`IPR2014-01463, Paper No. 48
`IPR2014-01544, Paper No. 49
`December 1, 2015
`
`
`Trials@uspto.gov
`
`
`
`
`
`571-272-7822
`UNITED STATES PATENT AND TRADEMARK OFFICE
`____________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`____________
`
`CISCO SYSTEMS, INC., QUANTUM CORP., ORACLE
`CORP., DOT HILL SYSTEMS CORPORATION,
`Petitioners,
`
`v.
`
`CROSSROADS SYSTEMS, INC.,
`Patent Owner.
`____________
`
`Cases: IPR2014-01226, IPR2014-01463, IPR2014-01544
`Patents: 6,425,035 B2, 7,934,041 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 10:00 a.m., at the U.S. Patent and
`Trademark Office, 600 Dulany Street, 9th Floor, Hearing Room
`D, Alexandria, Virginia.
`
`
`
`
`
`Cases: IPR2014-01226, IPR2014-01463, IPR2014-01544
`Patents: 6,425,035 B2, 7,934,041 B2, 7,051,147 B2
`
`
`APPEARANCES:
`
`ON BEHALF OF THE PETITIONERS:
`
`
`
`ON BEHALF OF THE PATENT OWNER:
`
`
`
`
`KEITH A. RUTHERFORD, ESQUIRE
`JAMES H. HALL, ESQUIRE
`Blank & Rome, LLP
`717 Texas Avenue, Suite 1400
`Houston, Texas 77002
`
`
`
` 2
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`ANDREW S. EHMKE, ESQUIRE
`DAVID L. McCOMBS, ESQUIRE
`SCOTT T. JARRATT, ESQUIRE
`Haynes and Boone, LLP
`901 Main Street, Suite 3100
`Dallas, Texas 75202-3789
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`Cases: IPR2014-01226, IPR2014-01463, IPR2014-01544
`Patents: 6,425,035 B2, 7,934,041 B2, 7,051,147 B2
`
`
`
`P R O C E E D I N G S
`- - - - -
`JUDGE POWELL: Good morning. This is the oral
`hearing for three related cases: IPR2014-01226, which involves
`patent 6,425,035 B2; IPR2014-01463 which involves patent
`7,943,041 B2; and IPR2014-01544 which involves patent
`7,051,147 B2.
`We have also some IPRs that have been joined with
`these proceedings. IPR2015-00825 has been joined with
`IPR2014-01226. IPR2015-00854 has been joined with
`IPR2014-01463. And IPR2015-00852 has been joined with
`IPR2014-01554.
`In the hearing room with me I have Judges Lee and
`Cherry. And joining us from Denver we have Judge Kalan. With
`that, can counsel please state your names for the record.
`MR. McCOMBS: Your Honors, I'm David McMombs,
`lead counsel for Cisco and Quantum in this proceeding, along
`with the joint petitioners Oracle and Dot Hill. With me is
`Andrew Ehmke and Scott Jarratt. Andy Ehmke will be doing our
`presentation today.
`With us we also have client representatives for Cisco
`and Oracle and also their District Court trial counsel in
`attendance as well. Thank you.
`JUDGE POWELL: Thank you.
`
` 3
`
`
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`Cases: IPR2014-01226, IPR2014-01463, IPR2014-01544
`Patents: 6,425,035 B2, 7,934,041 B2, 7,051,147 B2
`
`
`MR. RUTHERFORD: Your Honor, Keith Rutherford
`for Crossroads, the patent owner. With me at counsel table is
`also James Hall. We likewise have members of Crossroads here.
`Our CEO and COO are here as well. Sorry, lead counsel is Steve
`Sprinkle, by the way, also here.
`JUDGE POWELL: Thank you. Let me start by
`addressing the joint list of objections to demonstrative exhibits.
`Having reviewed and considered the list, we are going to allow
`both parties to use any of their demonstratives in today's hearing.
`We will carefully consider which evidence and arguments have
`been properly presented when we prepare our final decisions for
`these cases.
