` UNITED STATES PATENT AND TRADEMARK OFFICE
` BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`APPLE, INC.,
` Petitioner,
`vs. No. IPR2016-01114
`PARTHENON UNIFIED MEMORY
`ARCHITECTURE LLC,
` Patent Owner.
`___________________________ )
`
` DEPOSITION OF
` ROBERT COLWELL, Ph.D.
` __________________________________
` Wednesday, August 2, 2017
`
`Reported by:
`COREY W. ANDERSON
`CSR No. 4096
`
`212-279-9424
`
`Veritext Legal Solutions
`www.veritext.com
`
`212-490-3430
`
`1
`2
`
`34
`
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`PUMA Exhibit 2012
`APPLE v. PUMA, IPR2016-01114
`Page 1 of 55
`
`
`
`Page 2
` UNITED STATES PATENT AND TRADEMARK OFFICE
` BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`APPLE, INC.,
` Petitioner,
`vs. No. IPR2016-01114
`PARTHENON UNIFIED MEMORY
`ARCHITECTURE LLC,
` Patent Owner.
`___________________________ )
`
` Deposition of ROBERT COLWELL, Ph.D. taken on
`behalf of Patent Owner, at HAYNES BOONE, 525 University
`Avenue, 4th Floor, Palo Alto, California, beginning at
`9:02 A.M. and ending at 9:48 A.M. on Wednesday, August
`2, 2017, before COREY ANDERSON, Certified Shorthand
`Reporter No. 4096.
`
`1
`2
`
`34
`
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`212-279-9424
`
`Veritext Legal Solutions
`www.veritext.com
`
`212-490-3430
`
`PUMA Exhibit 2012
`APPLE v. PUMA, IPR2016-01114
`Page 2 of 55
`
`
`
`Page 3
`
`APPEARANCES:
`
`FOR THE PETITIONER:
`
`FEINBERG, DAY, ALBERTI & THOMPSON LLP
`DAVID ALBERTI, ESQ.
`1600 El Camino Real
`Suite 280
`Menlo Park, CA 94025
`(650) 618-4361
`dalberti@feinday.com
`
`SIDLEY AUSTIN LLP
`CURT HOLBREICH, ESQ.
`555 California Street
`San Francisco, CA 94104
`(415) 772-7446
`cholbreich@sidley.com
`
`1
`
`23
`
`45
`
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`212-279-9424
`
`Veritext Legal Solutions
`www.veritext.com
`
`212-490-3430
`
`PUMA Exhibit 2012
`APPLE v. PUMA, IPR2016-01114
`Page 3 of 55
`
`
`
`Page 4
`
`HAYNES and BOONE, LLP
`MICHAEL S. PARSONS, ESQ.
`DAVID W. O'BRIEN, ESQ.
`RAGHAV BAJAJ, ESQ.
`2505 North Plano Road
`Suite 4000
`Richardson, TX 75082-4101
`(972) 739-8611
`michael.parsons@haynesboone.com
`
`FOR THE PATENT OWNER:
`
`AHMAD ZAVITSANOS ANAIPAKOS ALAVI MENSING
`MASOOD ANJOM, ESQ.
`1221 McKinney
`Suite 3460
`Houston, TX 66010
`(713) 655-1101
`MANJOM@AZALAW.COM
`
` --oOo--
`
`ALSO PRESENT: JESSICA HANNAH
`
`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
`
`212-279-9424
`
`Veritext Legal Solutions
`www.veritext.com
`
`212-490-3430
`
`PUMA Exhibit 2012
`APPLE v. PUMA, IPR2016-01114
`Page 4 of 55
`
`
`
` I N D E X
` INDEX OF EXAMINATIONS
`EXAMINATION Page
`
`Page 5
`
`EXAMINATION BY MR. ANJOM 6
`
` EXHIBITS
`(None marked)
`
`1
`2
`3
`
`456
`
`789
`
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`212-279-9424
`
`Veritext Legal Solutions
`www.veritext.com
`
`212-490-3430
`
`PUMA Exhibit 2012
`APPLE v. PUMA, IPR2016-01114
`Page 5 of 55
`
`
`
` Palo Alto, California
` Wednesday, August 2, 2017
` 9:02 A.M.
