throbber
HAROLD STONE
`
`Page 1
` UNITED STATES PATENT AND TRADEMARK OFFICE
` BEFORE THE PATENT TRIAL AND APPEAL BOARD
` HTC AMERICA INC., HTC CORP.,
` LG ELECTRONICS, INC., LG
` ELECTRONICS U.S.A., INC., LG
` ELECTRONICS MOBILECOMM
` U.S.A., INC., SAMSUNG Case No. IPR2015-01500
` ELECTRONICS CO., LTD., and Patent (7,321,368 B2)
` SAMSUNG ELECTRONICS AMERICA,
` INC., Case No. IPR2015-01501
` Patent (7,777,753 B2)
` Petitioner,
` Case No. IPR2015-01502
` vs. Patent (7,542,045 B2)
` PARTHENON UNIFIED MEMORY
` ARCHITECTURE LLC,
` Patent Owner.
` ____________________________/
`
` DEPOSITION OF HAROLD STONE, PH.D.
` San Francisco, California
` Thursday, August 11, 2016
`
`REPORTED BY:
`LESLIE ROCKWOOD, RPR, CSR 3462
`
`212-279-9424
`
`Veritext Legal Solutions
`www.veritext.com
`
`212-490-3430
`
`1
`
`23
`
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`HTC and LG v. PUMA, IPR2015-01500
`Patent Owner - Exhibit 2013, p. 1
`
`

`
`HAROLD STONE
`
`Page 2
`
` UNITED STATES PATENT AND TRADEMARK OFFICE
` BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
` HTC AMERICA INC., HTC CORP.,
` LG ELECTRONICS, INC., LG
` ELECTRONICS U.S.A., INC., LG
` ELECTRONICS MOBILECOMM
` U.S.A., INC., SAMSUNG
` ELECTRONICS CO., LTD., and Case No. IPR2015-01500
` SAMSUNG ELECTRONICS AMERICA, Patent (7,321,368 B2)
` INC.,
` Case No. IPR2015-01501
` Petitioner, Patent (7,777,753 B2)
` vs. Case No. IPR2015-01502
` Patent (7,542,045 B2)
` PARTHENON UNIFIED MEMORY
` ARCHITECTURE LLC,
`
` Patent Owner.
` ____________________________/
`
` Deposition of HAROLD STONE, PH.D., taken on behalf
`of the Patent Owner, at the offices of Sidley Austin LLP,
`555 California Street, Suite 2000, San Francisco,
`California, beginning at 8:57 A.M. and ending at 11:56
`A.M., on Thursday, August 11, 2016, before Leslie
`Rockwood, RPR, CSR No. 3462.
`
`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
`
`HTC and LG v. PUMA, IPR2015-01500
`Patent Owner - Exhibit 2013, p. 2
`
`

`
`HAROLD STONE
`
`Page 3
`
`APPEARANCES:
`
`FOR THE PATENT OWNER PARTHENON UNIFIED MEMORY
`ARCHITECTURE LLC:
` AHMAD, ZAVITSANOS, ANAIPAKOS, ALAVI & MENSING
` BY: JUSTIN CHEN, ESQ.
` 1221 McKinney, Suite 2500
` Houston, Texas 77010
` (713) 655-1101
` jchen@azalaw.com
`
`FOR THE PETITIONER HTC AMERICA INC. AND HTC CORP.:
` SIDLEY AUSTIN, LLP
` BY: STEPHEN M. EVERETT, ESQ.
` 555 California Street, Suite 2000
` San Francisco, California 94104
` (415) 772-1235
` stephen.everett@sidley.com
`
`1
`
`23
`
`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
`
`HTC and LG v. PUMA, IPR2015-01500
`Patent Owner - Exhibit 2013, p. 3
`
`

`
`HAROLD STONE
`
`Page 4
`
` I N D E X
`
`THURSDAY, AUGUST 11, 2016
`
`WITNESS EXAMINATION
`HAROLD STONE, PH.D.
`
` BY MR. CHEN 6
`
`QUESTIONS WITNESS INSTRUCTED NOT TO ANSWER:
` (NONE)
`
`1
`
`234
`
`56
`
`7
`
`89
`
`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
`
`HTC and LG v. PUMA, IPR2015-01500
`Patent Owner - Exhibit 2013, p. 4
`
`

`
`HAROLD STONE
`
`Page 5
`
` DEPOSITION EXHIBITS
` HAROLD STONE, PH.D.
`NUMBER DESCRIPTION IDENTIFIED
`Exhibit 7 Reply Declaration of Harold 15
` S. Stone, Ph.D, 7.15.16
`Exhibit 8 Recent Developments in the 35
` Design of Image and Video
` Processing ICs, Konstantinos
` Konstantinides, et al.
`Exhibit 9 DSP3210 Information Manual 52
` Introduction
`Exhibit 10 US Patent Number: 5,546,547, 57
` 8.13.96
`Exhibit 11 Interactive Video from 71
` Desktops to Settops,
` Frederick Kitson, et al.
`Exhibit 12 Declaration of Mitchell A. 72
` Thornton, Ph.D., PUMA Exhibit
` 2009
`
`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
`
`HTC and LG v. PUMA, IPR2015-01500
`Patent Owner - Exhibit 2013, p. 5
`
`

