`
`
`
`
`IN THE UNITED STATES DISTRICT COURT
`IN AND FOR THE DISTRICT OF DELAWARE
`- - -
`
`
`
`1
`
`CIVIL ACTION
`
`NO. 18-966 (CFC)
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`:::::::::
`
`VLSI TECHNOLOGY LLC,
` Plaintiff,
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
` vs.
`INTEL CORPORATION,
` Defendant.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`1 2 3 4 5 6 7 8 9
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
` - - -
` Wilmington, Delaware
` Tuesday, November 6, 2019
` 9:00 o'clock, a.m.
` - - -
`BEFORE: HONORABLE COLM F. CONNOLLY, U.S.D.C.J.
` - - -
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`APPEARANCES:
`
` FARNAN LLP
` BY : BRIAN E. FARNAN, ESQ.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
` Valerie J. Gunning
` Official Court Reporter
`
`1 of 77 sheets
`
`Page 1 to 1 of 193
`
`11/08/2019 01:57:12 PM
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
` -and-
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
` VLSI Ex. 2013-p. 1
`
` IPR2019-01196
`
`
`
` P R O C E E D I N G S
`
`4
`
` (Proceedings commenced in the courtroom,
`beginning at 9:00 a.m.)
`
`THE COURT: Good morning. Please be seated.
`(Counsel respond, "Good morning, Your Honor.")
`THE COURT: All right. So I guess, Mr. Farnan?
`MR. FARNAN: Yes, Your Honor. Good morning.
`THE COURT: Welcome.
`MR. FARNAN: Good morning, Your Honor. Brian
`Farnan on behalf of the plaintiff. And with me is Chris
`Abernathy, Dominik Slusarczyk, Adina Stohl and Ben
`Hattenbach, all from Irell Manella.
`THE COURT: All right. Thank you.
`MR. FARNAN: Thank you.
`THE COURT: Mr. Blumenfeld?
`MR. BLUMENFELD: Good morning, Your Honor.
`THE COURT: Good morning.
`MR. BLUMENFELD: Jack Blumenfeld from Morris
`Nichols for Intel.
`At counsel table are Gregory Lantier and Dominik
`Massa from Wilmer Hale. Behind them is Mashood Rassam from
`Intel. Behind Mr. Rassam is Kimberly Schmidt, also from
`Intel. And in the corner, Richard Crudo and Calvin Walden
`
`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
`
`APPEARANCES (Continued):
`
`2
`
` IRELL & MANELLA LLP
`BY: CHRISTOPHER ABERNETHY, ESQ.
` DOMINIK SLUSARCZYK, ESQ.,
` S. ADINA STOHL, ESQ. and
` BEN HATTENBACH, ESQ.
` (Los Angeles, California)
`
` Counsel for Plaintiff
`
` MORRIS, NICHOLS, ARSHT & TUNNELL LLP
` BY: JACK B. BLUMENFELD, ESQ.
`
` -and-
`
` WILMER CUTLER PICKERING HALE AND DORR LLP
`BY: DOMINIC E. MASSA, ESQ.,
` (Boston, Massachusetts)
`
` -and-
`
`WILMER CUTLER PICKERING HALE AND DORR LLP
`BY: JOSHUA L. STERN, ESQ.
` RICHARD A. CRUDO, ESQ.
` (Washington, D.C.)
`
` -and-
`
`WILMER CUTLER PICKERING HALE AND DORR LLP.
`BY: S. CALVIN WALDEN, ESQ.
` (New York, New York)
`
` -and-
`
`APPEARANCES (Continued):
`
` INTEL CORPORATION
` BY: MASHOOD RASSAM, ESQ.
`
` Counsel for Defendant
`
` - - -
`
`3
`
`5
`
`1
`from Wilmer Hale. And behind them, Joshua Stern, also from
`2
`Wilmer Hale.
`3
`And with Your Honor's permission, Mr. Lantier
`4
`and Mr. Massa and I are going to split the presentation
`5
`today. And I can -- well, I will wait until we start before
`6
`handing up our slides.
`7
`THE COURT: All right.
`8
`MR. BLUMENFELD: Thank you.
`9
`THE COURT: Thank you very much.
`10
`MR. BLUMENFELD: Oh, and, Your Honor, Mr. Lee
`11
`wanted to be here today. He's two floors up trying a case
`12
`before Judge Stark, so was unable to be here.
