Thank you very much, Your Honor.
`Going back to the slide presentation, please, at slide 29, if we
`might, another argument that is made by the patent owner is that
`in the Bergsten/Hirai combination, it wouldn't work to provide
`the mapping allegedly because the host identity would be
`removed before reaching the map. And there is simply no
`support for that assertion, and what's more, it's counter to the
`express disclosure of Bergsten. We know from Figure 7 in
`Bergsten as shown on slide 30 that the input to that storage
`controller is the host interface ID with a host block number. And
`that information, that host interface ID is actually used and
`essentially the question is asked, does the host map exactly to the
`logical device.
`So in Bergsten specifically the host ID that is being sent
`from the host is expressly being used in that mapping function.
`So we have an express disclosure of that. There's no evidence
`that it is somehow stripped away or discredited.
`And with that, let me turn to the final combination,
`which is the Kikuchi/Bergsten combination, unless Your Honors
`have further questions about the Bergsten/Hirai combination.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
` 19
`
`
`
`
`
`
`
`Cases: IPR2014-01197, IPR2014-01207, IPR2014-01209
`Patents: 6,425,035 B2, 7,051,147 B2
`
`
`And Judge Lee, to address your comment, I apologize it
`took so long to find the reference. After paragraph 249 of
`Professor Chase's declaration, he said -- and this is Exhibit 1010.
`He said, quote, To the extent that Bergsten lacks access controls,
`configurable through the mapping table, a skilled engineer would
`be motivated to combine the teachings of Hirai with the address
`map of Bergsten to manage access controls to groups of host
`computers within the system.
`And in paragraph 250 he says, quote, “In the combined
`system the partition control table of Hirai would be merged with
`the map of Bergsten.” But the ordinary skill language is in
`paragraph 249.
`JUDGE CHERRY: Which petition is that?
`MR. BOBROW: 1207. Now, in the Kikuchi/Bergsten
`combination starting at slide 36, the Kikuchi reference also
`discloses substantially all of the elements of the claimed
`inventions. It is a storage system. It discloses multiple hosts. It
`discloses a router and routing logic. It discloses access controls
`and it provides mapping. And it discloses, however, a single
`storage device. There is one large disk that these multiple hosts
`communicate with remotely. And there are access controls that
`are described by Kikuchi in two ways. One is an access control
`that looks at a host ID and asks, does this host have any rights to
`access the storage device? That's one type of access control, do
`you have rights to access the disk?
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
` 20
`
`
`
`
`
`
`
`Cases: IPR2014-01197, IPR2014-01207, IPR2014-01209
`Patents: 6,425,035 B2, 7,051,147 B2
`
`
`The other type of access control that it provides is
`essentially allocation and partitioning. What I mean by that is
`that Kikuchi says that you can essentially allocate storage in the
`disk to particular hosts by creating partitions such that the disk
`apparatus is able to allocate a different disk partition to each host.
`The idea is that each host gets its own storage through this
`partitioning mechanism. And an offset system essentially is used
`to provide for these partitions so that when a particular host
`requests storage, there will be a table that essentially says host
`number 1 gets a particular offset and it will then direct that host to
`a particular partition of storage in the disk. That amounts to an
`access control.
`Now, in terms of what is missing from Kikuchi, it does
`not disclose the multiple storage devices that are in the claims of
`the patents at issue. And we rely upon Bergsten to provide that
`functionality. What Bergsten says, and we discussed this briefly
`before, is that it provides for a multiplicity of storage devices and
`mapping from the host to the logical storage level to the physical
`storage level.
`What the proposed combination is, is to take the
`Kikuchi