`
`Page 6
`
` ROBERT COLWELL, Ph.D.
` having been sworn by the Reporter,
` testified as follows:
` THE WITNESS: Yes, I do.
` EXAMINATION
` BY MR. ANJOM:
` Q. Good morning, Dr. Colwell.
` A. Good morning.
` Q. Could you please state your full name and
`address for the record?
` A. Full name is Robert P. Colwell, my address is
`3594 Northwest Bronson Crest Loop, Portland, Oregon.
` Q. And you understand that you are testifying in
`your capacity as an expert today?
` A. Yes.
` Q. You understand that you are testifying under
`oath?
` A. Yes.
` Q. You understand that you are here to provide
`complete and accurate testimony today?
` A. Yes.
`
`1
`2
`3
`
`45
`
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`212-279-9424
`
`Veritext Legal Solutions
`www.veritext.com
`
`212-490-3430
`
`PUMA Exhibit 2012
`APPLE v. PUMA, IPR2016-01114
`Page 6 of 55
`
`
`
`Page 7
` Q. And you understand that if you don't fully
`understand a question I'm asking you can ask me to
`rephrase the question for you?
` A. I do.
` Q. And if I ask you a question and you respond,
`would it be fair for me to assume you understood the
`question?
` A. In general, yes.
` Q. What do you mean? Are there instances where
`that won't be true?
` A. I just know in the past there have been
`instances where in retrospect the question didn't mean
`what I thought. I don't just mean this case, I just
`mean in general.
` Q. Okay. And you are acting as an expert for
`petitioners in multiple IPRs.
` Correct?
` A. Multiple IPRs meaning multiple patents?
` Q. Multiple IPR proceedings?
` A. I think so, yes.
` Q. And one of those is IPR2016-01114.
` Is that correct?
` A. I don't know that number.
` MR. ALBERTI: I think it's this (indicating).
`So there are two -- looks like we have two different
`
`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
`
`212-279-9424
`
`Veritext Legal Solutions
`www.veritext.com
`
`212-490-3430
`
`PUMA Exhibit 2012
`APPLE v. PUMA, IPR2016-01114
`Page 7 of 55
`
`
`
`Page 8
`sets of declarations, this was for the first one, this
`for the second (indicating).
` THE WITNESS: Okay. And the number?
` BY MR. ANJOM:
` Q. Just to make this easier, there are multiple
`patents that you addressed.
` Right?
` A. Right.
` Q. One of them was the '753 patent?
` A. Correct.
` Q. And IPR2016-01114 is the IPR first to identify
`the patent.
` Correct?
` A. Yes.
` Q. And you are prepared to discuss the opinions
`you expressed in the declarations that you submitted in
`that IPR.
` Correct?
` A. Correct.
` Q. What did you do to prepare for your deposition
`today?
` A. I reread the -- my original declaration, and I
`reread the reply declaration from June, and I met with
`counsel yesterday.
` Q. And how long was that meeting?
`
`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
`
`212-279-9424
`
`Veritext Legal Solutions
`www.veritext.com
`
`212-490-3430
`
`PUMA Exhibit 2012
`APPLE v. PUMA, IPR2016-01114
`Page 8 of 55
`
`
`
`Page 9
`
` A. About five hours.
` Q. Is there anything that you feel like you need
`to do to get ready for the deposition today?
` A. I'm not sure I understood that question.
`Like...
` Q. Do you feel like you have done everything you
`had to do to get ready for your deposition today or was
`there anything else?
` A. Oh. You know, I think I -- I think the
`preparation was reasonable.
` Q. Okay. Let's start with the '753 patent and
`then we can move on to the other IPR. Now, you
`submitted two declarations in IPR2016-01114.
` Correct?
` A. Yes.
` Q. I'm just going to refer to this as the '753
`IPR.
` Would that be okay?
` A. Sure.
` Q. Just so we are on the same page, I know you
`have some of them here, but I want to make sure we have
`the same thing. So I'm going to hand you a copy of your
`original declaration which is Exhibit 1003 in the '753
`IPR.
