`
`Page 1
` 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 P. COLWELL, Ph.D.
` Menlo Park, CA
` Monday, February 27, 2017
` 9:27 a.m.
`
`Reported by: SUSAN F. MAGEE, RPR, CCRR, CLR
`CSR No. 11661
`
`1
`2
`
`345
`
`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 2010
`APPLE v. PUMA, IPR2016-01114
`Page 1 of 76
`
`
`
`ROBERT P. COLWELL
`
`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 P. COLWELL, Ph.D. taken on
` behalf of Patent Owner at FEINBERG, DAY,
` ALBERTI & THOMPSON LLP, 1600 El Camino Real,
` Suite 280, Menlo Park, CA 94025, beginning at
` 9:27 a.m. and ending at 11:18 a.m. on Monday,
` February 27, 2017, before Susan F. Magee, RPR,
` CCRR, CLR, Certified Shorthand Reporter
` No. 11661.
`
`1
`2
`
`345
`
`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 2010
`APPLE v. PUMA, IPR2016-01114
`Page 2 of 76
`
`
`
`ROBERT P. COLWELL
`
`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
`
` HAYNES and BOONE, LLP
` MICHAEL S. PARSONS, ESQ.
` ADAM C. FOWLES, ESQ.
` CHRISTIAN B.E. HINES, ESQ.
` 2505 North Plano Road
` Suite 4000
` Richardson, TX 75082-4101
` (972) 739-8611
` michael.parsons@haynesboone.com
` adam.fowles@haynesboone.com
` christian.hines@haynes.boone.com
`
`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 2010
`APPLE v. PUMA, IPR2016-01114
`Page 3 of 76
`
`
`
`ROBERT P. COLWELL
`
`Page 4
`
`APPEARANCES (continued):
` For the Patent Owner:
` AHMAD ZAVITSANOS ANAIPAKOS ALAVI MENSING
` JUSTIN CHEN, ESQ.
` 1221 McKinney
` Suite 2500
` Houston, TX 77010
` (713) 655-1101
` jchen@azalaw.com
`
` For HTC Corp. and HTC America, Inc.:
` SIDLEY AUSTIN LLP
` CURT HOLBREICH, ESQ.
` 555 California Street
` San Francisco, CA 94104
` (415) 772-7446
` cholbreich@sidley.com
` --o0o--
`
`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 2010
`APPLE v. PUMA, IPR2016-01114
`Page 4 of 76
`
`
`
`ROBERT P. COLWELL
`
`Page 5
`
` I N D E X
`DEPOSITION OF ROBERT P. COLWELL, Ph.D.
`
`EXAMINATION BY PAGE
`BY MR. CHEN 7
` --o0o--
`
`1
`2
`
`34
`
`5
`6
`
`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 2010
`APPLE v. PUMA, IPR2016-01114
`Page 5 of 76
`
`
`
`ROBERT P. COLWELL
`
`Page 6
`
` E X H I B I T S
`NUMBER DESCRIPTION PAGE
`Exhibit 1 Declaration of Robert P. Colwell 21
` Ph.D. Under 37 C.F.R. Section 1.68
` (123 pages)
`Exhibit 2 U.S. Patent 7,777,753 (21 pages) 23
`Exhibit 3 AT&T DSP3210 Digital Signal 28
` Processor, The Multimedia Solution
` (40 pages)
`Exhibit 4 US Patent No. 5,584,038 (18 pages) 31
`Exhibit 5 US Patent No. 5,546,547 (25 pages) 37
`Exhibit 6 US Patent No. 5,787,264 (10 pages) 52
` --o0o--
`
`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 2010
`APPLE v. PUMA, IPR2016-01114
`Page 6 of 76
`
`
`
`ROBERT P. COLWELL
`
`Page 7
` Menlo Park, CA, Monday February 27, 2017
` 9:27 a.m.
`
` ROBERT P. COLWELL, Ph.D.,
`having been administered an oath, was examined and
`testified as follows:
`
` MR. CHEN: Should we start with attorney
`appearances?
