UNITED STATES PATENT AND TRADEMARK OFFICE
`__________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`__________
`
`EMC CORPORATION, LENOVO (UNITED STATES), INC.,
`and NETAPP, INC.,
`Petitioner,
`
`v.
`
`INTELLECTUAL VENTURES I, LLC,
`Patent Owner.
`__________
`
`Case IPR2017-00429
`Patent 6,775,745 B1
`__________
`
`Record of Oral Hearing
`Held: March 5, 2018
`__________
`
`
`
`
`Before JEFFREY S. SMITH, GEORGIANNA W. BRADEN, and DANIEL
`J. GALLIGAN, Administrative Patent Judges.
`
`
`
`
`
`
`
`
`
`
`
`
`__________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`__________
`
`EMC CORPORATION, LENOVO (UNITED STATES), INC.,
`and NETAPP, INC.,
`Petitioner,
`
`v.
`
`INTELLECTUAL VENTURES I, LLC,
`Patent Owner.
`__________
`
`Case IPR2017-00429
`Patent 6,775,745 B1
`__________
`
`Record of Oral Hearing
`Held: March 5, 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, ESQ.
`THEODOROS KONSTANTAKOPOULOS
`Wilmer Cutler Pickering Hale and Dorr, LLP
`60 State Street
`Boston, Massachusetts 02109
`(617) 526-6466
`peter.dichiara@wilmerhale.com
`
`
`ON BEHALF OF THE PATENT OWNER:
`
`
`BRENTON R. BABCOCK, ESQ.
`TED M. CANNON, ESQ.
`Knobbe Martens
`2040 Main Street, 14th Floor
`Irvine, California 92614
`(949) 760-0404
`brent.babcock@knobbe.com
`
`
`
`
`The above-entitled matter came on for hearing on Monday, March 5,
`
`2018, commencing at 1:30 p.m. at the U.S. Patent and Trademark Office,
`600 Dulany Street, Alexandria, Virginia.
`
`
`
`
`
`2
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`
`Case IPR 2017-00429
`Patent 6,775,745 B1
`
`
`P R O C E E D I N G S
`- - - - -
`
`
`
`1:29 p.m.
`JUDGE SMITH: Good afternoon. Welcome to the Patent Trial
`and Appeal Board. We're here this afternoon for an inter parties review
`matter, IPR2017-00429 for Patent 6,775,745, EMC Corporation, Lenovo and
`NetApp for Petitioner. Intellectual Ventures is the Patent Owner.
`I'd like to start by getting appearance of counsel. Petitioner, please
`step up to the podium and make your appearance.
`MR. DICHIARA: Hi, my name is Peter Dichiara from Wilmer
`Cutler Pickering Hale & Dorr. And with me today is Theodoros
`Konstantakopoulos, also from Wilmer Cutler Pickering Hale & Dorr. And
`we're representing the Petitioners in this matter.
`JUDGE SMITH: Welcome. Who do we have on behalf of the
`Patent Owner?
`MR. BABCOCK: Good morning, Your Honors. Brent Babcock
`with Knobbe Martens. With me is Ted Cannon, also with Knobbe Martens
`representing Patent Owner Intellectual Ventures.
`Also, is James Hietala and Tim Seeley, Counsel. And Intellectual
`Ventures are here with us.
`JUDGE SMITH: Thank you. And I would like to go over a few
`administrative details, quickly, before we begin. Our trial hearing order
`indicated there would be 30 minutes of argument for each side.
`
`3
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`
`Case IPR 2017-00429
`Patent 6,775,745 B1
`
`
`Petitioner, you'll go first presenting your case. Patent Owner, you'll
`be allowed to respond to Petitioner. And if Petitioner you wish to reserve
`time for rebuttal you may do so. Do you wish to reserve time for rebuttal?
`MR. DICHIARA: Yes, we will be. Approximately ten minutes,
`depending on the questions, Your Honor.