` Do you want 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
`
`212-279-9424
`
`Veritext Legal Solutions
`www.veritext.com
`
`212-490-3430
`
`PUMA Exhibit 2012
`APPLE v. PUMA, IPR2016-01114
`Page 9 of 55
`
`
`
`Page 10
`
` MR. ALBERTI: I do not have --
` MR. ANJOM: Give you one.
` MR. ALBERTI: Thank you.
` THE WITNESS: Okay.
` BY MR. ANJOM:
` Q. And that's a correct copy of your original
`declaration?
` A. Yeah, it looks like it.
` Q. Okay. I'm going to hand you a copy of your
`declaration that was submitted for the petitioner's
`reply, and that's Exhibit 1028. Is that a correct copy
`of your reply declaration?
` A. Yes.
` Q. With respect to your reply declaration that's
`Exhibit 1028, are the opinions contained in this
`declaration complete and accurate?
` A. I think so.
` Q. Who drafted this declaration?
` A. (No response).
` Q. By think declaration, I mean the reply
`declaration 1028?
` A. It was basically a joint effort.
` Q. Who -- I'm sorry, go ahead.
` A. I wrote a lot of it. Some of it I asked for
`help with. And we mutually edited it to my
`
`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
`
`212-279-9424
`
`Veritext Legal Solutions
`www.veritext.com
`
`212-490-3430
`
`PUMA Exhibit 2012
`APPLE v. PUMA, IPR2016-01114
`Page 10 of 55
`
`
`
`Page 11
`
`satisfaction.
` Q. And who helped you with that?
` A. Counsel.
` Q. And which sections of the declarations did you
`personally prepare and which sections did counsel help
`you with?
` A. Well, I personally worked on all of them. And
`I got help with the structure of it, I asked for what
`topics needed to be covered.
` Excuse me.
` Q. That's all?
` A. That's all I remember at the moment.
` Q. When did you start writing this declaration?
` A. When I was asked to. But I think that was
`around May of this year.
` Q. Okay. And I see that you signed the
`declaration on June 8th.
` Correct?
` A. Correct.
` Q. So between May when you started working on the
`declaration and June 8th when you signed it,
`approximately how much time did you spend preparing the
`declaration?
` A. I don't know.
` Q. Was it less than 50?
`
`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
`
`212-279-9424
`
`Veritext Legal Solutions
`www.veritext.com
`
`212-490-3430
`
`PUMA Exhibit 2012
`APPLE v. PUMA, IPR2016-01114
`Page 11 of 55
`
`
`
`Page 12
`
` A. I don't know.
` Q. More than 100?
` A. I don't know how many hours it was.
` Q. Was it more than 200?
` A. Seems unlikely, but I just -- I don't keep
`track of my hours that way. My wife does. So I just
`don't know.
` Q. Did you submit a bill for the work you did on
`this declaration?
` A. I submitted a monthly invoice for my hours.
` Q. And based on that, can you give me a
`guesstimate of how much time you may have spent on this
`declaration?
` A. I -- no, because I just don't remember the
`numbers.
` Q. Now, one of the references that you address in
`this declaration is the Bowes reference.
` Correct?
` A. Yes.
` Q. Are you familiar with the Bowes reference?
` A. Yes.
` Q. I'm going to hand you a copy of Bowes, if it's
`easier. That's Exhibit 1005. And Exhibit 1005 is a
`copy of the Bowes reference.
` Correct?
`
`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
`
`212-279-9424
`
`Veritext Legal Solutions
`www.veritext.com
`
`212-490-3430
`
`PUMA Exhibit 2012
`APPLE v. PUMA, IPR2016-01114
`Page 12 of 55
`
`
`
`Page 13
`
` A. Correct.
` Q. And if you can turn to figure 2 of Bowes now,
`this is one of the figure that you analyzed and opined
`on.
` Correct?
` (Pause)
` A. I'm looking, to answer your question I'm
`looking for where I specifically talked about figure 2,
`and I'm not seeing that. But I know that it's -- it's
`the figure that I had in mind when I did the analysis.
` Q. Do you remember analyzing that figure apart
`from the Exhibit 1028?