`Per the trial hearing order each party will have
`60 minutes to present arguments for the subject cases. Petitioners
`will go first and may reserve time for rebuttal. Patent owner will
`then respond to petitioners’ presentation and petitioners may use
`any remaining time to respond to patent owner's presentation.
`One important point to note is, as noted in the hearing
`order, while you are presenting, you must identify each
`demonstrative exhibit clearly and specifically such as by slide
`number or screen number. That's particularly important for Judge
`Kalan who cannot see the projection screen from Denver.
`With that, do we have any questions before we start?
`MR. RUTHERFORD: No, Your Honor.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
` 4
`
`
`
`
`
`
`
`Cases: IPR2014-01226, IPR2014-01463, IPR2014-01544
`Patents: 6,425,035 B2, 7,934,041 B2, 7,051,147 B2
`
`
`MR. HALL: Your Honor, we have our hard copies of
`our presentation for you.
`JUDGE POWELL: Sure. Why don't you bring those
`
`up.
`
`MR. JARRATT: We do as well.
`JUDGE POWELL: Okay. Is everybody ready? Let's
`start with petitioner. Would you like to reserve time?
`MR. EHMKE: Yes, Your Honor, I would like to
`reserve 20 minutes for rebuttal.
`JUDGE POWELL: When you are ready.
`MR. EHMKE: Again, my name is Andy Ehmke of
`Haynes and Boone here on behalf of the petitioners, Cisco
`Systems and Quantum Corporation, as well as the consolidated
`petitioners Oracle Corporation and Dot Hill Systems. As Your
`Honor mentioned, we are discussing U.S. patent number
`6,425,035, 7,934,041 and 7,051,147.
`And the two main issues we'll be talking about today, as
`we proceed to slide number 2, is it is our position that patent
`owner has improperly narrowed the map and access control
`limitations found in the claims across the three patents using their
`arguments that are importing additional limitations into the
`claims that we think are not the proper broadest reasonable
`contraction of the claim terms.
`But regardless of the interpretation, whether we apply
`the broadest reasonable construction of patent owner's improperly
`
`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: IPR2014-01226, IPR2014-01463, IPR2014-01544
`Patents: 6,425,035 B2, 7,934,041 B2, 7,051,147 B2
`
`narrowed construction, that the prior art of record, the CRD
`manual and the HP Journal and their teachings disclose and
`render obvious the claims of the '035, '041 and '147 patents.
`As we proceed, to just briefly discuss the three patents
`on slide 3 and then quickly moving to slide 4, what we have here
`is a figure from the patents. Now, we are referring to the '035
`patent here, but the patents have a common specification. So we
`are using the '035 as shorthand for today's proceedings.
`This is Figure 3. And what we see in Figure 3 are a
`number of workstations on the left designated by Element 58.
`They are coupled to a storage router in the middle designated as
`Element 56. And on the right-hand side are a number of storage
`devices, Elements 60, 62 and 64. And the focus of today's
`dispute is about the storage router and some of its features, in
`particular its mapping limitation as well as its access controls of
`how its mapping workstations on the left and storage on the right
`and how it's controlling access of the storage by the workstations
`on the left.
`Proceed to the next slide, we see how this manifests
`itself in the claim. We are on slide 5. Here is an example claim
`from the '035. Now, what we are focusing on here are the last
`key remaining issues, the mapping limitation and the implement
`access controls limitation. And we see the language here is quite
`broad, it's operable to map between. There's no specificity with
`respect to the mapping. And we see the access controls is
`
`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: IPR2014-01226, IPR2014-01463, IPR2014-01544
`Patents: 6,425,035 B2, 7,934,041 B2, 7,051,147 B2
`
`implement access controls with no specificity as to how to
`implement the access controls. You simply implement the access
`controls.
`If we move on to slide 6 we see in the '041 patent we
`are maintaining a map to allocate storage. Now, this map is using
`representations of hosts and representations of storage. And then
`the access control is controlling access without any specificity as
`to how the access is controlled.
`Again, if we turn to slide 7 in the '147 patent, we see
`again the mapping limitation is simply maps between the device
`and the controls are implement access controls.
