`
`
`
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`____________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`____________
`
`EMC CORPORATION, LENOVO (UNITED STATES), INC. and NETAPP,
`Petitioner,
`
`v.
`
`INTELLECTUAL VENTURES I, LLC,
`Patent Owner.
`____________
`
`Case IPR2017-00429
`Patent 6,775,745 B1
`____________
`
`Record of Oral Hearing
`Held: September 7, 2018
`
`
`
`
`Before JEFFREY S. SMITH, GEORGIANNA W. BRADEN, and DANIEL
`J. GALLIGAN, Administrative Patent Judges.
`
`
`
`
`
`
`
`
`
`
`Case IPR2017-00429
`Patent 6,775,745 B1
`
`
`
`APPEARANCES:
`
`ON BEHALF OF THE PETITIONER:
`
`
`PETER M. DICHIARA, ESQUIRE
`DANA O. BURWELL, ESQUIRE
`Law Office of Wilmer Hale
`60 State Street
`Boston, Massachusetts 02109
`(617) 526-6000
`
`THEODOROS KONSTANTAKOPOULOS, ESQUIRE
`7 World Trade Center
`250 Greenwich Street
`New York, New York 10007
`(212) 230-8800
`
`
`
`ON BEHALF OF THE PATENT OWNER:
`
`
`BRENTON R. BABCOCK, ESQUIRE
`TED M. CANNON, ESQUIRE
`Law Office of Knobbe Martens
`2040 Main Street, 14th Floor
`Irvine, California 92614
`(949) 760-0404
`
`and
`
`TOM BROWN
`JAMES HIETALA
`
`
`
`
`The above-entitled matter came on for hearing on September 7, 2018,
`
`commencing at 1:01 p.m., at the U.S. Patent and Trademark Office, Madison
`Building, 600 Dulany Street, Alexandria, Virginia, 22314.
`
`2
`
`
`
`Case IPR2017-00429
`Patent 6,775,745 B1
`
`
`P R O C E E D I N G S
`- - - - -
` JUDGE SMITH: Please be seated. We're here today
`for IPR 2017-00429 for U.S. Patent Number 6,775,745.
`Petitioner, will you please step up to the podium and state
`your appearance?
` MR. DICHIARA: Good afternoon. My name is Peter
`Dichiara and I represent the petitioners.
` JUDGE SMITH: And who do you have with you,
`Petitioner?
` MR. DICHIARA: With me today I have Theodoros
`Konstantakopoulos and Dana Burwell. They're both from
`Wilmer, Cutler, Pickering, Hale and Dorr representing
`petitioners, and I also have Tom Brown from petitioner EMC.
` JUDGE SMITH: Thank you. Patent owner, will you
`please step up to the podium and state your appearance?
` MR. BABCOCK: Good afternoon, Your Honors. I'm
`Brent Babcock with Knobbe Martens. I'm representing the
`patent owner, Intellectual Ventures. With me is Ted Cannon
`also with Intellectual Ventures. We also have James Hietala,
`in-house counsel at Intellectual Ventures.
` JUDGE SMITH: Thank you. So both sides have 60
`minutes to present their case. Petitioner, you'll present
`your case first. Patent owner, you'll respond. Petitioner,
`you may reserve time for rebuttal if you desire, and then
`patent owner, you can have a brief surreply.
`
`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
`26
`
`3
`
`
`
`Case IPR2017-00429
`Patent 6,775,745 B1
`
` When we go through the arguments if you reference
`the slides please -- when you discuss a slide please indicate
`which slide number you're on. That way our colleagues in
`Dallas can follow, and it's also when we review the
`transcript we can follow on the transcript.
` Petitioner, when you're ready you may begin. Do
`you wish to reserve time for rebuttal?
` MR. DICHIARA: Yes. Yes, I do, Your Honor.
` JUDGE SMITH: How much?
` MR. DICHIARA: I'm predicting it might be like 20
`minutes. It depends on the length of the initial
`presentation.