` A. A apart from, sorry?
` Q. Apart from what's the text in Exhibit 1028
`that you are flipping through, do you remember analyzing
`figure 2 of Bowes?
` A. Figure 2 of Bowes, yes, I do.
` Q. Okay. Now, you identified DSP 20 of Bowes as
`corresponding to the claimed decoder of the '753 patent.
` Is that correct?
` (Pause)
` A. Yes.
` Q. And you identified the CPU 10 of Bowes as
`corresponding to the central processing unit that's in
`the claims of the '753 patent.
`
`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
`
`212-279-9424
`
`Veritext Legal Solutions
`www.veritext.com
`
`212-490-3430
`
`PUMA Exhibit 2012
`APPLE v. PUMA, IPR2016-01114
`Page 13 of 55
`
`
`
`Page 14
`
` Is that correct?
` A. Yes.
` Q. And there is a memory bus 110 in Bowes.
` Do you see that?
` A. I do.
` Q. And that memory bus allows a number of bus
`masters including the CPU 10 and the DSP 20 to access
`the memory 14.
` Is that correct?
` A. Correct.
` Q. And which component did you identify as the
`arbiter in Bowes, do you remember that?
` A. I believe that's box 200 MCA.
` Q. And the function of the MCA 200 is to
`arbitrate the memory bus 110 between the various bus
`masters.
` Correct?
` A. Yes.
` Q. Now, Bowes discloses four potential bus
`masters.
` Correct?
` A. Yeah, I think that's right.
` Q. What are those four bus masters?
` A. CPU 10, DSP 20, the new bus controller 40, and
`the I/O interface 30.
`
`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
`
`212-279-9424
`
`Veritext Legal Solutions
`www.veritext.com
`
`212-490-3430
`
`PUMA Exhibit 2012
`APPLE v. PUMA, IPR2016-01114
`Page 14 of 55
`
`
`
`Page 15
` Q. Now, each of those four bus masters may at
`some point require access to the memory bus.
` Correct?
` A. Correct.
` Q. Now, let's take the DSP 20 as an example.
`When the DSP 20 wants to access the memory bus, it sends
`a bus request signal to the arbiter 200.
` Correct?
` A. Yeah. Right.
` Q. When the memory bus is available for
`assignment, the arbiter 200 would issue a bus grant
`signal to the DSP 20 in accordance with the Bowes
`priority scheme.
` Is that correct?
` A. In accordance with this priority scheme,
`that's right, yes.
` Q. Now, after the bus grants signal is issued,
`the DSP 20 can use the memory bus 110 to communicate
`with the memory 14.
` Correct?
` A. Right.
` Q. Now, the DSP 20 cannot read from the memory 14
`or write to the memory 14 unless and until the MCA 200
`grants the DSP 20 access to the memory bus.
` Is that correct?
`
`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
`
`212-279-9424
`
`Veritext Legal Solutions
`www.veritext.com
`
`212-490-3430
`
`PUMA Exhibit 2012
`APPLE v. PUMA, IPR2016-01114
`Page 15 of 55
`
`
`
`Page 16
`
` A. Correct.
` Q. The DSP 20 cannot communicate with the memory
`14 via the memory bus at all until after the MCA 200
`issues a bus grant signal.
` Is that correct?
` A. Until the MCA 200 has issued the bus grant
`signal to the DSP.
` Q. Correct?
` A. Not just any grant signal.
` Q. Right. So with that correction, is that
`statement correct?
` A. Yes.
` Q. Let's go to figure 4 of Bowes. Now, figure 4
`shows the MCA 200 from figure 2.
` Is that correct?
` A. Yes.
` Q. And that's what you have identified as the
`arbiter?
` A. Correct.
` Q. Now, you see that the arrow that says DSP REQ
`on the left that goes into the box 240?
` A. I do.
` Q. What is that?
` A. That's the request, bus request signal coming
`from the DSP.
`
`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
`
`212-279-9424
`
`Veritext Legal Solutions
`www.veritext.com
`
`212-490-3430
`
`PUMA Exhibit 2012
`APPLE v. PUMA, IPR2016-01114
`Page 16 of 55
`
`
`
`Page 17
` Q. Okay. So that's a signal path for the bus
`requesting from the DSP 20 to the MCA 200.