`JUDGE SMITH: Ten minutes for rebuttal, thank you. One more
`administrative detail. When you are referencing your slides, please identify
`the particular slide number so that our colleagues in Dallas can identify
`which slide you're referencing. And also, it makes it easier when we're
`reviewing the transcript to determine what slide you were on.
`Petitioner, when you are ready you may begin.
`MR. DICHIARA: Good afternoon. May it please the Board, as I
`mentioned, my name is Peter Dichiara and I'll be presenting on behalf of the
`Petitioners. With us today is Mr. Tom Brown for Petitioner EMC.
`Turning to Slide 2. We are here today to discuss the 745 patent and
`why the challenged claims aren't patentable. This patent concerns
`something known as a hybrid caching mechanism.
`Hybrid cache is considering both the recency and frequency of use of
`a file in determining which entries will stay in the cache and which ones will
`be evicted if space is needed.
`Turning to Slide 3. This is an annotated version that was in our
`petition concerning one of the embodiments and shows how this
`embodiment works in action.
`
`4
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`
`Case IPR 2017-00429
`Patent 6,775,745 B1
`
`
`And what you're seeing on the left-side of the figure in blue, is what
`the patent refers to as frequency factors. And the frequency factors indicate
`how often a file is being requested or accessed.
`And when the file was full and that cache needs to decide which
`entry in the cache should be evicted to make space for a new file, the 745
`embodiments will consider both the frequency and the recency of the cache
`entries to determine which entry essentially gets the boot.
`And what you're seeing here in this example, and as it's described in
`the patent, it's actually the orange entity F6 that is the one that's evicted.
`And you'll notice that that F6 is in fact not the absolute least recently used
`file in the whole cache, that would be F7. And it's not the absolute least
`frequently used file in the whole cache, that would be the one at top, F2.
`Instead, this embodiment and all of the embodiments in the 745
`patent try to consider both recency and frequency to evict something. But
`as both parties agree, a person of skill in the art would appreciate that there
`is no guarantee that the file is going to contain something that is the absolute
`least recent in the whole file, and simultaneously, the absolute least recent.
`So instead, the embodiments try to consider both factors in
`determining the eviction candidate. And they do that in various ways.
`So, if I can turn to Slide 4 I'd like to talk briefly about the file
`history. Because we think this provides some important context. As you
`know from our papers, the Petitioners maintain that the Lee reference
`discloses certain of the claim limitations.
`
`5
`
`
`
`Case IPR 2017-00429
`Patent 6,775,745 B1
`
`
`And the decision to institution, the Board likewise at least
`preliminarily agreed with that. And what we're seeing on this slide, and as
`we explained in our papers, is the Patent Office actually agreed with that
`during normal prosecution.
`And what happened during normal prosecution is the claims were
`rejected, the Applicant's response, rather than saying Lee doesn't teach
`frequency factors or least frequently and recently used file, instead they took
`a narrower claim. And they didn't even attempt to try and make these
`arguments that Lee doesn't teach it.
`So we think this is relevant, at least to the weight and the credibility
`of the arguments.
`If I can turn to Slide 5.
`JUDGE SMITH: Is it relevant beyond that in terms of estoppel or --
`MR. DICHIARA: We think there is law that would support it. We
`put that in our papers, the cases like Lemelson would support it. But we
`think it's at a minimum, relevant at least, to the weight that the testimony
`deserves in connection with Lee.
`JUDGE SMITH: Okay.
`MR. DICHIARA: So turning to Lee, on Slide 5, as its title says, it's
`an implementation and performance evaluation of something called the
`LRFU replacement policy. And LRFU is an acronym that stands for least
`recently frequently used.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`
`6
`
`
`
`Case IPR 2017-00429
`Patent 6,775,745 B1
`
`
`Slide 6, displaying some of the relevant disclosure, similar to the 745
`Patent. The Lee paper considers both recency and frequency of the cache
`entries.