` This is Justin Chen with AZA representing the
`patent owner Parthenon Unified Memory Architecture.
` MR. ALBERTI: David Alberti with Feinberg Day
`Alberti & Thompson representing Apple.
` MR. FOWLES: Adam Fowles from Haynes and Boone
`representing Apple.
` MR. PARSONS: Michael Parsons from Haynes and
`Boone representing Apple.
` MR. CHEN: All right. And so at this point
`I'd like to open the record for the proceeding labeled
`IPR2016-01114.
`
` EXAMINATION BY MR. CHEN
`
` Q. And could you please state your name and
`address for the record.
`
`1
`2
`
`34
`
`5
`6
`
`78
`
`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 2010
`APPLE v. PUMA, IPR2016-01114
`Page 7 of 76
`
`
`
`ROBERT P. COLWELL
`
`Page 8
` A. My name is Robert P. Colwell. My address is --
`street address or just city?
` Q. Street address and city, please.
` A. 3594 Northwest Bronson Crest Loop, Portland,
`Oregon.
` Q. And have you ever had your deposition taken
`before?
` A. I have.
` Q. When?
` A. About two weeks ago.
` Q. And what was that in connection with?
` A. It was a different case.
` What would -- is that all you want to know
`about it?
` Q. Yeah, that's fine.
` And I guess, how many times have you been
`deposed before?
` A. Twenty-five maybe.
` Q. Okay. So I'm sure you're familiar with all the
`ground rules of a deposition, but I'll just briefly go
`over them again.
` Do you understand that you've taken an oath to
`tell the truth today?
` A. I do.
` Q. And the court reporter will be transcribing
`
`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 2010
`APPLE v. PUMA, IPR2016-01114
`Page 8 of 76
`
`
`
`ROBERT P. COLWELL
`
`Page 9
`everything we say, so it's important that we try not to
`talk over each other.
` So I'll do my best to let me -- let you finish
`your answers before I start my next question, and will
`you please let me finish my questions before you begin
`to answer?
` A. I will try.
` Q. You're doing a great job so far of giving me
`verbal responses, so can you please continue to do that?
` A. As opposed to --
` Q. As opposed to nodding your head or shaking your
`head?
` A. Oh, yes.
` Q. And if at any point you don't understand a
`question that I'm asking, will you please let me know?
` A. I will do.
` Q. And if at a later point you remember something
`that you would like to add to a previous answer, will
`you please let me know?
` A. Okay.
` Q. Is there anything that could prevent you from
`offering truthful and accurate testimony today?
` A. Nope.
` Q. Okay. Are you currently employed?
` A. No. Well, I'm a consultant in my own
`
`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 2010
`APPLE v. PUMA, IPR2016-01114
`Page 9 of 76
`
`
`
`ROBERT P. COLWELL
`
`Page 10
`
`consulting firm.
` Q. So you own your own consulting firm?
` A. My wife and I do.
` Q. And do you have a title in this consulting firm
`that you own?
` A. Like personal title?
` Q. Yeah.
` A. Consultant.
` Q. Okay.
` A. Well, also president, and my wife's vice
`president.
` Q. How long have you owned this consulting company
`for?
` A. I think we started it in 2002.
` Q. And so since 2002, is this the only thing that
`you've been, I guess, employed with?
` A. No.
` Q. Where else have you worked since 2002?
` A. I was employed by DARPA, D-A-R-P-A, Defense
`Advance Research Projects Agency, from 2011 to 2014.
` Q. And what did you do before DARPA?
` A. What -- you mean immediately before?
` Q. Yeah. Immediately before.
` A. Well, as I mentioned, I was a consultant for my
`company from 2002 until when I started at DARPA in 2011.
`
`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 2010
`APPLE v. PUMA, IPR2016-01114
`Page 10 of 76
`
`
`
`ROBERT P. COLWELL
`
`Page 11
` Q. So from 2002 to 2011 as in terms of the work
`you were doing, it was just through that consulting firm
`that you own?
` A. Yes.