` Right?
` A. Right.
` Q. And there is an arrow that has, it's in bold
`on the top right-hand side and it has four different
`items listed on it, one of which is DSP BGM.
` Do you see that?
` A. No, I don't, actually.
` Q. So on the right-hand side of box 240, there
`are three different arrows. The first one goes out of
`the box 240. Do you see that, the first arrow that goes
`out of the box 240?
` A. The one that's got four names on it?
` Q. Yes.
` A. That -- and it's bold?
` Q. Yes.
` A. Yeah, I see that one.
` Q. Now, one of the names on it is DSP BGN.
` Do you see that?
` A. I do.
` Q. What is that?
` A. That's the bus grant signal going back to the
`DSP.
` Q. So that's the signal path for the bus grant
`
`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
`
`212-279-9424
`
`Veritext Legal Solutions
`www.veritext.com
`
`212-490-3430
`
`PUMA Exhibit 2012
`APPLE v. PUMA, IPR2016-01114
`Page 17 of 55
`
`
`
`Page 18
`
`signal from the MCA 200 to the DSP 20.
` Correct?
` A. Yes.
` Q. And that's what tells the DSP 20 that it is
`permitted to access the memory bus.
` Correct?
` A. Correct.
` Q. Now, going back to figure 2, you see the DSP
`20?
` A. Yes.
` Q. What are the components in this figure that
`are coupled to the DSP 20?
` MR. ALBERTI: Objection to form.
` THE WITNESS: I'm not sure what you mean by
`coupled to.
` BY MR. ANJOM:
` Q. Have you used the term "coupled" before?
` A. Yes.
` Q. What does it mean to you?
` A. Well, it depends on the context. It can mean
`different things.
` Q. What are some of the different meanings?
` A. Well, it could mean something as specific as
`directly connected to via a direct wire, or it could be
`something more general, depending on the context. You
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`212-279-9424
`
`Veritext Legal Solutions
`www.veritext.com
`
`212-490-3430
`
`PUMA Exhibit 2012
`APPLE v. PUMA, IPR2016-01114
`Page 18 of 55
`
`
`
`Page 19
`know, able to communicate with via an intermediary, for
`example.
` Q. So let's look at, just a minute, let's look at
`Exhibit 1028. That's your reply declaration.
` A. Okay.
` Q. Page 8, paragraph 14. Can you just read that
`paragraph 14? You can read it to yourself, you don't
`need to read it out loud.
` A. Okay.
` (Pause)
` A. Okay. I have read it.
` Q. Do you see that you have used the term
`"coupled to" in that paragraph?
` A. I do.
` Q. What does "coupled to" mean in that context
`that you have used it?
` MR. ALBERTI: Objection, form.
` (Pause)
` THE WITNESS: So I think in this context the
`intention was to say it's connected to. It's
`communicating with.
` BY MR. ANJOM:
` Q. So coupled to means it's communicating with
`and it's connected to?
` A. Well, kind of means both in this instance.
`
`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
`
`212-279-9424
`
`Veritext Legal Solutions
`www.veritext.com
`
`212-490-3430
`
`PUMA Exhibit 2012
`APPLE v. PUMA, IPR2016-01114
`Page 19 of 55
`
`
`
`Page 20
`
`It's communicating because -- or it's able to
`communicate because they are connected together.
` Q. Okay. So using that definition that you just
`gave me for couples to, going back to figure 2, is the
`MCA 200 coupled to the DSP 20?
` A. Yes.
` Q. Is CPU 10 coupled to the DSP 20?
` A. Well, yes and no. The fact that they're
`sharing a bus means that they are connected together,
`but they are not -- there is no indication in Bowes that
`the -- well, Bowes simply doesn't address the question
`of whether the CPU can directly write to the DSP using
`that memory bus. It just doesn't address the issue.
`And it's not -- it's plausible either way, whether it
`does or not. That's a system implementation choice.
` Q. So as a person of ordinary skill in the art,
`would it be accurate to say that it's possible that the
`DSP CPU in figure 12 Bowes are coupled or they may not
`be depending upon design choice.