`And it does so via these entities called CRF. Each of the files in the
`cache has a metric called CRF, which is an acronym for combined recency
`and frequency.
`And as its name suggests, that's a quantification of the files recency
`and frequency of the use. Each of the entries will have a different CRF
`value.
`
`And when space is needed, as you can see in the last block, when
`space is needed, the one that's evicted is the entity with the smallest CRF
`value. The smallest quantification of recency and frequency.
`Just briefly, these are the limitations that the Patent Owner has
`disputed or taught in its Patent Owner response. I'll turn it to the frequency
`factors in Slide 8 and I'll be brief.
`Judging from the Patent Owner's slides, it doesn't look like they're
`strongly pressing this issue any longer. I might be wrong, but I didn't see
`any slides on this so, on this opening part I'll be briefer than I otherwise
`would have. And if it is raised I'll circle back on rebuttal.
`The key thing here, as you see the claim limitation is that the
`frequency factors indicate, they indicate how often each of the
`corresponding files are accessed.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`
`7
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`
`Case IPR 2017-00429
`Patent 6,775,745 B1
`
`
`And as we maintained in our papers, and as you see in Slide 9, this is
`taught and disclosed by Lee's CRF values. The CRFs indicate how often
`each file is accessed. That's why F is in the name.
`And if I can turn to Slide 11, oh, strike that. I want to just say a
`brief word about Slide 10. Slide 10, I have some of the testimony from
`Patent Owner’s expert Dr. Chong.
`And at a high level what's going on here is he's agreeing that the
`frequency factors in the patent indicate. They don't calculate the frequency,
`they just indicate it. And they do it by not counting how often. You can't
`count how often, they do it by counting how many times a block is accessed.
`So, if I said you had blocks that were accessed ten times, you might
`question and say, okay, is that over the same period of time or not, are they
`different times and so forth. But it indicates frequency, it doesn't calculate
`frequency. That's the takeaway from this slide.
`So, let me turn to Slide 11. Slide 11 is our annotations of an
`exhibit that the Patent Owner's expert, Dr. Chong, created. And what we're
`showing here in this slide is exactly the way that Lee is just like one of the
`embodiments in the 745 patent that's called the self-tuning embodiment.
`Now, what you're seeing in yellow is that each time the file is
`accessed, the CRF value increments by one. That's the yellow annotation.
`And the prior, but at the same time the prior values, because of
`aging, they decrease. So if you look at red, that file was accessed before
`one time, but the value isn't one in red it's decreased to .9933.
`
`8
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`
`Case IPR 2017-00429
`Patent 6,775,745 B1
`
`
`And that's just like the self-tuning embodiment in the patent where
`the frequency factors increment with accesses and they decrease with age.
`And they do so, so that files don't get stuck in the cache.
`And we know that the claims should be covering an embodiment like
`that, both on, we believe the claim language. But even if you looked at the
`claim set, Claim 4 we think is broad.
`If you look at dependent Claim 7 it's saying, Claim 4 wherein itself
`tuning. We think ordinary understanding would be that Claim 4 should at
`least cover self-tuning in the other embodiments because Claim 7 is
`demanding that. So, we think we have a clear hit on the frequency factors.
`I'm going to jump ahead to Slide 17 because most of the other slides
`concern the disputes revolving frequency factor claim construction, which I
`didn't see in your slides.
`JUDGE SMITH: Before you jump in, just real quickly.
`MR. DICHIARA: Sure.
`JUDGE SMITH: If the claim 4 did not encompass the self-tuning
`embodiment, what would that do to the proposed grounds of unpatentability?
`MR. DICHIARA: We think we still hit it. I mean, the only reason
`we bring up the self-tuning is there was a denial about whether we had the
`frequency factors and we're like that's crazy. Pardon my language.
`It's not only that the claim reads on it where like one of your
`embodiments that are properly covered. So I don't know how we would get
`Claim 4 not to cover self-tuning given Claim 7.