` Q. What did you work on while you were at DARPA?
` A. Lots and lots of things, but I was hired as a
`deputy director, and I was -- I was that for a year, and
`then I became a director of an agent -- I'm sorry. I
`should explain. DARPA has five or six what's called
`technical offices, and in each office there are order of
`20 program managers, and each office is led by a
`director and a deputy.
` So one of those six technical offices was MTO,
`Microsystems Technology Office, and I was hired as the
`deputy for that, and I was in that role for a year. And
`then I took over as director for the next two years
`until I left.
` Q. So you mentioned you were part of the -- is it
`micro technology?
` A. Microsystems Technology.
` Q. Microsystems Technology Office.
` And what did that office focus on?
` A. We were expected to identify opportunities for
`the Department of Defense in various technical areas
`that were -- I don't know how to put a border on them,
`
`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 2010
`APPLE v. PUMA, IPR2016-01114
`Page 11 of 76
`
`
`
`ROBERT P. COLWELL
`
`Page 12
`but I can give you examples. We were worried about
`high-powered radars, very high-powered lasers, more
`efficient computing systems. We even made
`blast-detection gauges to put on soldiers that could be
`exposed to road-side bombs. We started some genetic
`engineering efforts. There's -- I'm forgetting some. I
`mean, there were lots. It's a pretty wide range.
` Q. So while you were at DARPA, did you work on
`anything relating to shared memory architectures?
` A. Not per se. We did have some computing systems
`work, but it wasn't -- that wasn't the focus of it. The
`focus was energy efficiency.
` Q. And I guess from 2002 through 2011 in your
`consulting work, did you work on anything relating to
`shared memory architectures?
` A. That's hard to answer. I mean, I worked on
`computing systems in general, and they all had memory
`systems. By "shared memory" I presume you mean shared
`in the sense of the cases at question here --
` Q. Yes.
` A. -- which means there's graphics and stuff
`coming out of the same memory as the -- I don't
`remember. I don't remember if we had systems that did
`that anymore or not.
` Q. And before you started your consulting firm in
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`212-279-9424
`
`Veritext Legal Solutions
`www.veritext.com
`
`212-490-3430
`
`PUMA Exhibit 2010
`APPLE v. PUMA, IPR2016-01114
`Page 12 of 76
`
`
`
`ROBERT P. COLWELL
`
`Page 13
`
`2002, were you employed?
` A. Yes.
` Q. And where were you employed?
` A. Before -- from 1990 to 2001 I was employed
`full-time by Intel.
` Q. And what did you work on when you were at
`Intel?
` A. For the -- from '90 until '92 I was a senior
`architect on a chip that we called P6, and the product
`name became the Pentium Pro as it productized in -- in
`1995.
` Then in the summer of '92 until '95 -- well,
`from '92 until I left the place, I was put in charge of
`all the architecture efforts for the x86 at the company,
`but I was still directly in charge of the architecture
`effort of the P6 and then of the Pentium 4. We called
`it Willamette. It became known as Pentium 4's product.
` Q. And in your work at Intel, did you work on
`anything that related to triggered memory architectures?
` A. Well, I was in charge of the CPU part, so
`the -- that's more of a systems-y question. People
`could take the chip that I was making or leading the
`charge on and build a shared memory system out of it if
`they wanted to. That was outside my purview.
` Q. So while you were working on the Pentium
`
`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 2010
`APPLE v. PUMA, IPR2016-01114
`Page 13 of 76
`
`
`
`ROBERT P. COLWELL
`
`Page 14
`processors at Intel, did you ever draft any diagrams or
`technical diagrams in your work?
` A. Undoubtedly. In fact, yes, the answer is yes
`for sure.
` Q. What kinds of diagrams would that entail?
` A. Lots. The first one was I wanted to do a
`better job than other -- than I thought other chip
`design efforts at Intel had done. So I was trying to
`convince a design team of engineers who are historically
`loathed to document what they're doing -- they'd much
`rather sling gates around than write down what it is
`they're trying to do. And so in order to get them to do
`that, I started doing the diagrams myself, and then it's
`a lot easier to get someone to say, "No, I hate that
`diagram. It should really look like this," than to give
`them a blank sheet of paper and say, "Please draw what
`you think your piece of the machine should look like."