` Would that be fair?
` A. Hard to say. I mean, in the sense that I
`meant coupled to as we just discussed, it is plausible
`either way that the CPU might have a write path, for
`example, into the control registers of the DSP. You
`can't tell. Bowes doesn't address it, so there is no
`
`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
`
`212-279-9424
`
`Veritext Legal Solutions
`www.veritext.com
`
`212-490-3430
`
`PUMA Exhibit 2012
`APPLE v. PUMA, IPR2016-01114
`Page 20 of 55
`
`
`
`Page 21
`
`way to know.
` Q. Right. So as a person of ordinary skill in
`the art, looking at figure 2, it's possible that they
`are coupled to each other and it's possible that they
`are not.
` Is that fair?
` A. I can't really answer that.
` Q. Why not?
` A. As stated. Because I don't have enough
`information. I mean, Bowes calls that thing a memory
`bus and he spends most of his patent talking about how
`he is going to arbitrate it so as to share the bandwidth
`to memory, but that's all he focuses on. So how he does
`an awful lot of the rest of the housekeeping that a
`system has to be able to do, he just leaves up in the
`air, doesn't address it at this patent.
` So I just -- I don't know.
` Q. So apart from what Bowes specifically
`addresses in the patent, you did your analysis -- strike
`that.
` You did your analysis from the perspective of
`a person of ordinary skill in the art.
` Is that correct?
` A. True.
` Q. So you were not bound by the specific texts of
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`212-279-9424
`
`Veritext Legal Solutions
`www.veritext.com
`
`212-490-3430
`
`PUMA Exhibit 2012
`APPLE v. PUMA, IPR2016-01114
`Page 21 of 55
`
`
`
`Page 22
`
`Bowes.
` Correct?
` MR. ALBERTI: Objection to form.
` THE WITNESS: I guess I didn't think of it
`that way. I don't know what bound by means exactly
`here.
` BY MR. ANJOM:
` Q. Was your analysis limited to the specific text
`that is in Bowes, or do you your analysis from the
`perspective of a person of ordinary skill in the art?
` A. Well, both. I mean, I tried to approach it as
`of what -- what -- how a person of ordinary skill would
`approach it, but I don't want to invent things that
`aren't present. I mean, that's -- if he doesn't address
`something, it's -- it's -- I think a person of ordinary
`skill would do the same what I'm doing, which is I
`simply just don't know the answer.
` Q. So in your view if he doesn't specifically
`address something, you would have to invent things to
`make sense of it?
` MR. ALBERTI: Objection, form.
` THE WITNESS: Not always. But in this case I
`think you would.
` BY MR. ANJOM:
` Q. Is the DSP 20 of Bowes coupled to the ROM 12
`
`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
`
`212-279-9424
`
`Veritext Legal Solutions
`www.veritext.com
`
`212-490-3430
`
`PUMA Exhibit 2012
`APPLE v. PUMA, IPR2016-01114
`Page 22 of 55
`
`
`
`Page 23
`
`in figure 2?
` A. That's even less -- I mean, I don't know why
`the DSP 20 would have to communicate with the ROM. I
`don't remember seeing that addressed in Bowes.
` Q. Is the DSP of Bowes coupled to the memory 14?
` A. The DSP 20 definitely has to communicate with
`memory. So in that sense I think you could say they are
`coupled.
` Q. Is the DSP 20 coupled to the new bus
`controller in figure 2?
` A. I don't know.
` Q. What about the DMA controller? Is the DSP 20
`coupled to the DMA controller?
` A. I don't know.
` Q. What additional information would you need to
`be able to determine if they are coupled or not?
` A. Well, for instance, if -- if -- the easiest
`way would be if this was a real system that was deployed
`in the field, there would be documentation in addition
`to this patent disclosure, and the kind of thing you
`might see in that documentation would explain how a
`person would program the DSP, for instance.