` JUDGE SMITH: Okay. Let me see. Let me --
` MR. DICHIARA: And I might not even need the full
`20.
` JUDGE SMITH: So I'll set the clock for 40
`minutes, and when you're ready you may begin.
` MR. DICHIARA: Sure. And one brief housekeeping
`matter.
` JUDGE SMITH: Yes.
` MR. DICHIARA: I've already provided a slide deck
`to the court reporter. I don't know if you would like a hard
`copy, if you would find it helpful.
` JUDGE SMITH: I think (inaudible).
` MR. DICHIARA: Okay. And would you like hard
`copies?
`
`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
`26
`
`4
`
`
`
`Case IPR2017-00429
`Patent 6,775,745 B1
`
` MR. BABCOCK: We're fine, Peter. Thank you.
` JUDGE SMITH: When you're ready you may begin.
` MR. DICHIARA: Good afternoon. May it please the
`board, as I mentioned earlier my name is Peter Dichiara. I
`represent the petitioners in this matter, and we are here
`today to discuss the '745 patent. Turning to slide 2, we're
`here to discuss why the challenge to claims 4 through 6 are
`unpatentable.
` The '745 patent concerns something called a hybrid
`cache, and it uses the term hybrid cache because it's a cache
`that considers both frequency and recency of a file's use in
`determining which entry to evict from the cache. Turning
`briefly to slide 3, which is one of the slides we had in the
`initial phase of the trial, this is an annotated figure 2A
`from the patent showing one of the embodiments in action.
` What you're seeing on the left side of the slide
`in blue is what the patent called frequency factors, and the
`frequency factors, as the patent explains, indicate how often
`a file is used, and you may recall from our papers that when
`the '745 patent was drafted the applicants thought that the
`prior art caches were, in their terms, strictly used MRU or
`LRU techniques, not frequency techniques, and you'll see that
`in column 2 of the patent.
` And what they thought they were providing to the
`state of the art was this consideration of frequency and the
`patent provided several embodiments that had frequency
`
`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
`26
`
`5
`
`
`
`Case IPR2017-00429
`Patent 6,775,745 B1
`
`factors which would indicate how often a file was used in
`different ways. Turning to slide 4, these are the issues I'm
`hoping to discuss this afternoon.
` Briefly just ground 1 and ground 4, you may recall
`that the initial petition had six grounds but three of those
`grounds were conditional premised on an argument that the
`patent owner might make for claim construction at the
`petition time. We didn't know what claim construction they
`were going to take so we had three grounds that relied on
`something called the Sweeney reference.
` In case the patent owner argued that extended
`segment meant one thing we had three other grounds relying on
`a different reference, Karedla if they argued extended
`segment meant another. At no point has patent owner proposed
`the one construction, so it's never really been a live
`dispute. We think those three grounds are mute.
` Ground 3 is the one we've already tried. We
`already had the hearing fully briefed, and the remaining two
`are grounds 1 and 4. So turning to --
` JUDGE SMITH: So for the three grounds that you're
`not arguing then you're suggesting that we can enter judgment
`in favor of patent owner on those grounds?
` MR. DICHIARA: I think the proper judgment would
`be that the grounds are moot because they report in as a
`conditional premise. It's never just been a live dispute, so
`it never got live.
`
`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
`26
`
`6
`
`
`
`Case IPR2017-00429
`Patent 6,775,745 B1
`
` JUDGE SMITH: I see. Okay.
` MR. DICHIARA: Okay. So turning briefly to slide
`5, this is the claim limitation from claim 4 which is at the
`center of the dispute for Burton in ground 1. A critical
`issue is whether Burton teaches frequency factors, and you'll
`probably recall that this is the same limitation that we
`briefed and argued in the initial phase of the trial.
` The critical inquiry here is that the frequency
`factors, as the claim states, indicates how often the file is
`accessed. It doesn't calculate it. Turning to slide 6 here
`is Burton. This is the primary reference for ground 1, and
`as you can see starting right from the title of Burton,
`Burton is directed to the very same problem and solution that
`the '745 patent is.