` So I started drawing those just -- and some of
`them lived. Some of them are still the ones that Intel
`uses to describe what's inside of a P6, for example.
`Anyway, that was one reason. Internal documentation was
`one reason.
` I was asked to give a lot of public
`appearances. I gave like 20 university talks in 1996 or
`so on the P6, and all of that documentation was done 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
`
`212-279-9424
`
`Veritext Legal Solutions
`www.veritext.com
`
`212-490-3430
`
`PUMA Exhibit 2010
`APPLE v. PUMA, IPR2016-01114
`Page 14 of 76
`
`
`
`ROBERT P. COLWELL
`
`Page 15
`
`me.
` Q. So those diagrams were about, I guess, the
`physical layout of things on the chip, or did they
`represent something else?
` A. Yeah. It wasn't -- I don't -- the physical
`layout of things on a chip was done by another part of
`the team, and I didn't have a whole lot to do with that.
`My part was to explain logically why various functional
`units were in there, why they did what they did and how
`they interacted with other ones.
` Q. Okay. So I guess you kind of mention that you
`did something logically.
` What does that mean in contrast to something
`that's physical?
` A. Well, architects tend to live in a more
`abstract level of existence maybe. I have known people
`that call themselves architects that don't -- didn't go
`through electrical engineering and don't understand
`necessarily how circuits even work, but there is still a
`level -- and by the way, I don't condone that. I
`actually think you're a better engineer if you
`understand all the levels.
` But the point is you can -- it's at least --
`there's at least existence cases for people that just
`live on the level of this box does this function. Like
`
`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 2010
`APPLE v. PUMA, IPR2016-01114
`Page 15 of 76
`
`
`
`ROBERT P. COLWELL
`
`Page 16
`a branch predictor, for instance. You -- you get --
`tell me what program counter you're at, and I'm going to
`try to give you a prediction of do I think that's a
`branch. If it is, do I think you should take it or not
`and predict take it or not, and the decisions that you
`make at that level will flow down through various
`aspects of the machine and will impact performance.
` And you can have discussions, like, in fact,
`architects do. You can have discussions like that all
`day long without ever talking about the chip
`organization physically or how the circuits will
`implement the functions you're envisioning.
` Q. Okay. I guess to a nonengineer, I guess, how
`would you explain the difference between sort of a
`logical representation of something and a physical
`representation?
` A. Well, it depends on the audience. But for
`example, I guess everybody -- I would use analogies, I
`suppose, and everybody knows -- most people know how to
`drive a car, but they don't necessarily know how the
`engine works or what internal combustion means or all
`the nuts and bolts of how the engineer who designed that
`engine had to think about in order to make the car
`reliable and efficient and so on.
` They just know the interface they have is this
`
`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 2010
`APPLE v. PUMA, IPR2016-01114
`Page 16 of 76
`
`
`
`ROBERT P. COLWELL
`
`Page 17
`pedal does that, and this pedal does that. And the
`physics behind it don't have to be visible.
` And in the same sense on a chip, architects can
`consider various concepts that they can worry about the
`implementation of later. They do have -- in my personal
`opinion, they do have to have a sense of what they're
`envisioning will eventually be realizable. I mean, it's
`easy to conjure up visions that actually can't be built,
`and that doesn't do anybody any good. But as long as
`people have some experience and know what they're doing,
`you tend to stay away from that condition, and you can
`have, you know, long fruitful discussions without ever
`referring to a circuit.
` Q. So if you had a logical representation of a
`chip, would that tell you about the physical connections
`on the chip as it's manufactured?
` A. There is some correspondence there. I mean,
`when you -- when you've encapsulated or you've
`envisioned a function on the machine, well, a simple one
`might be an ALU. There's -- everyone knows inside of a
`computer there must be something in there that knows how
`to add, for example. It takes two numbers, add them
`together and produces the result.