`Now, if we look at the next slide, as we look at what the
`prior art teaches at a high level, slide 8, moving on to slide 9, we
`see the disclosure from the CRD manual. Now, what the CRD
`manual teaches is for your central controller, in this case, the
`CRD-5500 device, it provides a host LUN mapping feature. With
`this mapping feature, you can assign redundancy groups to a
`particular host. Now, in the terminology of the CRD manual a
`redundancy group is a subset of storage. It could be a partition.
`It could be a RAID set. It could be a number of hard drives. So
`when we see redundancy groups, we are talking about a bit of
`storage. So the mapping feature is assigning storage to a
`particular host. And if we look under the access control
`paragraphs, we see this manifests itself again. We see that the
`host LUN mapping feature makes it possible to map rate sets,
`
`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: IPR2014-01226, IPR2014-01463, IPR2014-01544
`Patents: 6,425,035 B2, 7,934,041 B2, 7,051,147 B2
`
`storage, to map rate sets differently to each host. That's the
`teaching of the CRD manual.
`Now, what this feature provides, because we have the
`mapping of the RAID sets differently to each host is that we can
`now make some of these storage, these redundancy groups visible
`to one host but not to another. We can control the access.
`JUDGE CHERRY: Excuse me. So how does the CRD
`manual define host? Do they tell us anything about what a host
`is?
`
`MR. EHMKE: The host in the CRD are simply
`computers or workstations, Your Honor.
`JUDGE CHERRY: I mean, well, your opponent has
`argued that while it does say this, what it's really teaching is that
`it's teaching mapping to a particular channel. And so why would
`a person of ordinary skill understand this to mean a computer
`when it's -- when they are saying that it's talking about channels?
`MR. EHMKE: Absolutely. In the mapping of the
`CRD, what it's doing is it has this mapping, we've got the
`redundancy groups. This is the map feature and I'm sorry, we are
`referring to slide 25. I'll pause for a second there for Judge
`Kalan. We are looking at slide 25.
`In the mapping of the CRD, it's mapping it -- this is a
`host LUN mapping feature and it's mapping it to a representation
`of the host that it calls a channel number. It's a number that's
`used to identify the host associated with that channel. We have a
`
`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
`
` 8
`
`
`
`
`
`
`
`Cases: IPR2014-01226, IPR2014-01463, IPR2014-01544
`Patents: 6,425,035 B2, 7,934,041 B2, 7,051,147 B2
`
`host, we have a cable, we have a number that identifies that host
`because it's associated with that cable with that host. So we have
`a map for the host represented by channel 0. We have a map
`represented by channel 1 for another host.
`So the nomenclature of mapping to the device versus
`mapping to the channel, I think, is a little bit of wordsmithing,
`because when we look at the patent with respect to the mapping
`to the device, we are not taking the physical computer and
`inserting it into a virtual map inside the controller. We are not
`taking the device. So it's not really mapping to the device. It's
`mapping to a representation of the device, an identifier of the
`device. The identifier could be a SCSI address, a fiber channel
`address, some other identifier. When we are talking about
`mapping to a device, we are talking about mapping to a
`representation of the device, an identifier of the device so that we
`can facilitate routing access controls.
`JUDGE KALAN: Excuse me. Is the device of the
`claims different than the host or the workstation you just
`discussed?
`MR. EHMKE: No, Your Honor. The term device host
`and workstation are used interchangeably throughout the
`specification and the claims. So the notion of mapping to the
`device is really mapping to an identifier or a representation of the
`device. And the channel number in the CRD performs the same
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
` 9
`
`
`
`
`
`
`
`Cases: IPR2014-01226, IPR2014-01463, IPR2014-01544
`Patents: 6,425,035 B2, 7,934,041 B2, 7,051,147 B2
`
`feature. It's a number, the mapping used to route the storage
`information to the respective hosts.