` The Burton reference uses the term demoting
`instead of evicting, but it's the same thing. They're trying
`to evict the entry from the cache based on its recency and
`frequency of access, trying to evict the least recently and
`least frequently. Turning to slide 7, to do this Burton uses
`something that -- a metric that it points, it calls the
`value, sometimes it calls it a ranking or a variable.
` Sometimes it calls it an LRU rank, but as you're
`seeing here in these two specific blurbs or papers mentioned
`others, it's quite clear that their ranking indicates
`frequency. You're seeing this on the bottom blurb quite
`explicitly, that the value indicates how frequently, and we
`
`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
`26
`
`7
`
`
`
`Case IPR2017-00429
`Patent 6,775,745 B1
`
`believe this is very powerful evidence and that under any
`reasonable interpretation the board should find that Burton
`satisfies the frequency factor.
` One thing I'd like to note here is that in the
`patent owner's papers and their supplemental response they
`don't address these passages at all, and from my review I
`don't think it shows up in their demonstratives either. We
`were a little curious about that at the time of the
`supplemental response and we had asked their expert, Dr.
`Chong (phonetic), what do these terms mean.
` If it says it's indicating frequency what else
`could it mean, and that back and forth went on for 20 pages
`without an answer and eventually it concluded us saying what
`would be your response if the PTAB was curious about a
`question to that -- an answer to that question, and Dr.
`Chong's response was that he thought it wouldn't be a
`meaningful question.
` JUDGE SMITH: The question being what do these
`passages in Burton mean?
` MR. DICHIARA: Yeah, what else could they mean
`other than that the rank indicates frequency because that's
`explicitly what it says multiple times.
` JUDGE SMITH: I see.
` MR. DICHIARA: Let me turn to slide 8. This is a
`side by side. You'll have to look at the fine print to
`figure out which one is from which because they're so
`
`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
`26
`
`8
`
`
`
`Case IPR2017-00429
`Patent 6,775,745 B1
`
`similar. On the top you're seeing the '745 patent saying that
`what the frequency factor does is it's increased with each
`use of a file, and you're seeing essentially verbatim
`disclosure in Burton that its rank has increased with access.
` So turning to slide 9 putting on the screen what
`the patent owner has proposed as a construction for the term,
`it's the same construction that they proposed in the initial
`phase of the case, and as I mentioned before this is
`something that has already been fully litigated in this case,
`briefed, orally argued.
` It's likewise the same construction that was
`briefed, argued and rejected in the unified patent IPR
`2016-1463 was the IPR number there. We don't think that this
`construction is proper under BRI. We don't think it's based
`on the specification even though we all know that broadest
`reasonable interpretation is supposed to be in view of the
`specification.
` Instead of having a construction that's accounting
`for the various embodiments in the patent that actually use
`the term frequency factor the patent owner tries to jump
`around and include some, exclude others relying on this
`language instead of relying on the specification.
` As I mentioned before, in a way it doesn't matter
`because it's our position that regardless Burton satisfies
`this limitation under any reasonable construction, both from
`its explicit teachings and just from the way it operates.
`
`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
`26
`
`9
`
`
`
`Case IPR2017-00429
`Patent 6,775,745 B1
`
` JUDGE SMITH: When you say any reasonable
`construction are you including patent owner's proposed
`construction?
` MR. DICHIARA: Yes, we believe it even satisfies
`their construction because Burton shows with the same degree
`of certainty that the '745 patent does.
` JUDGE SMITH: Okay.
` MR. DICHIARA: So I'm turning to slide 10 here,
`and this is just another example about why we think this
`proposed construction from the patent owner isn't a good one,
`and what you're seeing here, and we had a little of this in
`the initial phase, is this is the construction they're
`proposing, and their own expert, Dr. Chong, can't tell you
`what this construction means.
` Turning to slide 11 they double down IV, patent
`owner doubles down on this bad construction by in turn
`relying on, as we outlined in our papers, what we believed to
`be a bad hypothetical. The top blurb here is a -- are the
`results of their hypothetical where they created an access
`pattern and they tried to suggest that Burton's rank doesn't
`indicate with a fair degree of certainty.