` And they're right. There's going to be a box
`in there that does that. And in an architectural level,
`
`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 2010
`APPLE v. PUMA, IPR2016-01114
`Page 17 of 76
`
`
`
`ROBERT P. COLWELL
`
`Page 18
`we call that a functional unit as a generic thing, and
`we say, "What are the inputs and the outputs?"
` So yes, it knows how to add, but add what? So
`you have to pass coming in to bring you the bits you'd
`like to add, and there has to be a path going out, and
`it takes the result, puts it where you want it.
` So yeah. In that sense there's a
`correspondence between what the function does, and you
`have to envision what the flows in and out are going to
`be.
` Q. Okay. So is there perfect correspondence
`between, I guess, the flow shown in a logical
`representation of a chip and the signal lines on the
`physical representation of a chip?
` A. At the highest level, no, because there's
`circuits -- well, there's wires, for example, that you
`just don't need to show because everybody knows they're
`there.
` Power and ground. You wouldn't -- at an
`architectural level of description you wouldn't show
`power and ground because it's redundant. People know
`you don't have that, the thing's not going to work.
`What you wouldn't probably show at all times are major
`buses. Data paths are usually shown because those tend
`to be large and have implications for layout and speed.
`
`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 2010
`APPLE v. PUMA, IPR2016-01114
`Page 18 of 76
`
`
`
`ROBERT P. COLWELL
`
`Page 19
` Depending on the level of discussion of the
`chip, you might or might not show various control
`signals. If they're significant to explaining how the
`thing works, you'll show them. Otherwise, for example,
`a clock, you might or might not show a clock. If you
`leave it out, it will still be understood there is one.
`It just isn't particularly interesting. But you might
`show it if there's something special about it that is
`relevant to the functioning of the box.
` Q. Are there things that would be shown on a
`logical representation of a chip that would not be shown
`on a physical representation?
` A. I can think of one from the case I was deposed
`on two weeks ago.
` Q. Can you tell me about that, or is it
`confidential?
` A. No. It's --
` MR. ALBERTI: Objection. Scope.
` THE WITNESS: It's -- you know, it's fairly
`simple. It's if you have multiple cores on a chip --
`let's pick Intel. I don't know much about anyone else's
`chip, but I only know certain things about theirs at
`this point. But they may have multiple cores. You can
`buy a four-core chip from Intel, and inside the box of a
`chip that they gave you, you'll find four things shown
`
`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 2010
`APPLE v. PUMA, IPR2016-01114
`Page 19 of 76
`
`
`
`ROBERT P. COLWELL
`
`Page 20
`as cores. And yet even if they show you a die photo,
`you'll be able to see which ones. You can see because
`they're organized that way.
` But there's another thing that's called the
`uncore, and it's the thing that these four cores share
`to get outside and talk to memory, talk to a bus.
` But uncore is shared among all four. It's not
`four separate uncores just because there's four cores,
`but there's still conceptual correspondence between what
`one of the cores does and what it -- some of the
`resources it needs in the uncore.
` So while there's no physical connection between
`a core and the piece of the uncore that it wants, there
`is a conceptual one. So it would make sense under
`certain circumstances to draw sort of a dotted box
`around the core and a part of the uncore even though
`there's no physical reality corresponding to that.
` BY MR. CHEN:
` Q. What did you do to prepare for today's
`deposition?
` A. Met with counsel for yesterday and the day
`before.
` Q. And who did you meet with?
` A. Various people, but --
` Q. Everyone at the table?
`
`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 2010
`APPLE v. PUMA, IPR2016-01114
`Page 20 of 76
`
`
`
`ROBERT P. COLWELL
`
`Page 21
`
` A. Not everyone but most.
` Q. Could you tell me who?
` A. Well, now I'll see if I can remember everyone's
`names. David was there. Adam was there. Mike was
`there. Curt was there.
` Who am I forgetting? Oh, Jake. Jake was there
`on Saturday.
` Q. And how long did you meet for?