`13
`THE COURT: Okay. Judge Stark has two trials
`14
`today going on.
`15
`All right. Why don't we begin? Let's go claim
`16
`by claim, or term by term, I should say.
`17
`So I got this joint claim construction chart
`18
`with many, many attachments, and then I got the joint
`19
`appendix in support of the joint claim construction brief,
`20
`so when I was just really looking at the appendix of the
`21
`brief just for clarity in case anybody wants to know, I
`22
`basically ignored the attachment to the joint claim
`23
`construction chart.
`24
`So the other thing is, I didn't look at any
`25
`extrinsic evidence, and so for each term limitation at the
`Page 2 to 5 of 193
`2 of 77 sheets
`
`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
`
`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
`
`11/08/2019 01:57:12 PM
`
` VLSI Ex. 2013-p. 2
`
` IPR2019-01196
`
`
`
`74
`
`76
`
`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
`
`addresses -- just leave it as it says, they don't get
`further construction?
`MS. LANTIER: We would prefer that over VLSI's
`proposed construction.
`THE COURT: VLSI says they can live with that.
`That's how I'm not going to construe it. I'm not going to
`construe the term. VLSI has already said they can live with
`that. And it's not your first choice, Intel, but you can
`live with it. Right?
`MS. LANTIER: It's not our first choice. I
`understand Your Honor's ruling. I am not disputing it. For
`the record, we would preserve our position that the right
`construction is the one that we proposed.
`THE COURT: That's understood.
`MR. LANTIER: Yes.
`THE COURT: VLSI already said at the outset they
`don't need construction of this material, so I'm going to go
`with VLSI's proposed or secondary proposal, which is --
`actually, their primary proposal, it did not need
`construction, I will go with that, and you can argue your
`position to the Federal Circuit if need be. And, of course,
`if something comes to my attention, because it sounds like
`there's going to be expert testimony about virtual
`addressing, then I will revisit claim construction at that
`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
`25
`
`other phrases of the claim language as well for which
`construction is not being sought.
`So just starting with this term determining,
`VLSI proposes that no construction is needed for --
`THE COURT: Actually, can I just ask you: Could
`you live with quantifying or quantify?
`MS. STOHL: Your Honor, I don't believe that
`that would be an appropriate construction here.
`THE COURT: So you wouldn't agree to that?
`MS. STOHL: We would not agree to that at this
`
`time.
`
`THE COURT: Okay.
`MS. STOHL: And I'm happy to explain --
`THE COURT: Yes.
`MS. STOHL: -- further why.
`So here -- well, first of all, aside from the
`fact that determining is the plain and ordinary understood
`word, in fact, as we noted in the briefs, they don't
`actually argue for the term determining to be the
`construction of other terms in other cases.
`Intel's proposal of sense or sensing is much
`more limited. It's a subset of determining and doesn't stay
`honest to the patent.
`Specifically, Intel's proposal imports
`limitations from a preferred embodiment. Here, so the
`
`75
`
`77
`
`MS. LANTIER: Yes, Your Honor. I understand.
`THE COURT: VLSI is good with that? You
`understand my ruling?
`MR. ABERNATHY: Yes, Your Honor.
`THE COURT: All right. Let's go to the next
`
`term.
`
`record?
`
`MS. STOHL: Good morning, your Honor.
`THE COURT: What's your name just, for the
`
`1
`1
`2
`2
`3
`3
`4
`4
`5
`5
`6
`6
`7
`7
`8
`8
`9
`9
`10
`10
`MS. STOHL: Adina Stohl for the plaintiff.
`11
`11
`THE COURT: Ms. Stohl? Okay. Thank you.
`12
`12
`MS. STOHL: THE '027 patent discusses a
`13
`13
`determination of an analog variation parameter that is
`14
`14
`representative of an integrated circuit fabrication process
`15
`15
`variance of the integrated circuit, and an operational
`16
`16
`temperature as well as a digital variation parameter with
`17
`17
`the purpose being to optimize power consumption on an
`18
`18
`IC-by-IC basis.
`19
`19
`Intel is taking construction of three terms for
`20
`20
`this patent. The first two are very similar and are put
`21
`21
`together. Determining an analog or a digital variation
`22
`22
`parameter, and the third being determining an operational
`23
`23
`MS. STOHL: So I would say that what they do is
`temperature.