` The second blurb, this is the patent owner's
`supplemental response. This is their paper, not ours, and
`they're stating that this hypothetical is – when the time
`counter is one, the state of the cache after the first access,
`this is when the cache is first initializing. It's not when
`
`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
`26
`
`10
`
`
`
`Case IPR2017-00429
`Patent 6,775,745 B1
`
`it's up and running and you're starting to evict anything.
` And then the third blurb is perhaps the most
`telling because we -- again, we examined their expert Dr.
`Chong on this, and here he's agreeing that eviction is
`something that occurs much later in time, many cycles after
`initialization. So one reason this hypothetical is bad is
`it's just simply not representative. It's not considering
`the cache at the appropriate time.
` Let me turn to the next slide, 17 -- 12, I'm
`sorry, I misspoke, and here is some further cross-examination
`testimony of Dr. Chong, and we're asking him did you analyze
`this later in time, and he didn't, and then we asked him
`would you be surprised if you got a different result at a
`later point in time, and he says he would not. He would not
`be surprised.
` And I just want to pause on that for a moment. If
`this is a hypothetical and you're getting a directly opposite
`conclusion from the same hypothetical you have to ask
`yourself whether this is a reliable hypothetical, and on top
`of that what we're asking him about is a more relevant time
`frame, something that is after initialization when you're
`more in a cache state where you're actually going to start
`evicting entries, which is what the whole claim is about,
`evicting entries, not initializing a cache.
` So let me turn to slide 13 and show what actually
`happens with the same access pattern that Dr. Chong and IV
`
`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
`26
`
`11
`
`
`
`Case IPR2017-00429
`Patent 6,775,745 B1
`
`used but used at a later point in time, and what you see is
`that the analysis shows that Burton behaves exactly like the
`way it said it would. Its explicit text going back to slide
`7 said the rank is going to indicate frequency.
` And what you're seeing on this analysis is that
`the entries that were accessed more, like the top entry A is
`accessed three times, it has a higher rank than the ones that
`were accessed two or the ones that were accessed once, so
`it's behaving exactly the way the text that they ignored said
`it would behave.
` And again this was using their access pattern. We
`didn't create one. So we believe that Burton should be a
`simple case for claim 4. It discloses indicating frequency
`under any reasonable construction. It uses the exact same
`phrase as the claim. It's like the self-tuning embodiment in
`the '745 patent that we explained in our papers, how it
`accessed the self-tuning implementation.
` And then as we mentioned, not that we're relying
`on this hypothetical, we're mostly using it to show that
`their hypothetical is bad, but even if you used it at the
`same point in time you're seeing that their own access
`pattern shows that it behaves exactly the way the explicit
`text said it would.
` So let me turn to slide 14 and claim 5 just
`briefly because it is in ground 1. As you can see claim 5 is
`simply stating what types of files are we're talking about.
`
`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
`26
`
`12
`
`
`
`Case IPR2017-00429
`Patent 6,775,745 B1
`
`Slide 15 we laid out the evidence in our petition, where
`that's met. At no point in time has the patent owner tried
`to independently defend claim 5. They've always relied on
`their analysis for claim 4, and you're seeing that quite
`clearly in slide 16. This is one of their initial papers.
` JUDGE BRADEN: And excuse me, Counselor, but for
`claim 5 you're relying on Karedla, correct, you're not
`relying on Burton or the dependent claim limitation?
` MR. DICHIARA: I believe that's right, Your Honor,
`yes.
` JUDGE BRADEN: Okay. Thank you.
` MR. DICHIARA: So turning to slide 16 this is --
`again this is their papers, their language, not ours. It's
`at the time of the popper. There hasn't even been an
`institution decision yet, and they refer to their claim 5 as
`a relatively insignificant depending claim, again completely
`consistent with the idea that it's not independently bringing
`anything to the table.