` A. Yesterday, I don't know. 10:00 to 5:00 or so.
`And the day before just really the afternoon.
` (Exhibit 1, Declaration of Robert P. Colwell
`Ph.D. Under 37 C.F.R. Section 1.68 (123 pages), was
`marked for identification by the court reporter and is
`attached hereto.)
` BY MR. CHEN:
` Q. Okay. You've been handed what's been marked as
`Exhibit 1, and is Exhibit 1 the declaration that you
`submitted in the proceeding labeled IPR2016-01114?
` A. Yes.
` Q. And if you look at the lower right, this has
`also been labeled Exhibit 1003 in this proceeding;
`right?
` A. I see that.
` Q. And if you'll turn to the very last page, this
`is your signature on page 123; 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 2010
`APPLE v. PUMA, IPR2016-01114
`Page 21 of 76
`
`
`
`ROBERT P. COLWELL
`
`Page 22
`
` A. That's correct.
` Q. And you read this declaration carefully before
`you signed it?
` A. I sure did.
` Q. And is everything in here accurate as far as
`you know?
` A. Yes.
` Q. Is there anything that needs to be corrected to
`make it accurate?
` A. Not that I know of.
` Q. How long did it take you to prepare this
`declaration?
` A. Oh, man. This is like a year ago; right?
`August. You know, I don't remember. It's too long ago.
` Q. Remember a ballpark estimate or nothing at all?
` A. I -- it's hard to give an estimate. It's just
`too long ago. Too much stuff has happened after that.
` Q. Do you remember if you discussed the
`declaration with anyone other than Apple's counsel?
` A. You know, I don't remember anyone else besides
`Apple's counsel.
` Q. And at any point did anyone tell you what
`should not be in your declaration?
` A. I'm not sure I understood that one.
` Q. Did anybody ever tell you that, you know,
`
`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 2010
`APPLE v. PUMA, IPR2016-01114
`Page 22 of 76
`
`
`
`ROBERT P. COLWELL
`
`Page 23
`
`something should not be in your declaration?
` A. Oh.
` MR. ALBERTI: Objection. Form.
` THE WITNESS: No, I don't remember anything
`like that.
` BY MR. CHEN:
` Q. And did you rely on anything for this
`declaration that's not cited in the declaration?
` A. I don't think so. Everything should be in
`here.
` (Exhibit 2, U.S. Patent 7,777,753 (21 pages),
`was marked for identification by the court reporter and
`is attached hereto.)
` BY MR. CHEN:
` Q. Okay. You've been handed what's been marked as
`Exhibit 2.
` And this is U.S. Patent No. 7,777,753; right?
` A. Yes.
` Q. If you look at the lower right-hand corner,
`it's also been identified as Exhibit 1001 in this
`proceeding; right?
` A. Right.
` Q. If you could flip to Figure 2.
` Now, Figure 2 shows on kind of the left-hand
`side First Device 42; is that right?
`
`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 2010
`APPLE v. PUMA, IPR2016-01114
`Page 23 of 76
`
`
`
`ROBERT P. COLWELL
`
`Page 24
`
` A. Yes.
` Q. It also shows a decoder that's labeled with a
`number 44; right?
` A. Yes.
` Q. That Decoder 44 is part of a larger block
`labeled 80 that's an encoder slash -- sorry.
`Decoder/encoder; is that correct?
` A. Yes.
` Q. What is the memory interface of Decoder 44?
` A. That's the box labeled 76.
` Q. Is the box labeled 76 implemented using
`circuitry?
` A. I guess I don't see where the patent text
`speaks to that question.
` Q. Well, decoder -- well, step back.
` Block 76 is part of Block 80; right?
` A. Yes.
` Q. Okay. If you look at column 6, line -- around
`42 -- so at the end of that line it says, "The preferred
`embodiment of the invention, the first device 42
`decoder/decoder 80, are in one integrated circuit;
`however, they can be on separate integrated circuits in
`any combination."
` Do you see that?