`JUDGE LEE: As I understand it, there is no real
`dispute about what the CRD manual teaches and how that works
`as between the parties. Both parties seem to agree that the CRD
`manual teaches mapping between a redundancy group and a host
`channel. And it's simply your argument that that meets the
`limitation of what mapping is and the opposing party disagrees
`with that. So it seems to me to really boil down to the claim
`construction of mapping. So I would appreciate if you could
`address what does the intrinsic evidence indicate the proper scope
`of mapping would be?
`MR. EHMKE: Absolutely, Your Honor. So we'll turn
`to slide 13, then, moving on to 14. The first piece of evidence is
`going to be the claim language itself. We see here with respect to
`the '035 patent that the claim language is saying it's operable to
`map between devices. There's no specificity in this claim
`language about what has to be in the map, what type of
`characteristics the identifiers have to have in the map, what
`features or functionality or even how it's operable to map
`between. Likewise, we have maps between the device. No
`specificity at all with respect to the claim language of what's
`required to meet the map limitation.
`And then we have the '041 which has extra words but
`isn't really saying a whole lot more. It's allocating storage space
`
`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: IPR2014-01226, IPR2014-01463, IPR2014-01544
`Patents: 6,425,035 B2, 7,934,041 B2, 7,051,147 B2
`
`by associating representations of the devices. And if we look to
`the specification, we'll go to slide 15, the specification is likewise
`as broad. The specification is saying allocate storage to each
`workstation using mapping tables or other mapping techniques.
`There's no specificity here. They are relying on the knowledge of
`a person of ordinary skill in the art as to just use whatever
`mapping technique you desire. They don't claim it and they don't
`require it in their specification.
`We asked their expert during the deposition, was there
`anything that we are missing with respect to this? Is there
`anything specific in the specification about these mapping tables
`or mapping techniques? He says, no, there's nothing specific
`disclosed in the patent.
`That is why our position is the broadest reasonable
`construction because what the Board has adopted, if we go to
`slide 16, is simply allocating storage. There's no requirements.
`There's no special techniques. There's no characteristics required
`in the claims or described in the specification.
`JUDGE CHERRY: Well, I don't think the dispute is
`necessarily about what particular technique is used. It's about
`what is actually mapped in terms of does it have to be a particular
`device or not. And I think to the extent it does say each attached
`workstation, would that indicate that that's a particular device?
`MR. EHMKE: I would argue no, Your Honor. There's
`nothing wrong with having a representation of a plurality of
`
`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: IPR2014-01226, IPR2014-01463, IPR2014-01544
`Patents: 6,425,035 B2, 7,934,041 B2, 7,051,147 B2
`
`devices. Each device could be represented by a single
`representation. For example, you could have the accounting
`department represented by a single representation and control
`their access to the accounting department's data while restricting
`their access to the HR department's data. There's no requirement
`within the claims nor description specification that mandates that
`the mapping have a unique representation for each host nor that
`each host must have its own specific unique representation.
`JUDGE LEE: Is there anything in the specification that
`indicates that the claims contemplate or include the claimed
`device that's being mapped to be multiple hosts or multiple
`workstations?
`MR. EHMKE: I'm sorry, that requires that the
`representation applied to multiple devices?
`JUDGE LEE: Not requires necessarily, but does the
`intrinsic record indicate whether or not the term "device," the one
`that's being mapped in these claims, whether that could
`encompass a group of hosts or a group of workstations all
`together comprising the claimed device? Is that something that's
`within the scope of the claims?
`MR. EHMKE: If we could turn to the claim language
`
`again --
`
`JUDGE CHERRY: I think what Judge Lee is trying to
`get at is the group, is there any -- I mean, you have got to point
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
` 12
`
`
`
`
`
`
`
`Cases: IPR2014-01226, IPR2014-01463, IPR2014-01544
`Patents: 6,425,035 B2, 7,934,041 B2, 7,051,147 B2
`
`out that it doesn't exclude that. Is there anything that indicates
`that it's encompassed within that?
`MR. EHMKE: I do not recall there is any specific
`disclosure of representing a plurality of hosts. But I also would
`point out there's nothing in the specification that requires patent
`owner’s desired limitation of having unique representations for
`every host in all situations, particularly given the breadth of the
`claims.