`
`9
`
`
`
`Case IPR 2017-00429
`Patent 6,775,745 B1
`
`
`But if you didn't want to go there we think it's just a straight reading
`that CRF, as the name says, and you can see it on Slide 17 on that top blurb,
`it's indicating frequency. That's why its name is CRF. That's what the F
`stands for.
`So I'll turn to the other claim limitation, if that makes sense. So the
`other limitation they dispute is a least frequently and least recently used file.
`And as the claim says, this phrase actually shows up in the
`identifying step. You are identifying a least frequently and least recently
`used file.
`And as we explained in our papers, we think we satisfy it under any
`of the constructions in the case. As we put forth consistent with the way it
`is in the decision institute, we think we satisfy it even under IV's newly
`proposed construction.
`And the reason we satisfy is, again, getting back to this last blurb on
`the screen, that the entity that is evicted in Lee, is the file or the block with
`the smallest CRF value.
`And recall that CRF stands for combined recency and frequency.
`It's a quantification of recency and frequency. And the one evicted is the
`one with the smallest quantification of recency and frequency.
`And I'd like to circle back that Lee's paper is called LRFU for a
`reason. It's concerned with less recently frequently used files. Those are
`the guys who are evicted.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`
`10
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`Case IPR 2017-00429
`Patent 6,775,745 B1
`
`
`Let me turn to Slide 18. And this displays IV's proposed
`construction for the term that they made in their Patent Owner response.
`And all of their arguments rely on this construction.
`As we explained in our papers, we think this construction is
`improper for a few reasons and should be rejected. We think the analysis
`that relies on it, in the POR, should likewise be rejected.
`But again, at the end of the day it doesn't really matter that much
`other than we think it's not good for the record. Because Lee satisfies even
`this construction.
`So let me turn to Slide 19 and deal with one of the ways in which we
`think this claim construction is improper. And that's namely that it's
`inconsistent with the spec. Violates BRI.
`And we now know after cross examination that both experts agree
`that the 745 patent has an embodiment where the MRU is an eviction
`candidate. And we know that the federal circuit has confirmed that claim
`construction should cover preferred embodiments.
`It's rarely, if ever correct, to exclude preferred embodiments. But
`it's much stronger than that here. We say Claim 4 is broad enough to cover
`it because they're considering everything up to the MRU. We think, as we
`said in our papers, we think that's reinforced by Claim 6.
`And then you can turn to Claim 8 and Claim 8 shows that the
`patentee knew how to write a claim that would avoid the MRU if that was
`the intent. Because if you look at Claim 8's language, they're making it
`specific. We're not considering the MRU.
`
`11
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`Case IPR 2017-00429
`Patent 6,775,745 B1
`
`
`So we think this is covered just by looking at the claims. And we
`think this is covered by the spec. And that their construction that they're
`proposing in that last sentence in their construction is inconsistent with that.
`And that's one reason alone that this construction should be rejected.
`If I turn to Slide 20, here is some of the testimony I was talking
`about. This wasn't in their papers. This came out from cross examination
`of their expert Dr. Chong.
`You can tell from the questioning we're talking about the very same
`embodiment. When I turn back to Slide 19, the very same embodiment
`we're talking about in Slide 19.
`Slide 20 I'm asking him about that. And I say, Doctor, there is at
`least one embodiment that you would concede that the MRU is an eviction
`candidate and he agrees. And he's right about that.
`Let me turn to Slide 21 briefly. Because this isn't the only reason
`that we think the construction --
`JUDGE SMITH: You have about five minutes left.
`MR. DICHIARA: I will only need five.
`JUDGE SMITH: Yes. If I could real quickly, let's say we accept
`your premise that the MRU is within the scope of the claim, why is that an
`issue we are, is that --
`MR. DICHIARA: I know --
`JUDGE SMITH: -- is that construction necessary in order to --
`MR. DICHIARA: We don't think it's necessary, Your Honor. We
`think it's not a good construction, it wouldn't be good for the record. But
`
`12
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`
`Case IPR 2017-00429
`Patent 6,775,745 B1
`
`even if it were adopted, it would be incorrect to say that Lee evicts the
`MRU.