` 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 2010
`APPLE v. PUMA, IPR2016-01114
`Page 24 of 76
`
`
`
`ROBERT P. COLWELL
`
`Page 25
` Q. So does that mean that block 80 is implemented
`using circuitry?
` A. It -- I would read that as circuitry is
`involved, but there are -- sometimes there's either
`software or firmware also involved, so if the
`distinction matters we have to be more specific.
` Q. Is circuitry involved in the implementation of
`block 76 on Figure 2?
` A. There would have to be some circuitry for sure.
`Whether it's all circuitry or whether there's any kind
`of software/firmware is just not really spoken to here.
` Q. What is the memory interface circuit of Decoder
`44?
` A. I thought you asked me that already.
` Q. So maybe I should clarify. So before I asked
`you what's the memory interface of Decoder 44, and now
`I'm asking you what is the memory interface --
` A. Circuit.
` Q. -- circuit of Decoder 44.
` A. Well, so this circuit, there has to be circuit
`meaning transistor-level stuff associated with the
`interface from Box 80 to Bus No. 70 in Figure 2, so
`there's -- that's guaranteed to be circuits. It has to
`be transistors driving those lines. These other boxes,
`it doesn't speak to that directly.
`
`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 2010
`APPLE v. PUMA, IPR2016-01114
`Page 25 of 76
`
`
`
`ROBERT P. COLWELL
`
`Page 26
` Q. So are you saying that the memory interface
`circuit of Decoder 44 is Block 76 plus circuitry between
`Block 76 and the bus?
` A. I -- I'm assuming that the circuitry that I
`just referred is actually inside of 76. But this figure
`is not accurate enough to be certain. It wouldn't be --
`it wouldn't be crazy, for example, to have the inventor
`intend to mean that there's -- that the I/O drivers
`themselves are actually outside of 76. You could look
`at it either way. I just know that they're in the path.
` Q. What's the difference between the memory
`interface of Decoder 44 and the memory interface circuit
`of Decoder 44?
` A. I don't know if there is one. I mean, the -- I
`don't think the spec is as clear as it could be, but the
`claim 7, for instance, refers to an arbiter included in
`the memory interface circuit of the decoder, and so the
`decoder was -- if the decoder is this Box 80, then the
`memory interface circuit is the thing that has to
`include the arbiter, and the arbiter is shown as Box 82,
`so therefore it seems to be suggesting that the memory
`interface Box 46 and the memory interface circuit are
`the same.
` Q. So are you saying that Arbiter 82 is included
`in the memory interface circuit of Decoder 44?
`
`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 2010
`APPLE v. PUMA, IPR2016-01114
`Page 26 of 76
`
`
`
`ROBERT P. COLWELL
`
`Page 27
` A. That's the -- that is an inference being -- I'm
`drawing from the claim language from 7, claim 7.
` Q. What is the memory interface of first device
`42?
` A. Well, as shown in Figure 2 it would be Box 42
`labeled "Memory Interface."
` Q. Is circuitry involved in the implementation of
`Block 72?
` A. There has to be some circuits involved in
`there. It may not be the only circuits.
` Q. What is the memory interface circuit of First
`Device 42?
` A. The spec -- the spec in column 7 at line 27
`refers to memory interfaces 76, 72 of the
`decoder/encoder 80 and the first device 42. So there I
`believe the spec is calling Box 42 from -- sorry.
`Box 72 from Figure 2 a memory interface.
` Then in the same claim 7, column 16 around line
`28 it says -- the claim language says, The central
`processing unit has a memory interface circuit.
` So again, I'm going to draw the same inference
`that apparently they mean -- they seem to mean that both
`things are the same thing. Memory interface is a memory
`interface circuit according to these guys.
` (Exhibit 3, AT&T DSP 3210 Digital Signal
`
`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 2010
`APPLE v. PUMA, IPR2016-01114
`Page 27 of 76
`
`
`
`ROBERT P. COLWELL
`
`Page 28
`Processor, The Multimedia Solution (40 pages), was
`marked for identification by the court reporter and is
`attached hereto.)
`