` And in that -- in that discussion, they would
`tell you where the program for the DSP comes from, how
`it gets loaded, and what the constraints are for 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
`
`212-279-9424
`
`Veritext Legal Solutions
`www.veritext.com
`
`212-490-3430
`
`PUMA Exhibit 2012
`APPLE v. PUMA, IPR2016-01114
`Page 23 of 55
`
`
`
`Page 24
`And maybe it's -- it's -- the DSP has some means of
`talking to one of these other agents that are on that
`bus.
` But without that kind of documentation, you
`can't -- I mean, just going from the patent there is
`just not enough information here.
` Q. So whether the DSP 20 talks to the other
`components that I asked you about is an implementation
`choice or a design choice.
` Would that be fair?
` A. At a system level, yeah. It's like the system
`architecture choice that whoever made this entire box,
`that's the kind of choice they would make.
` Q. So the MCA 200 of Bowes merely controls access
`to the memory bus, but does not itself access or utilize
`the memory bus.
` Is that accurate?
` A. Yes. That's my understanding.
` Q. Did you review Dr. Thornton's deposition
`transcript in preparing your reply declaration?
` A. I did.
` Q. Let's turn to page 11 of Exhibit 1028, which
`is your reply declaration. Do you see that you have
`cited to a portion of deposition transcript from
`Dr. Thornton?
`
`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
`
`212-279-9424
`
`Veritext Legal Solutions
`www.veritext.com
`
`212-490-3430
`
`PUMA Exhibit 2012
`APPLE v. PUMA, IPR2016-01114
`Page 24 of 55
`
`
`
`Page 25
`
` A. In the middle of page 11?
` Q. Yes.
` A. Yes, I do.
` Q. And you see that there is a question posed to
`Dr. Thornton on lines 14 through 17, and the question
`asks "A person of ordinary skill in the art would want
`the arbiter to control access to memory, but not
`necessarily take up memory bus bandwidth itself."
` Right? That's the question. Do you see that?
` A. I do.
` Q. Now, what does the term "arbiter" refer to in
`that question, is that the arbiter from a specific
`system or is that arbiter in the abstract, do you
`remember?
` A. I don't at the moment remember without looking
`at the depo again. So the way -- when I wrote the
`reply, when reading the paragraph above the quoted text,
`it's clear what I had in mind was this, this specific
`system, as you know, because I am referring to MCA 200.
`And that's a Bowes.
` Q. So it's your understanding the term the
`arbiter in that cited portion of the transcript is
`referring to the MCA 200 of Bowes?
` A. I think that's what I had in mind when I wrote
`this. But you know, by extension it's like any time you
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`212-279-9424
`
`Veritext Legal Solutions
`www.veritext.com
`
`212-490-3430
`
`PUMA Exhibit 2012
`APPLE v. PUMA, IPR2016-01114
`Page 25 of 55
`
`
`
`Page 26
`have an arbiter controlling the only access to memory,
`the same conditions apply.
` Q. So are all arbiter -- I'm not sure I
`understand. Are you saying all arbiters in all systems
`are the same?
` A. No, they are certainly not. But any system
`that has a shared access path to memory and the need for
`multiple agents on that access path to memory, to take
`turns accessing memory, under those conditions
`controlling access to the bus is controlling access to
`the memory.
` Q. But so in the context of the paragraph
`preceding this cite of testimony, is your recollection
`that the term "the arbiter" here is referring to the MCA
`200 of Bowes?
` A. Well, it's the example that I had in mind.
`But as I said, I think it's -- it's generally -- this is
`just an instance of a general class of machines. And I
`believe Dr. Thornton understood that from reading his
`depo.
` Q. Now, let's turn to page 10 of Exhibit 1028.
`You see the heading which says "A POSITA Would Have
`Considered Colocating the Arbiter with the DSP."
` Do you see that heading?
` A. I do.
`
`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
`
`212-279-9424
`
`Veritext Legal Solutions
`www.veritext.com
`
`212-490-3430
`
`PUMA Exhibit 2012
`APPLE v. PUMA, IPR2016-01114
`Page 26 of 55
`
`
`
`Page 27
` Q. What does the term "Colocating" mean in that
`heading?
` A. The arbiter is a pile of gates at some level
`of abstraction, and those gates go together in order to
`accomplish the function of arbitration. In a real
`system that pile of gates has to be physically realized
`somewhere in order to work, and so by co-locating I
`meant as a system implementer you have a choice where to
`put those gates, and one of the choices -- one of the
`logical places would be with the DSP.