`24
`24
`they -- you do some sort of analyzation, whatever that may
`As Your Honor will notice, all three of these
`25
`25
`terms use the word determining, which is a word used in
`be. So it might be -- it might be, you know, studying and
`11/08/2019 01:57:12 PM
`Page 74 to 77 of 193
`20 of 77 sheets
`
`specification is replete with numerous ways in which
`determining is discussed. Intel hones in on the word
`sensing sometimes with determining, sometimes without, but
`as Your Honor will notice, the inventors knew the word
`sensing. In fact, they used it in the specification, but
`when it came time to actually claim their invention, they
`chose to use the broader word determining. You see it here
`in claim 1.
`
`THE COURT: Help me out. I'm a juror and I'm
`reading your claim. An analog variation parameter. You're
`accusing Intel of doing that. What are you going to say
`Intel does to determine an analog variation parameter?
`MS. STOHL: There are -- we have multiple
`theories of infringement, Your Honor, but specifically, the
`separate issue, but an analog variation parameter which will
`be discussed shortly as another --
`THE COURT: Help me out with the verb. What are
`you going to say to the jury? Ladies and gentlemen, Intel
`determines in implementing this method for power supply
`optimization, it determined an analog variation parameter.
`What is it they do? What are you going to tell
`
`the jury?
`
` VLSI Ex. 2013-p. 3
`
` IPR2019-01196
`
`
`
`78
`
`80
`
`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
`
`reaching a conclusion. It might be --
`THE COURT: Wait. They studied study and reach
`a conclusion?
`MS. STOHL: There are multiple examples.
`THE COURT: When you are in front of the jury,
`what are you going to say to the jury when you say, ladies
`and gentlemen, Intel infringed this claim because they
`determined an analog variation parameter. Here's what they
`did to determine an analog variation parameter. Tell me,
`what are you going to tell them?
`MS. STOHL: So one of the options is that they
`went and they read that parameter.
`THE COURT: They read it. What does that mean?
`MS. STOHL: They read the value of it.
`THE COURT: So read, they measured it?
`MS. STOHL: It can be a measurement.
`THE COURT: Well --
`MS. STOHL: And the patent explicitly includes
`that as measuring.
`THE COURT: What else?
`MS. STOHL: It can be a calculation. It can be
`a generation as discussed in determining --
`THE COURT: They detected it?
`MS. STOHL: They detected it. If I just --
`THE COURT: So why not go with measure?
`
`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
`
`THE COURT: But you want to read determining to
`mean causing. Isn't that really what you want?
`MS. STOHL: As causing?
`THE COURT: Yes. Like in the passive.
`MS. STOHL: No, I would not agree with that,
`Your Honor.
`THE COURT: No?
`MS. STOHL: And then part of my pushback on
`measuring is that, well, the term is also used in relation
`to the adjustment signal. That's not something that can be
`measured. Right?
`Looking at the claim language, the adjustment
`signal is actually determined using the analog variation
`parameter in relation to the operational temperature and/or
`with the digital variation parameter. None of that can be
`done through measuring. It's a metric that didn't exist
`before.
`
`Can you put me back at 68, please.
`Your Honor, may I continue?
`THE COURT: I'm sorry?
`MS. STOHL: May I continue?
`THE COURT: Yes, please.
`MS. STOHL: So here what is very clear is going
`back to the same principle, determining does not mean
`sensing. To limit it would be improper.
`
`79
`
`81
`
`MS. STOHL: Because --
`THE COURT: What is the difference between read
`and measured?
`MS. STOHL: So measure I agree is encompassed
`within the term determining, but it doesn't define it.
`THE COURT: I'm on read.
`MS. STOHL: Sorry?
`THE COURT: I'm on read. You are going to argue
`to the jury they read some parameter. Right?
`MS. STOHL: Well, I would argue that the
`claim word determining is used repeatedly throughout the
`patent.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`And to do that, we have to employ some analysis.
`17
`We have to do something. So what do you have to do to
`18
`determine an analog variation parameter?
`19
`MS. STOHL: So here, just as an illustrated
`20
`example, kind of, there are all of these options that can be
`21
`determining, and the patent actually uses these other words
`22
`for determining. So to limit it to any one of these words,
`23
`I would actually argue is improper. Sensing is one way
`24
`which determining is used. Measuring is another, as Your
`25
`Honor stated.