`
`And I can turn to Slide 22 to make this point. After they proposed
`their construction they tried to deny Lee has it, based on a hypothetical that
`they created, five-entry cache. And they created some access patterns.
`I have a blurb from the expert here. You would see in Dr. Chong's
`written testimony, Paragraph 86, he would say essentially the same thing.
`That they are contending that Lee fairly often evicts the MRU.
`Let me just pause on that for a moment, okay, because what they're
`saying is that sometimes it evicts the MRU. They're not saying all the time.
`And we have a method claim here. And under the law, from cases
`like Unwired Planet, that's sufficient. If you sometimes satisfy a method
`claim, that's enough. That's what Unwired Planet stands for.
`Now, I'll turn to Slide 22 because they are -- Slide 23, strike that.
`For the remote to Slide 23, their facts are wrong. That's not the way Lee
`works.
`
`Their thesis statement that Lee evicts an MRU, sometimes, fairly,
`often, whatever, is wrong. What you're seeing in the upper right is a picture
`from Lee, particularly for when they're describing evicting an entity from the
`cache.
`
`And what you see here is their implementation uses two structures,
`link list in red, the heap in blue. And the way it works, and they describe it
`is the first thing, Number 1.
`
`13
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`Case IPR 2017-00429
`Patent 6,775,745 B1
`
`
`So you're trying to bring a new file into the cache and need to make
`room. The first thing they do is get rid of the file with the smallest CRF
`value. Completely consistent with the text I put on the screen earlier.
`But where does that file live? It's at the left end of the link list.
`And just adjacent to that you have the testimony from Dr. Chong, again, this
`came out on cross examination.
`And the one at the left end of the link list is the guy who was
`released, and he agrees, yes, that's right. That's the first operation.
`Once that happens, an entity is moved from the heap onto the right
`end of the link list. That's Number 2 in the figure. They use the term
`demoted. It really just means moved.
`And then after that, now that there is some room in the heap, the new
`file goes into the heap. That's the yellow disclosure.
`So the most recently used file, the one that just came in, is in the
`heap. It's not on the link list, it's certainly not on the left end of the link list.
`And it's just incorrect to come up with a hypothetical that so flatly
`contradicts what Lee expressly disclosed.
`So, unless there are questions on that, I would reserve the rest of my
`time for rebuttal.
`JUDGE SMITH: Okay, you have two minutes left.
`MR. DICHIARA: Is it okay for me to take those two other on
`rebuttal?
`JUDGE SMITH: Does anybody have any questions?
`MR. BABCOCK: No.
`
`14
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`Case IPR 2017-00429
`Patent 6,775,745 B1
`
`
`JUDGE SMITH: Okay.
`MR. DICHIARA: Thank you, Your Honor.
`MR. BABCOCK: Good afternoon, Your Honors. Complainants
`present, Petitioner's presentation is very much like the petition, very
`superficial without actually looking at the words of the claim and
`understanding what the claim requires.
`This is Petitioner's slide which identifies the issues raised by the
`Patent Owner. And this is close to being correct. There are two key --
`JUDGE BRADEN: I'm sorry, what slide are you on?
`MR. BABCOCK: Slide 7 of Petitioner's slides. Two key --
`JUDGE BRADEN: Okay, thank you.
`MR. BABCOCK: Two key limitations at issue. Not just frequency
`factors but an entire limitation.
`Petitioner simply says frequency factor takes into account, takes into
`account frequency. That's not what the claim says. We'll look at it in a
`second.
`What about a least frequently and least recently used file. The claim
`requires that twice. That you identify it and then you actually eliminate it.