` Q. Would be within the DSP, is that what you
`said?
` A. Well, I don't know within. But suppose you
`are -- suppose your DSP was a single integrated circuit,
`one of the -- and your system is comprised of other such
`integrated circuits doing other functions; CPU, for
`example. By co-locating under those conditions I mean
`putting the pile of gates that form the arbiter onto the
`same piece of silicon as the rest of the DSP.
` Q. Okay. So would it be fair that when you say
`co-locating here you are saying that the pile of gates
`that make up the arbiter like you just mentioned and
`whatever circuits are associated with the DSP would all
`be on the same integrated circuit or the same chip?
` A. Yeah, for that set of assumptions, that's what
`
`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
`
`212-279-9424
`
`Veritext Legal Solutions
`www.veritext.com
`
`212-490-3430
`
`PUMA Exhibit 2012
`APPLE v. PUMA, IPR2016-01114
`Page 27 of 55
`
`
`
`Page 28
`
`I meant.
` Q. What else could be located on the same
`chipset?
` A. I'm not sure what you mean, what else.
` Q. So you have an integrated circuit where you
`said you can have the DSP and the arbiter located there.
`You could add other components there.
` Right?
` A. I'm still not sure. What do you mean by add
`other components?
` Q. Could you put other -- could you put any other
`components on the same integrated circuit or the same
`chip?
` A. You have to tell me what a component is to
`answer that.
` Q. Have you ever used the term "component"?
` A. Uh-huh.
` Q. What does the term "component" mean to you?
` A. Well, typically one might refer to a single
`chip as a component. So when you say you can add a chip
`to a chip, I'm not sure.
` Q. Well, so that obviously doesn't make sense.
`Right? So what else does the term "component" mean
`other than actual chip?
` A. Well, sometimes people speak of a complex --
`
`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
`
`212-279-9424
`
`Veritext Legal Solutions
`www.veritext.com
`
`212-490-3430
`
`PUMA Exhibit 2012
`APPLE v. PUMA, IPR2016-01114
`Page 28 of 55
`
`
`
`Page 29
`they'll decompose a complex design or a complex thing.
` Q. Uh-huh.
` A. Into pieces, and sometimes they'll call those
`"components."
` Q. So let's do that, let's call those pieces
`"components." Are there any other pieces or components
`that you could add in that sense to an integrated
`circuit?
` A. That question is ill-formed, I'm sorry.
` Q. That's fine.
` A. It's kind of the issue is you can -- subject
`to physical constraints and actually economic
`constraints, when you are designing integrated circuits,
`you -- you have a thing that you want to put onto the
`final product, and I'm going to -- I'll call that a chip
`for the moment because we are separating component to
`mean something else, and you have -- and you have to
`decide what goes into that chip and what will not go
`into it.
` Under no conditions would a reasonable person
`just randomly assign functions into that chip because it
`wouldn't work, either electrically or functionally and
`certainly not economically. There are general design
`guidelines that a designer follows to make the decisions
`as to what goes in and what does not go in to any given
`
`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
`
`212-279-9424
`
`Veritext Legal Solutions
`www.veritext.com
`
`212-490-3430
`
`PUMA Exhibit 2012
`APPLE v. PUMA, IPR2016-01114
`Page 29 of 55
`
`
`
`Page 30
`
`chip.
` Q. So you just said there are reasons they may
`not work and you said there might be electronic reasons,
`functional reasons, or economical reasons.
` Correct?
` A. Correct.
` Q. So what are some reasons why it may not work
`electronically?
` A. Let's see. For example, if you took -- so
`let's suppose it's a CPU, let's start with a CPU and
`there are certain functions inside that CPU that have to
`be there for it to be able to execute code and to do its
`job. It would be illogical and extremely weird to think
`about taking some internal function of that CPU and
`removing it and putting it somewhere else in the system.
` And even though I could probably arrange for
`the communication to be such that the function is still
`fulfilled, the fact that it's geographically far away
`might mean that the round trip delay of,