` Let me turn to slide 17 if you will in claim 6,
`which is a new claim. This hasn't been tried yet. It was
`not any part of the institution decision. As you can see
`from looking at the claim language it's further limiting the
`scanning limitation which first appeared in claim 4, and we
`believe like all the claims it should be viewed under BRI, so
`obviously it's further limiting claim 4 but it still has to
`be construed, understood consistently with the proper
`
`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
`26
`
`13
`
`
`
`Case IPR2017-00429
`Patent 6,775,745 B1
`
`principles that brought us reasonable interpretation.
` And one of the things that we would like to raise
`is that the record evidence that has developed in this case
`both during the initial phase of the trial and the
`supplemental phase, all after DI time, all after decision to
`institute time we believe clearly demonstrates that the
`scanning limitation is properly understood as providing a
`directional requirement for the scanning, which direction are
`you scanning.
` Now, the patent owner argues that it's narrower
`than that and it has to get to inclusivity, that it has to
`include the MRU. As I'll explain in the following slides, we
`think that construction is wrong, but we also think again it
`shouldn't matter because we would satisfy it either way,
`particularly with Burton as I'll get to in a moment. Oh, I'm
`sorry. I didn't have the right slide on the screen. So let
`me get to slide 19 and turn to the specification.
` So here's one passage from the specification that
`helps outline the point. This is from column 7 of the patent
`and it describes that you can -- how far forward -- they use
`the term search, but how far forward are you going to search
`from the LRU. So here's an embodiment that's talking about
`scanning from the LRU a certain amount, i.e. how far forward,
`and it's argued that claim 6 is reciting this directionality.
` Let me turn to the next slide, 20, and show why
`that's so. The immediately preceding clause is showing
`
`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
`26
`
`14
`
`
`
`Case IPR2017-00429
`Patent 6,775,745 B1
`
`another embodiment. In this case you're scanning in the
`other direction. You're scanning not from the LRU, but
`towards the LRU. So while claim 4 language is broad enough
`to cover both embodiments, claim 6 is just that given
`direction. That's our view.
` JUDGE SMITH: But let me ask you this: The claim
`doesn't say towards. It says to. I mean, can you speak to
`the -- are you asking us to read the word do to mean the word
`towards?
` MR. DICHIARA: No. No, Your Honor. What we're
`saying with the proper understanding of claim 6 is that it's
`broad enough to cover scanning to the MRU inclusive but not
`necessarily, and I'll show you exactly why there's support
`from that both from the spec and from their own expert, and
`--
` JUDGE SMITH: But, I mean, I just want to
`understand just --
` MR. DICHIARA: Yes.
` JUDGE SMITH: -- the directional aspects.
` MR. DICHIARA: Yes.
` JUDGE SMITH: I'm just asking you about with
`respect to the directional aspect.
` MR. DICHIARA: Exactly.
` JUDGE SMITH: The directional aspect would be -- I
`mean, the word towards means you're going in the direction of
`but not necessarily reaching?
`
`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
`26
`
`15
`
`
`
`Case IPR2017-00429
`Patent 6,775,745 B1
`
` MR. DICHIARA: Exactly right. One of the
`analogies I like to use on this is, you know, you might say
`you're going from your home to the grocery store, right, but
`let's say you're saying I'm going from my home to the grocery
`store but I'm looking for something. Now, you might go all
`the way to the grocery store or you might not, and that's
`exactly the way the patent behaves.
` JUDGE SMITH: But, I mean, in this context did the
`person who drafted this patent understand the distinction
`between the word to and the word towards? So with respect to
`just this directional aspect --
` MR. DICHIARA: Uh-huh.
` JUDGE SMITH: -- if we did say claim 6 is a
`directional aspect would we be -- so the claim says --
` MR. DICHIARA: Let me bring it back up on the
`screen so I'll go back to --
` JUDGE SMITH: This is 19 --
` MR. DICHIARA: Yeah.
` JUDGE SMITH: -- slide 19, or 17?
` MR. DICHIARA: 17 will show it larger.
` JUDGE SMITH: 17, all right.