`The third thing that's going to be important here for the Panel is to
`recognize that there needs to be an implementation of Lee that does both of
`these at the same time. Petitioner has not provided a single example of Lee
`in which both of these limitations are satisfied in a given implementation.
`Again, Lee is an implementation where you pick a lambda and a
`cache size and you come up with a cache structure. We aren't, it's
`
`15
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`Case IPR 2017-00429
`Patent 6,775,745 B1
`
`Petitioner's burden to show that Lee satisfies every limitation of this claim in
`an implementation.
`There is not a single example that Petitioner has provided to this
`Panel to show where every limitation is satisfied in Lee. They take each
`limitation separately, wave some hands, but they never show this Panel why
`Lee meets every limitation.
`And in fact, we will show you beyond our burden --
`JUDGE BRADEN: What about Lee, at 107 and 108, where it
`discusses the combined recency and frequency value, the CRF, and then
`discusses the least recently frequently used replacement policy where it
`kicks out the, I guess it's the file that has the CRF value that's the smallest in
`the cache. How does those not meet both of those limitations?
`MR. BABCOCK: So, Your Honor, that's a very good question.
`Let me jump to that discussion of Lee. And I'm going to jump all the way -
`-
`
`JUDGE BRADEN: And are we now moving to Patent Owner slides
`or are we still --
`MR. BABCOCK: We're now on Patent Owner slides.
`JUDGE BRADEN: Very good, thank you.
`MR. BABCOCK: Lee teaches that you pick a lambda --
`JUDGE SMITH: Wait, which slide are you on?
`JUDGE BRADEN: And what slide are you on?
`MR. BABCOCK: Slide 31.
`JUDGE BRADEN: Thank you.
`
`16
`
`
`
`Case IPR 2017-00429
`Patent 6,775,745 B1
`
`
`MR. BABCOCK: So, Lee teaches that you pick a lambda and then
`you, for a given cache size. In these examples where Lee picks a lambda
`around ten to the minus four, you're going to be able to include the MRU in
`the file.
`The MRU is going to be part of the analysis that's going to be
`included in the eviction statements.
`JUDGE BRADEN: So?
`MR. BABCOCK: Well --
`JUDGE BRADEN: I don't see anything in Claim 1 that discusses a,
`I'm sorry, where --
`MR. BABCOCK: Okay.
`JUDGE BRADEN: Okay, where are we, I guess I still don't see
`where --
`MR. BABCOCK: Okay.
`JUDGE BRADEN: -- that's going to be an issue.
`JUDGE SMITH: Yes, I --
`JUDGE BRADEN: I mean, before.
`MR. BABCOCK: Well, Your Honor, you jumped to the second
`issue which is whether or not Lee teaches what's claimed. The first issue is,
`what does the claim mean.
`What does it mean to identify the least frequently and least recently
`used file and then to evict that least recently and least recently used file.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`
`17
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`
`Case IPR 2017-00429
`Patent 6,775,745 B1
`
`
`Does it mean just to evict a file that has frequency and recently as a
`factor? Is CRF where frequency and recency are simply a factor, a
`component? Does that meet that limitation?
`That's not what the claim says. The claim could, I'm going to go to
`Slide 7. Our Slide 7.
`JUDGE BRADEN: Okay, I was going to say, then point, in Claim
`4, where does it discuss not using the, you know, disregarding the most
`recently --
`MR. BABCOCK: It says --
`JUDGE BRADEN: -- used file. Because I can see Claim 8 which
`specifically says, without considering a frequency factor corresponding to
`the most recently used file, but I don't see a similar limitation in Claim 4.
`MR. BABCOCK: Your Honor, I've got Claim 4 up, Slide 7. It
`says, identifying a least frequently and least recently used file. And then it
`says, evict the least frequently and least recently used file.
`So we need to look at the spec to say, what does it mean to be the
`least frequently and least recently used file. We either construed that to
`where it doesn't say evict or identify a file based upon frequency and
`recency, we could have said that.