`21 of 77 sheets
`
`THE COURT: I know. We're trying to determine
`what it means. We're trying to determine what it means.
`Right?
`
`1
`Intel's proposal also restricts the claim scope
`2
`contrary to law. What's very clear is that varied use of a
`3
`disputed term demonstrates the breadth of the term rather
`4
`than providing a limited definition. So I mentioned a
`5
`couple ways in which determining is discussed in the
`6
`specification, in the claims themselves. Those aren't
`7
`limiting and, in fact, again, are all just examples of the
`8
`way in which it can be done.
`9
`As one example here, we see in claims 5 and 11,
`10
`so it's the claims themselves, it requires determining a
`11
`digital variation parameter comprising measuring and
`12
`comparing. So determining includes comparing measured
`13
`values, something that's different from sensing.
`14
`Turning to the specification, we see that
`15
`determining can include counting iterations. We see that in
`16
`column 9, 6 through 27. As well we see that determining
`17
`includes calculating ratios and that's actually back in the
`18
`claims as well.
`19
`So claim 16 notes that determining the digital
`20
`variation parameter is done as a ratio. That also is
`21
`different than sensing. You don't sense a ratio.
`22
`Looking back again at the abstract, this is in
`23
`relation to the adjustment signal, and in column 12, 50 to
`24
`53, determining includes generating a signal. That's again
`25
`why we disagree with the proposal as narrowing the word
`Page 78 to 81 of 193
`11/08/2019 01:57:12 PM
`
` VLSI Ex. 2013-p. 4
`
` IPR2019-01196
`
`
`
`82
`
`84
`
`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
`
`determining to just measuring.
`And it goes even further than that. Not only
`does it's restrict the word contrary to what -- to the
`claims themselves and contrary to law, but it actually goes
`a step further and renders the claims unintelligible.
`So as I mentioned a moment ago, determining
`includes measuring and comparing as two examples. Very
`simply, sensing does not compare those two.
`And for claim 16, determining as a ratio, again,
`a ratio is not sensing.
`The Federal Circuit has been clear on this, that
`if a construction would render a claim nonsensical, it
`simply can't be correct, and here restricting it to sensing
`or measuring would do that.
`I'm happy to move on to the next portion of this
`
`claim.
`
`THE COURT: Sure.
`MS. STOHL: So the next dispute that relates to
`these first two terms, determining an analog or a digital
`variation parameter is this idea of, well, what is
`variation?
`
`So VLSI proposes that an analog variation
`parameter is an analog portion of the integrated circuit.
`That means the area on an IC-by-IC base basis. Similarly,
`the construction is the same for digital variation
`
`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
`
`THE COURT: So are you going to external
`evidence now?
`MS. STOHL: It's not necessary. It's just to
`further explain.
`THE COURT: If it's not necessary, don't make
`your argument using it.
`MS. STOHL: On the flip side, Intel's proposal
`address a non-claim limitation. So specifically, their
`argument is that the claim uses the plain meaning of
`variation to refer to a value that changes over time. Their
`only citation for this plain meaning requiring a change over
`time is their own expert. Variation doesn't require that.
`There's a multitude of ways that things can vary. Sizes
`and shapes vary. None of that requires over time.
`THE COURT: Intel essentially, they are accusing
`you of this read only thing, that you are just trying to
`read a fixed value. Can you maybe explain to me what that
`means in the process?
`MS. STOHL: What they are arguing?
`THE COURT: Well, what does it mean? A fixed
`
`value?
`
`MS. STOHL: So the purpose --
`THE COURT: As you understand it. I realize
`it's their argument. I'm going to ask them the same thing.
`I'm just curious. Can you explain to a non-integrated
`
`83
`
`85
`
`1
`parameter, obviously related to the digital portion. Intel,
`2
`however, proposes that variation somehow requires that it
`3
`vary during operation. That simply just doesn't exist.
`4
`VLSI's proposal is taken directly from the
`5
`patent itself if you look at the abstract. Going back
`6
`again, what is the purpose of the determination of an analog
`7
`parameter representative of an integrated circuit
`8
`fabrication process? It is to optimize power consumption on
`9
`an IC-by-IC basis. Whether or not it varies over time is
`10
`immaterial. It can or it cannot, but it's not required to
`11
`do so. What's required simply is that it's optimized on an
`12
`IC-by-IC basis.