` MR. DICHIARA: Yeah.
` JUDGE SMITH: So scanning from a frequency factor
`corresponding -- so the factor corresponds to the LRU. To a
`factor corresponding -- so this word to right in the middle
`--
`
`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
`26
`
`16
`
`
`
`Case IPR2017-00429
`Patent 6,775,745 B1
`
` MR. DICHIARA: Uh-huh.
` JUDGE SMITH: -- after the words LRU file, the
`word to --
` MR. DICHIARA: Correct.
` JUDGE SMITH: -- you would have us -- give this a
`meaning that that's similar to the meaning of towards; is
`that right?
` MR. DICHIARA: I --
` JUDGE SMITH: The meaning of to in this claim
`would be in the direction of?
` MR. DICHIARA: Yes, exactly, Your Honor, and we
`think that's supported both by the specification, by their
`expert's testimony, which I'll get to twice over both about
`specification examples and the behavior of the patent, and as
`I'll conclude with I'll show another reason at the end of the
`slide deck exactly why that is true, so --
` JUDGE SMITH: Okay. Okay.
` MR. DICHIARA: And it won't take long. It's only
`a few slides.
` JUDGE SMITH: Okay. Okay.
` MR. DICHIARA: So --
` JUDGE BRADEN: And also can you address the fact
`-- are you saying that searching and scanning are the same
`thing?
` MR. DICHIARA: Yes, the way the patent is using it
`at this point is saying that you can search in one direction
`
`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
`26
`
`17
`
`
`
`Case IPR2017-00429
`Patent 6,775,745 B1
`
`or scan in one direction, you can scan in the other, and
`there are reasons to do each just like they have lots of
`other embodiments. You'll recall that --
` JUDGE BRADEN: But there's not specific
`embodiments that are for searching and then other
`embodiments, such as those in figures 4 -- 2C and figures 3
`that would be for scanning?
` MR. DICHIARA: I think all of the embodiments,
`they're using searching and scanning somewhat
`interchangeably.
` JUDGE BRADEN: Interchangeably.
` MR. DICHIARA: The only difference is that
`searching is also folding into it what are you searching for,
`right. You're not just simply scanning but you're searching
`for something specifically, and what they're really looking
`for is this LR/LF candidate.
` JUDGE BRADEN: Okay. Okay. Thank you, Counsel.
` MR. DICHIARA: It's slightly different, not to
`nitpick. So let me turn to slide 21. This is the patent
`owner's construction in this phase of the case. The patent
`owner's construction insists that claim 6 requires that all
`files -- you see their emphasis -- must be scanned.
` Turning to slide 22, we're going back in time a
`little bit, this is what they said in the initial phase of
`the trial. This is Dr. Chong's written testimony. It's
`unsolicited because it's part of the patent owner response.
`
`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
`26
`
`18
`
`
`
`Case IPR2017-00429
`Patent 6,775,745 B1
`
`It's not in reply to anything, and he's noting that there are
`embodiments -- he's referring to a column 6 embodiment
`different than the one I just mentioned, by the way.
` The how far forward is from column 7. This is a
`column 6 embodiment. He's saying there's an embodiment that
`never scans the MRU file, so he's saying -- he's agreeing
`there are embodiments that don't scan everything. Let me
`turn to slide 23. This is again from this phase of the case
`when we had an opportunity to examine Dr. Chong's views with
`regard to the way the patent behaved, and you'll see at the
`top of this slide we're talking to him about a '745
`embodiment.
` That's what you're seeing in the lines 23, 24, 25,
`and we're asking him the way this would behave and to
`envision that the absolute least recently used entry, the one
`that's at the bottom of figure 2A, the LRU, happens to have a
`frequency factor of one, the absolute smallest value a
`frequency factor can be, and we asked him all right, well,
`here's a situation where the first thing you look at is the
`least recent, and there can't be anything less recent.