`But it says you first assign a frequency factor. Not any frequency
`factor, the factor must do something. It must indicate how often, or as
`Petitioner said, how recently each of the corresponding files are requested by
`the operating system.
`
`18
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`Case IPR 2017-00429
`Patent 6,775,745 B1
`
`
`JUDGE SMITH: It also, the claim also doesn't say, without
`considering a frequency factor corresponding to a most recently used file.
`It could have said that too. In fact, you did say that in Claim 8.
`MR. BABCOCK: No, the Claim --
`JUDGE SMITH: Claim 8 doesn't say anything?
`MR. BABCOCK: Well, it excludes the scanning. But the issue
`here is, the first claim construction --
`JUDGE BRADEN: In Claim 4, when you're talking about
`assigning the frequency factors it says the frequency factor indicating how
`often --
`MR. BABCOCK: That's right.
`JUDGE BRADEN: -- each of the corresponding files. It doesn't
`say how recently.
`MR. BABCOCK: I understand. So that's a separate limitation.
`JUDGE BRADEN: Okay.
`MR. BABCOCK: Has to be satisfied along with the identifying the
`least frequently, a least recently used file. So, to do it stepwise, Your
`Honor, you first have to construct these claim terms that are in dispute.
`And the question is, does the identification of a least frequently and
`least recently used file, the eliminating a least frequently, a least recently
`used file, doesn't permit the elimination of the most recently used file.
`If I said, I'm going to eliminate the least frequently and least recently
`used, can I submit the most recently used. And the answer is, in accordance
`with this specification the answer is no.
`
`19
`
`
`
`Case IPR 2017-00429
`Patent 6,775,745 B1
`
`
`Now, the task of this Board is to construe the claims in light of the
`specification. It's the broadest reasonable, but not the broadest possible.
`And I will show you that every single limitation in this patent explicitly
`avoids using the MRU.
`JUDGE SMITH: Is that, before we go on, is that the construction
`we should give to this term because earlier, in the related petition for this
`same patent, you gave us a construction of this term that is, I'm looking for
`the language here. But it was, indicate with, indicate with a fair degree of
`certainty --
`MR. BABCOCK: That's the --
`JUDGE SMITH: -- with a fair degree of certainty, is that --
`MR. BABCOCK: -- frequency factor.
`JUDGE SMITH: Is that the construction, because the construction
`you're giving us now is not that, the construction you're giving us now is,
`any file, as long as it's not the most recently --
`MR. BABCOCK: I'm sorry, Your Honor, with all due respect,
`you're confusing frequency factor up in the middle with least recently and
`least recently, least frequently used file.
`We're not changing our definition or our construction for assigning a
`frequency factor. And not just any frequency factor, it has to meet the
`claim, it has to indicate how often each of the corresponding files are
`requested by the operating system.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`
`20
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`Case IPR 2017-00429
`Patent 6,775,745 B1
`
`
`And then we'll say, with some degree of certainly. With a
`reasonable degree of certainty. And that was a dispute we had with the last
`case.
`
`JUDGE SMITH: Right.
`MR. BABCOCK: We haven't changed that. We're focusing now
`on two different limitations. Not the frequency factor of limitation but
`we're focusing on the least frequently and least recently used file. And
`what does that mean in light of the specification.
`And the specification has many embodiments, and I'll show them
`quickly. And we'll go through these embodiments and I will show you that
`every single time this patent discusses how this algorithm works, it always
`excludes the MRU from being eliminated.
`Why is that, because by including the MRU you have something
`called thrashing. And the patent teaches against thrashing.
`First, let's start with ordinary meaning. Petitioner wants to say, hey,
`least can be most. The least recently used file can't be the most recently
`used. Well, that kind of goes against commonsense.
`So at least they should be able to say, well, in light of the
`specification, there is some teaching in which the MRU is picked. And they
`can't, the only thing they point to is situations where the MRU is scanned but
`it's never, and I'll repeat, never elected and eliminated from the cache.