`13
`Intel will argue that the purpose of the claims
`14
`is this idea of determining adjustment signals, something
`15
`that happens over time. But the disconnect there is they're
`16
`ignoring that there's another element. They are ignoring
`17
`that there's an operational element from temperature or, for
`18
`example, there can be other elements that are expressed in
`19
`the variation parameters at all.
`20
`We're sitting in a room with walls. I can
`21
`determine the thickness of the walls, which has an effect on
`22
`the acoustics in this room, perhaps noise coming in from the
`23
`outside. That may change over time how much noise is in
`24
`this room, but it's not because of a change in the wall.
`25
`Dr. Conte, again --
`11/08/2019 01:57:12 PM
`
`circuit engineer, what does that mean?
`MS. STOHL: At the time that a chip is
`fabricated, there are a number of elements to that chip to
`identify that specific clip as different from other chips.
`So when a chip is fabricated, it's fabricated on this large
`wafer. Other chips, there might be hundreds, there might be
`thousands on a wafer, they are supposed to be identical.
`They are printed and they are supposed to be really carbon
`copies of each other. In reality, that doesn't happen.
`There are small natural variations, natural errors that
`create differences between two chips that are supposed to be
`the same.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`What had previously been done is that all of the
`14
`chips on a single wafer or all of the chips on a single, in
`15
`a single batch were set to operate such that they would work
`16
`even for the chip that was kind of not necessarily the dud,
`17
`but the one that was a little bit more imperfect than the
`18
`others, and that might be that, you know, the wires are a
`19
`little bit wider, so it requires more energy to push
`20
`through, or the transistor shape and size was a little bit,
`21
`a little bit different.
`22
`And what happens at that time is you test all of
`23
`these chips, and what you see is, okay. Well, for this
`24
`particular chip, we notice that it does have this slight
`25
`variation and we're going to program that into this chip so
`Page 82 to 85 of 193
`22 of 77 sheets
`
` VLSI Ex. 2013-p. 5
`
` IPR2019-01196
`
`
`
`86
`
`88
`
`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
`
`that we know when we run this chip, this chip needs a little
`bit more power, for example, than another chip. Right.
`If we didn't separately identify the differences
`between those two chips, we would always make sure that we
`gave enough power to all the chips, to make sure that they
`were above the threshold that would work for chip A, B, C
`and D. If we don't look at it that way and we said, look at
`it, what does chip A need? It needs a power over here while
`chip B needs something lower, what we can do is set chip B
`to work at a lower level and essentially conserve power in
`that way.
`
`THE COURT: So where in that process are you
`determining the analog variation parameter?
`MS. STOHL: There are multiple theories, but the
`one in which, that Intel is specifically discussing --
`THE COURT: No, no. Where are you telling me
`it's occurring?
`MS. STOHL: I believe it occurred at two points.
`THE COURT: And what are they? Explain it to
`
`me.
`
`MS. STOHL: The first one being that during that
`process, when the chips are set to figure out, well, is chip
`A or chip B, which one requires more power? At that point
`you determine what the difference is between that chip, and
`that would be determine.
`
`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
`
`THE COURT: Okay. So the first point, you said
`it. When and how is the analog variation parameter
`determined in the first of the two steps?
`MS. STOHL: So in the first of the two --
`THE COURT: Actually, counsel, give her a
`chance. I'm commenting on somebody coming up. Judge
`Andrews had mentioned this to me. He has a rule apparently,
`you are not allowed to come up to the podium to help out
`somebody. Anyway, my comments are for the court reporter's
`benefit. So go ahead, if you can. I lost my train of
`thought.
`
`MS. STOHL: Sure. So the first one is when it's
`determined when measured in the manufacturing process.
`Again, you have this wafer.
`THE COURT: Measured and written? So basically,
`are you saying that when somebody decides what spec to set,
`that's a determination of the analog variation parameter?
`MS. STOHL: It's not a spec. There are fuses in
`chips, so there are particular values that are fused or set
`into the chip. They are not set arbitrarily. These ones
`would need to be set to account for variations between one
`chip and another. Right.
`Some fuses are set consistently across all
`simply by virtue of, I don't know. I mean, I assume they
`all have something that says Intel in there, for example,
`
`87
`
`89
`
`THE COURT: How do you make that determination?
`MS. STOHL: Through the fabrication process.