`
`I would also argue that this argument by patent owner is
`importing a preferred embodiment into the specification which
`would be an improper narrowing of the claims.
`JUDGE KALAN: In the claims, in each of these
`separate patents, I'm looking at claim 1 in each of the patents,
`devices are identified differently, sometimes plural and in the
`case of the '147 patent singular. Is that significant?
`MR. EHMKE: I think it is, Your Honor. When we
`look at the language of the claims, again, that's a position we start
`with, the language of the claims. Because it says map between
`devices connected to the first transport medium and the storage
`devices, again, there's nothing that limits the claim that the
`mapping has to be a one-to-one correlation. We are mapping
`between the devices. So I think even with that claim construction
`or claim terminology, we can map between two devices to the
`first transport media in the storage devices. There's nothing
`wrong with saying we can map two devices to one storage device.
`
`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
`
` 13
`
`
`
`
`
`
`
`Cases: IPR2014-01226, IPR2014-01463, IPR2014-01544
`Patents: 6,425,035 B2, 7,934,041 B2, 7,051,147 B2
`
`We could map a department to its data. The claims support that
`reading, yes.
`JUDGE CHERRY: I think they have kind of pointed us
`to Figure 1 which shows one device to one storage group. What
`is your response to that -- I'm sorry, not Figure 1. Figure 3.
`MR. EHMKE: Our response to that is Figure 3 is
`specifically designated as an embodiment of the invention. It's
`not the invention itself. They even actually refer to Figure 3 and
`say there are many different configurations associated with it.
`You can change the type of protocol associated with it. You can
`change the type of hosts. It's not the invention. The specification
`does not limit the invention to that embodiment. We are looking
`at the claims as they are written and they are written broader in
`the embodiments disclosed in Figure 3.
`JUDGE LEE: I would like to ask you a question about
`the CRD manual and the system it describes. You put up a figure
`earlier, I believe it was slide 25, from that manual and you point
`to Channel 0. This entire table relates to Channel 0. You point to
`Channel 0 as corresponding to the host or the claimed device. I'm
`just trying to understand how this works. What does a host LUN
`represent if the entire channel is a single host?
`MR. EHMKE: Yes, Your Honor. What the host LUN
`is, from the host's perspective, this is a virtualized system. So
`from the host's perspective, it thinks, as much as a computer can
`think, it thinks it's writing to a piece of storage. And when it's
`
`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: IPR2014-01226, IPR2014-01463, IPR2014-01544
`Patents: 6,425,035 B2, 7,934,041 B2, 7,051,147 B2
`
`writing, it's referring to host LUN. So the host on Channel 0 is
`going to say, I want to write data to host LUN 5. That comes in,
`the mapping for that host, because the host is on Channel 0. That
`mapping is then looked up and it says, this host wanted to write
`data or read data from host LUN 5. Let's see what actual piece of
`storage it can read or write to. And it sees, oh, wait, 5 is not
`associated with a particular redundancy group. I'm going to
`control its access and not permit that read or write request.
`Now, if the host on Channel 0 says I want to read from
`host LUN 14, that request comes in on Channel 0. The system
`says, oh, I should use the mapping for Channel 0. It grabs the
`mapping and says this request was for host LUN 14. Which hard
`drive, which partition is that associated with? For host LUN 14,
`its correspondence is to redundancy group 14. And then it
`executes the read or write command with respect to that particular
`hard drive.
`So it's virtualizing the hard drive on the right-hand side,
`if you will, using the redundancy group representations. And
`then the host, which is represented by Channel 0, is using another
`virtualizing layer of the host LUN map numbers. Does that
`answer your question?
`JUDGE LEE: Yes, it does. Thank you.
`JUDGE CHERRY: Wait. Just so I understand, the host
`LUNs, would those correspond to kind of, you know, like a D
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
` 15
`
`
`
`
`
`
`
`Cases: IPR2014-01226, IPR2014-01463, IPR2014-01544
`Patents: 6,425,035 B2, 7,934,041 B2, 7,051,147 B2
`
`drive and would those be various drives that are represented on
`the host?