`Figures 2a and --
`JUDGE BRADEN: Okay, Counselor, help me with this because I'm
`looking at the specification and in each of those instances it's very specific to
`
`21
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`
`Case IPR 2017-00429
`Patent 6,775,745 B1
`
`say, in one embodiment of the invention, the frequency factor for the MRU
`file is not considered when determining the LFU file. I mean, this is
`Column 6 looking at Lines 39 through 44.
`So, how do we get away from the very narrowing language that's in
`the specification that limits it to specific embodiments, to preferred
`embodiments, in one embodiment?
`MR. BABCOCK: Because the patent repeatedly says, in this
`embodiment, in this, it provides many different embodiments. And I agree.
`JUDGE BRADEN: It always uses that type of language to try and
`get it, say, in this embodiment, in this one embodiment, in the preferred
`embodiment. But then in the claim language it appears to use much broader
`language and doesn't have that limiting language of what we see in Claim 8.
`How do we get around that?
`MR. BABCOCK: The way you get around it, Your Honor, is you
`do what the federal circuit has said you do. You look at the specification
`and you look for any teaching that would contradict the plain and ordinary
`meaning of this claim.
`This claim says --
`JUDGE BRADEN: But aren't you going against federal circuit
`guidance --
`MR. BABCOCK: No.
`JUDGE BRADEN: -- that prohibits us from improperly importing
`limitations into the claim?
`
`22
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`
`Case IPR 2017-00429
`Patent 6,775,745 B1
`
`
`MR. BABCOCK: Every single embodiment says, don't pick the
`MRU. And we have slide after slide saying, don't pick the MRU. Why
`didn't you pick the MRU, because you have thrashing.
`The MRU comes in, it's got the least recently, the least frequently.
`It's the next one evicted so the next file that comes in is going to be the least
`recently.
`And you end up just cycling through that first cache entry just
`repeatedly, repeatedly.
`JUDGE SMITH: Okay.
`MR. BABCOCK: And that's where the patent discusses is
`thrashing, it says you don't want to do this.
`So if you have a situation, and the federal circuit has made this clear,
`if every embodiment says this is what we mean, this is how it works, and
`you're looking at the most, the simple ordinary meaning of least and you're
`trying to say, well, least could be most, you need to have some teaching --
`JUDGE SMITH: I'm sorry to interrupt, because I want to be clear
`on this point. I mean, the federal circuit says, absent an express definitio
Accessing this document will incur an additional charge of $.
After purchase, you can access this document again without charge.
Accept $ Charge
Still Working On It
This document is taking longer than usual to download. This can happen if we need to contact the court directly to obtain the document and their servers are running slowly.
Give it another minute or two to complete, and then try the refresh button.
A few More Minutes ... Still Working
It can take up to 5 minutes for us to download a document if the court servers are running slowly.
Thank you for your continued patience.
This document could not be displayed.
We could not find this document within its docket. Please go back to the docket page and check the link. If that does not work, go back to the docket and refresh it to pull the newest information.
Your account does not support viewing this document.
You need a Paid Account to view this document. Click here to change your account type.
Your account does not support viewing this document.
Set your membership
status to view this document.
With a Docket Alarm membership, you'll
get a whole lot more, including:
- Up-to-date information for this case.
- Email alerts whenever there is an update.
- Full text search for other cases.
- Get email alerts whenever a new case matches your search.
One Moment Please
The filing “” is large (MB) and is being downloaded.
Please refresh this page in a few minutes to see if the filing has been downloaded. The filing will also be emailed to you when the download completes.
Your document is on its way!
If you do not receive the document in five minutes, contact support at support@docketalarm.com.
Sealed Document
We are unable to display this document, it may be under a court ordered seal.
If you have proper credentials to access the file, you may proceed directly to the court's system using your government issued username and password.
Access Government Site