`THE COURT: That's where I lose you.
`MS. STOHL: So when you are making these chips,
`you make it on a big wafer. The wafer has a couple hundred
`to a couple thousand chips on a wafer. You dice them up.
`You then have each little chip that goes through a nice
`little assembly line and it tests them. And at that point,
`it's testing, well, what is -- what is this specific
`variation parameter associated with this chip? At that
`point it's determining it for the first time.
`There's another --
`THE COURT: How is it determining? What is it
`
`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
`23 of 77 sheets
`
`doing?
`
`MS. STOHL: What is it --
`THE COURT: And what is the it?
`MS. STOHL: The it is an assembly process. It's
`automated. There's a lot of machines that do it. You know,
`for example, at one point the machine applies different
`voltages and it sees kind of -- well, you know, what it's
`doing at different levels.
`THE COURT: Right. But you said there are two
`points in that process where it's determined what the analog
`variation parameter is. Am I using the right language?
`MS. STOHL: Yes.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`MS. STOHL: It does, Your Honor.
`14
`THE COURT: Okay. So at two points during this
`15
`process we're going to determine the analog variation
`16
`parameter. So I start with a wafer made of silicon. Is
`17
`that correct?
`18
`MS. STOHL: Yes.
`19
`THE COURT: And then I start laying various
`20
`layers of metal on it. Is that right?
`21
`MS. STOHL: Roughly, yes.
`22
`THE COURT: Okay.
`23
`MS. STOHL: It's a buildup. You put in a
`24
`transistor, you then build up the various layers of the
`25
`chip, yes.
`Page 86 to 89 of 193
`
`for Intel clips for another brand.
`THE COURT: Sorry. You've got to back up. I
`don't have your technical background.
`So you've got this die. We're going to put a
`thousand or so chips on it. We're going to make them in the
`machine.
`
`MS. STOHL: Correct.
`THE COURT: So far, so good?
`MS. STOHL: Yes.
`THE COURT: You said at two points during
`this process. This process involves a lot of steps. All
`right?
`
`11/08/2019 01:57:12 PM
`
` VLSI Ex. 2013-p. 6
`
` IPR2019-01196
`
`
`
`90
`
`92
`
`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
`
`THE COURT: Okay. So when do I get to this
`first determination of analog variation parameters?
`MS. STOHL: So after you take this wafer, you
`chop up, you chop it up into all the pieces, so each little
`chip or integrated circuit is a separate unit. It then goes
`through attesting process. During that testing process,
`there are measurements and calculations made on each of
`those chips.
`THE COURT: Okay.
`MS. STOHL: At that point, again, there's, you
`know, heat that's used in relation to it to find out what
`would be necessary at different points, but essentially, it
`finds out all of these elements --
`THE COURT: Parameters.
`MS. STOHL: Exactly. Parameters.
`THE COURT: So you're testing and measuring to
`find out the parameters of the specifications of all of the
`little chips on the dot? Is that fair?
`MS. STOHL: Of each particular chip.
`THE COURT: Okay. All right. Now, have we
`determined the analog variation parameters?
`MS. STOHL: Now you've determined it for the
`
`first time.
`
`THE COURT: Based on that testing and measuring?
`MS. STOHL: Yes.
`
`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
`
`You take this integrated circuit. You put it in a computer.
`You ship it out to a retailer? I mean, to a customer?
`MS. STOHL: I mean, it's could be anyone.
`THE COURT: Yes?
`MS. STOHL: It is when it is actually being
`
`used.
`
`THE COURT: When it's actually being used. Does
`it happen only once or does it happen every time the
`computer is turned on?
`MS. STOHL: It's accessed numerous times.
`THE COURT: Every single time the computer is
`
`used?
`
`MS. STOHL: I would assume so, yes.
`THE COURT: And so what happens then? Describe
`that process.
`MS. STOHL: So during various programs and
`calculations done by the, within this chip specifically to
`make sure that it's using the correct amount of power, it
`reaches out and it --
`THE COURT: What's the it?
`MS. STOHL: The central, like, processing units
`for this chip.
`THE COURT: Okay. Processor. All right.
`MS. STOHL: It reaches out and it grabs that
`value, which has now been fused into the chip, and at that
`
`91
`
`93
`
`THE COURT: So why don't I just adopt Intel's
`claim construction? I mean, when you describe the process,
`you talk about it, testing and measuring. I mean, what else
`is there?