`
`MR. EHMKE: Yes. Use an example, Your Honor, you
`could theoretically say that the host associated with Channel 0,
`that host might think it's writing to the D drive. We'll think of D
`drive as the host LUN value using this example. It's not quite
`correct, but I think it applies for this example. So that host is
`going to write, in its mind, to drive D. That drive D is the host
`LUN. So that's host LUN 14. Host LUN 14 corresponds to hard
`drive 14. And so when the host writes its command, it's pulled
`because it's on Channel 0. That's its representation. It then says,
`okay, that write command was for D which corresponds to host
`LUN 14 which corresponds to redundancy group 14, and the
`command is sent to the appropriate storage device at the other end
`of the system.
`Any further questions on the host LUN mapping?
`JUDGE POWELL: No, but I have a question about
`this. I'm looking at slide 14 again. Could you address, if I were
`to decide that I believe -- and I'm looking at differences between
`the claim languages in the various patents here. If I decide that
`the CRD reference manual maintains a map to allocate storage
`space on the remote storage devices, what about -- how do you
`address the limitation of representations of the devices connected
`to the first transport medium? So, I gather that in some fashion
`the channel number is a representation of the device.
`
`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: IPR2014-01226, IPR2014-01463, IPR2014-01544
`Patents: 6,425,035 B2, 7,934,041 B2, 7,051,147 B2
`
`
`MR. EHMKE: That's correct.
`JUDGE POWELL: Can you elaborate on that any?
`MR. EHMKE: Sure. A couple of points with respect to
`your question. We have the three claims up here and in terms of
`how they have been argued and presented in this proceeding, they
`have actually been argued in terms of a coextensive claim
`construction matter the way the patent owner has argued them.
`So while we do have different terminology, we believe the claim
`scope is essentially the same across all three.
`Now, with respect to your particular question on the
`'041 about the associating representations of the devices, our
`position on that is, again, the CRD has the mapping and there is a
`number in that mapping that is called the channel number. That
`number is a representation of a host. When that communication
`from the host comes in, it's going to use that channel number to
`grab the corresponding map and apply the redundancy group
`analysis and make sure that has access to the hosts.
`Yes, we believe that the channel number qualifies as the
`representation. And again, the claim language doesn't say what
`type of representation is required. The specification doesn't
`require a particular type or manner of representation. And that
`was the focus of my earlier discussion about the mapping
`techniques is, again, the specification isn't requiring or mandating
`a particular type of mapping technique. No requirement as to the
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
` 17
`
`
`
`
`
`
`
`Cases: IPR2014-01226, IPR2014-01463, IPR2014-01544
`Patents: 6,425,035 B2, 7,934,041 B2, 7,051,147 B2
`
`particular type of representation either. It's simply a number used
`in a mapping so that we know which host and which map to use.
`JUDGE LEE: Is the channel number a representation of
`the device or simply the device's location? And if it's the latter,
`does that matter? Because what Crossroads seems to be arguing
`is that it's a port number, essentially. It tells you where to route
`things. It tells you where the associated host might be. But if
`you take that computer, the computer number 5, you move it to a
`different channel, the storage router won't know the difference.
`MR. EHMKE: That is precisely their argument. And
`the reason why we believe that argument fails is because that's
`looking at the CRD in a vacuum. The proposed ground has never
`been anticipation over the CRD. It is not about the CRD-5500
`system as it exists. The argument has been from the beginning
`it's obvious in view of the teachings of the CRD when combined
`with the Hewlett-Packard Journal.
`So what we have with respect to the -- so the initial
`teaching, the first teaching that's part of the ground of rejection is
`the mapping feature that uses the channel number as a
`representation. So it's teaching user representation in the
`mapping. And with that representation, we assign the storage to
`particular hosts. We assign map rates differently for each host.