` It has the lowest frequency factor, there can't be
`anything lower than that, do you need to keep on searching
`all the way through, and he says no, underneath that, those
`assumptions of that hypothetical you wouldn't need to search
`anymore, so he's noting that a person skilled in the art
`would know that it's not necessarily going to be the case
`
`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
`26
`
`19
`
`
`
`Case IPR2017-00429
`Patent 6,775,745 B1
`
`that they're going to scan all the way to the MRU even in the
`embodiments that say they're going to scan to the MRU.
` This is the example I was trying to allude to
`earlier about if you say you're looking for something and I'm
`going to look for it on the street from my house to the
`grocery store, if I find it halfway up the street and my goal
`is searching for something I'm not going to keep on going to
`the grocery store. I already have what I want.
` JUDGE SMITH: Oh, I see. So yeah, if you lost
`your -- if you're walking home and you dropped your watch,
`and so once you find the watch you're not going to -- yeah, I
`see.
` MR. DICHIARA: Yeah.
` JUDGE SMITH: All right.
` MR. DICHIARA: And that's the way the patent
`behaves.
` JUDGE SMITH: I see. Okay.
` MR. DICHIARA: So let me get to your earlier
`point, Your Honor. You were asking about the claim
`limitation, the claim language about toward and to, right?
` JUDGE SMITH: Yes.
` MR. DICHIARA: Well, here's an example showing
`that the patentee, when they put this patent together and
`were writing the claims they knew how to write inclusivity
`when they wanted to. This is a scanning limitation sent
`unchallenged, claim 8, right, and it says we're scanning each
`
`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
`26
`
`20
`
`
`
`Case IPR2017-00429
`Patent 6,775,745 B1
`
`of the files.
` And we would submit that this is further evidence
`showing that the from and to language is directional in
`nature, not demanding inclusivity. Our view is it's broad
`enough to cover if you scanned all the way up to the MRU
`because there are certainly embodiments that do that, but
`it's also broad enough to cover ones that go less than MRU
`just as long as it's in the right direction. It has to be in
`that direction.
` JUDGE SMITH: So your point with claim 8 is that
`this claim is limited to scanning every file; is that right?
` MR. DICHIARA: Or the ones they recite. The point
`I'm making is that if the patentee, when they were interested
`in demanding inclusivity of certain things actually being
`scanned and not stopping halfway up, they used terminology
`consistent with that, they said scanning each of the files,
`right, so this gets to the point that we shouldn't be reading
`inclusivity into the scanning claim. It's just a directional
`limitation.
` Turning to slide 25, this is the evidence we
`relied on in Burton. In Burton they have these rank values
`we talked about which considers both recency and frequency,
`similar to what happened with the Lee reference last time.
`They have a variable that considers both and they're looking
`for the variable with the lowest number, lowest number
`reflective of least recent, least frequent, and they're
`
`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
`26
`
`21
`
`
`
`Case IPR2017-00429
`Patent 6,775,745 B1
`
`looking for the 32 with the lowest rank.
` So I had mentioned before that we believe that
`Burton teaches the claim under any of the constructions,
`including the inclusivity construction which we don't think
`is correct, and the reason is that their analysis assumes
`that you're only dealing with large caches. That's what
`their analysis relies on, and as we mentioned in our papers
`we believe that Burton renders claim 6 unpatentable for small
`cache embodiments.
` So I'll turn to slide 27. This is some of the
`testimony that we talked about in the initial phase for a
`slightly different reason, but here we are talking to Dr.
`Chong about his experience with file caches and we ask him
`what's the smallest one you've ever seen and he says he's
`seen them as small as in the tens to hundreds, so there's
`certainly small file caches.
` Dr. Kubiatowicz agrees with that. We have that in
`our papers in Exhibit 1028, paragraph 122. Let me turn to
`this phase of the case, slide 28. We're asking Dr. Chong
`about the '745 claims to see is there any reason not to be
`considering small caches, and he agrees that the '745
`(inaudible) would cover caches as small as five blocks, so
`there's no reason that the claims would disqualify a small
`cache.
` They certainly read on a small cache and, you
`know, just to circle back the hypothetical that the patent
`
`1
`2