`
`1
`1
`2
`2
`3
`3
`4
`4
`5
`5
`I think that's what Intel's concern is. That's
`6
`6
`certainly my concern, why don't we explain to the jury what
`7
`7
`determining means, because I get the impression, you know,
`8
`8
`that there's something lurking behind.
`9
`9
`MS. STOHL: So there the spec is time where the
`10
`10
`analog variation time is the determined.
`11
`11
`THE COURT: Okay.
`12
`12
`MS. STOHL: And that's after this entire
`13
`13
`fabrication process concludes.
`14
`14
`THE COURT: All right.
`15
`15
`MS. STOHL: So now when you have this chip and
`16
`16
`it sets all of its particular specifications and particular
`17
`17
`parameters during its use, so during its run time, it's
`18
`18
`determined again when the fuse, that calculation that was
`19
`19
`done previously is now retrieved, so it's now read. That
`20
`20
`parameter now is fused into the chip and you go and you
`21
`21
`retrieve it. You read it.
`22
`22
`MS. STOHL: It's retrieving or assessing, yes.
`THE COURT: So how does that work as a practical
`23
`23
`I don't know the value right now, so I have to go over there
`matter? Is the chip now in a computer?
`24
`24
`to find out what it is, to determine what it is, and that
`MS. STOHL: It would be, or some sort --
`25
`25
`THE COURT: You finish the fabrication process.
`over there is the fused value in the chip.
`11/08/2019 01:57:12 PM
`Page 90 to 93 of 193
`
`point, again, it's determining that analog variation
`parameter which has been fused into the chip.
`THE COURT: But it has already been fused into
`the chip. Isn't that what Intel's concern is? It's a fixed
`value. All you're doing is looking at a fixed value that's
`in the chip because it's the way you manufactured it.
`MS. STOHL: It is a fixed value. We agree with
`that at the second point.
`THE COURT: So what are you doing? What are you
`doing at that point? Other than retrieving it, what are you
`doing?
`
`MS. STOHL: Well, so I would say that that is
`encompassed by determining. You know, these walls are
`fixed. They've been here since, I don't know, since the
`courthouse has been built.
`THE COURT: That's why I said at the very
`beginning, you didn't really want to say it's causing. In
`other words, what are you doing at that point? You're
`retrieving, you're retrieving the fixed value. Is it
`retrieving? That's determining, just retrieving a fixed
`value?
`
`24 of 77 sheets
`
` VLSI Ex. 2013-p. 7
`
` IPR2019-01196
`
`
`
`94
`
`96
`
`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
`
`I would also argue that determining can't simply
`be read as Intel proposes, because the word determining is
`used throughout the claim, and that wouldn't make any sense
`in relation to, for example, determining an adjustment
`signal, which is something that previously didn't exist.
`Right? You can't measure something. You have to do some
`sort of calculations to get there.
`THE COURT: All right.
`MS. STOHL: And, again, I mean this was brought
`up earlier by Mr. Lantier in relation to deactivating, where
`he said that word was used elsewhere in the same claim and
`therefore should be used consistently. The same as here
`with determining. You can't construe it one way for one
`phrase within the same term and not for the other.
`So going back to the other portion of this
`variation idea, right. So variation, which is represented
`as this fabrication process, is something that as the patent
`discusses is pointed to, pointing out the potential
`differences between these different IPs, between these
`different little chips on that same wafer. It doesn't
`require that it change or vary during operation.
`So here, here are two examples. As I mentioned
`earlier, so a transistor shape and size can be a little
`different. That does not change during operation.
`Variation in wire thickness. Again, that's represented by
`
`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
`
`or be done during operation. Here, however, they didn't.
`They said a variation parameter. They did not require that
`it be done during operation.
`And then taking a step back, looking at the
`specifications, there, too, the inventors knew that
`variation parameters could possibly vary in operation and
`the specification discusses options where that does happen,
`but in the claims themselves, the inventors didn't use that
`language.
`
`Turning to the last term in the '027 patent,
`determining an operational temperature. I won't go over
`determining sensing again. The other element of this claim
`is determining an operational temperature.
`VLSI proposes that no construction is required
`or that an operational temperature is a temperature during
`operation.
`
`Intel, however, proposes that an, which is one
`or more, to be construed as the, which is singular, and
`temperature of the