`That's the baseline teaching of we are mapping storage differently
`to each host using representations of the hosts, but we are
`combining that with the Hewlett-Packard Journal. The HP
`
`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: IPR2014-01226, IPR2014-01463, IPR2014-01544
`Patents: 6,425,035 B2, 7,934,041 B2, 7,051,147 B2
`
`Journal is disclosing that you can take a plurality of SCSI devices
`and put them all on a single fibre channel loop. And so now we
`have a teaching where we have a combined system using the
`teachings of we have a host LUN mapping feature that's assigning
`redundancy groups differently to each hosts, but now those hosts
`are on a single fibre channel loop.
`The HP Journal also discloses that one of the features of
`it is that you can retain the initial input/output services associated
`with your underlying --
`JUDGE CHERRY: Is this in your petition?
`MR. EHMKE: It absolutely is in our petition.
`JUDGE CHERRY: Where is it in your petition?
`MR. EHMKE: So as we discussed, we see these are
`citations from our petition here with respect to the baseline
`teachings of the CRD. Then on page 19, Your Honor, we further
`discuss specifically the teachings -- and I have pulled up the
`petition. It's paper 3 in the 01226 proceeding, and we are looking
`on page 19 to start here. That's what's on the screen, page 19
`from the petition in the '035.
`Here we talk about the features and services provided
`by the CRD system. We are talking about the mapping, the
`allocation, the controlling of the access. Then if we scroll down,
`we say here in connection with the talking about the combination
`of applying the CRD's teachings with the Hewlett-Packard
`Journal, we specifically say that you would do this because you
`
`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: IPR2014-01226, IPR2014-01463, IPR2014-01544
`Patents: 6,425,035 B2, 7,934,041 B2, 7,051,147 B2
`
`are going to resolve the slots problem because it supports the
`same input/output services with a fewer number of slots. So we
`want to maintain the input/output services but just use less slots.
`And an additional feature of it is because we are taking
`all these individual computers and we are putting them all on the
`same loop, we need to handle addressability. That's a feature
`described in the Hewlett-Packard Journal. It's providing
`increased addressability even though we are putting them all on
`the same fibre channel loop.
`We went further. As we scroll down to page 25 of the
`petition, we said then on top of that, we have the baseline
`teachings. We are teaching using the mapping. We are teaching
`that you want the same input/output services. You are just going
`to use less cards. We said then you would make modifications to
`the components of the CRD-5500 necessary to keep them
`operating in their intended manner.
`Again, the whole point of this discussion has been about
`maintaining the mapping of the redundancy groups differently to
`each host and that a person of ordinary skill in the art would be
`able the make those necessary modifications.
`We outline all of this in the initial position of combining
`the teachings of the Hewlett-Packard Journal onto the CRD-5500.
`The resulting combined system would be a system that maps
`redundancy groups differently to each host where all those hosts
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
` 20
`
`
`
`
`
`
`
`Cases: IPR2014-01226, IPR2014-01463, IPR2014-01544
`Patents: 6,425,035 B2, 7,934,041 B2, 7,051,147 B2
`
`are on a single fibre channel loop. This was not a new argument.
`It was presented from the beginning.
`Now, what the patent owner is latching onto in their
`slides, Your Honors, is we asked further questions of their own
`expert with respect to would a person of ordinary skill in the art
`have this knowledge to make these modifications? And he
`actually walked through some of the things that he might do. He
`talked about, oh, yeah, the Tachyon chip, which is disclosed in
`the Hewlett-Packard Journal, does have the ability to distinguish
`and identify among the hosts. That's what we are talking about
`here. He confirmed the knowledge of a person of ordinary skill
`in the art, confirmed that a person of skill in the art would be able
`to make the modifications to the CRD to keep them operating in
`their intended manner.
`JUDGE CHERRY: But you didn't describe those
`modifications in your petition, right?
`MR. EHMKE: We described that we would make the
`modifications to the components. We described all of the
`teachings necessary. We didn't build the system. Patent owner's
`questions are -- one of their examples is did you modify the
`firmware. There's no requirement to implement the system. The
`teaching associated with this is that we have redundancy groups
`that were mapping the hosts using the structure. The test for
`obviousness is not whether the features may be bodily
`incorporated resulting in the system, and we actually cited this 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