`
`HAROLD STONE
`
`Page 6
` San Francisco, California; Thursday, August 11, 2016
` 8:57 A.M.
`
` PROCEEDINGS
`
` HAROLD STONE, PH.D.,
` called as a witness, having been duly sworn, was
` examined and testified as follows:
`
` EXAMINATION
`BY MR. CHEN:
` Q. Good morning, Dr. Stone.
` A. Good morning.
` Q. Could you please state your full name and
`address for the record.
` A. My name is Harold Stuart Stone. My address
`is 223 6th Avenue, Kirkland, Washington 98033.
` Q. And have you ever had your deposition taken
`before?
` A. Yes, I have.
` Q. How many times?
` A. Many times. I haven't been able to count.
`Okay.
` Q. All right.
` MR. EVERETT: Justin, before we go on, do you
`
`1
`2
`
`34
`
`56
`
`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
`
`HTC and LG v. PUMA, IPR2015-01500
`Patent Owner - Exhibit 2013, p. 6
`
`

`
`HAROLD STONE
`
`Page 7
`
`want to put appearances on the record?
` MR. CHEN: Yeah, yeah, sure. Justin Chen with
`AZA for the plaintiff Parthenon Unified Memory
`Architecture LLC.
` MR. EVERETT: And Stephen Everett, Sidley
`Austin, for defendant HTC -- excuse me -- for petitioner
`HTC.
` Q. BY MR. CHEN: All right. So I'm sure you're
`familiar with all the procedures relating to having your
`deposition taken?
` A. Yes, I am.
` Q. Great. And so you understand that you've taken
`an oath to tell the truth today?
` A. Yes, I have.
` Q. And if at any point you don't understand a
`question that I'm asking you, will you please let me
`know?
` A. I will.
` Q. And if at a later point you remember something
`you'd like to add to a previous answer, will you let me
`know?
` A. I will.
` Q. Okay. And we'll try to take a break every hour
`or so, but if you need to take a break earlier, will you
`let me 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
`
`HTC and LG v. PUMA, IPR2015-01500
`Patent Owner - Exhibit 2013, p. 7
`
`

`
`HAROLD STONE
`
`Page 8
`
` A. I will.
` Q. Is there anything that could prevent you from
`offering truthful and accurate testimony today?
` A. Not that -- not that I'm aware of, no.
` Q. What did you do to prepare for today's
`deposition?
` A. I did some reading. I met with counsel
`yesterday, and that's it.
` Q. Okay. What did you read in preparation for
`today's deposition?
` A. I read Dr. Thornton's report and my report, my
`reply.
` Q. Did you read anything else?
` A. Possibly. That was -- what I told you was where
`I spent my time.
` Q. And how long did you meet with counsel to
`prepare for today's deposition?
` A. Approximately, six -- let's see. Approximately,
`seven hours.
` Q. And who did you meet with?
` A. I met with Mr. Everett, with Kurt Holbine?
` MR. EVERETT: Holbreich.
` THE WITNESS: Hol- -- can you spell that for me?
` MR. EVERETT: H-O-L-B-R-E-I-C-H.
` THE WITNESS: C-H. Holbreich.
`
`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
`
`HTC and LG v. PUMA, IPR2015-01500
`Patent Owner - Exhibit 2013, p. 8
`
`

`
`HAROLD STONE
`
`Page 9
`
` MR. EVERETT: Holbreich.
` THE WITNESS: And Mike Bet- --
` MR. EVERETT: Bettinger.
` THE WITNESS: Bettinger, B-E-T-T-I-N-G-E-R.
` Q. BY MR. CHEN: Now, have you ever worked with
`video decoding prior to these IPR proceedings?
` A. Yes, I have.
` Q. And when did you do that?
` A. If you have my vitae, I can point it out.
` Q. I don't have it with me. I guess what can you
`remember off the top of your head?
` A. Okay. The -- there was a case I was involved in
`a few years ago. It was MicroUnity versus a number of
`defendants. I represented Qualcomm and others, and that
`case entailed several aspects of cell phones, including
`video decoders.
` Q. Okay. And did those video decoders use the MPEG
`standard?
` A. Yes, they do. Let me take a moment and turn off
`my phone.
` (Interruption in proceedings.)
` THE WITNESS: Okay. They -- they -- repeat your
`question, please, and I'll --
` Q. BY MR. CHEN: Oh, yes. I asked you if those
`video decoders in that case used the MPEG standard?
`
`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
`
`HTC and LG v. PUMA, IPR2015-01500
`Patent Owner - Exhibit 2013, p. 9
`
`

`
`HAROLD STONE
`
`Page 10
`
` A. They did.
` Q. And when else have you worked with video
`decoding that you can remember right now?
` A. Oh, I do. At my last deposition, I pointed out
`two patents that issued in the early 2000s, and both of
`them involved using image transformations, one was on the
`IDCT, inverse discrete cosine transform, and that's
`embedded in the MPEG standard, and the other one used
`related transform to do what I called image registration.
`And the -- these transforms and the types of operations
`that were disclosed in those two patents are common to
`the MPEG standard.
` Q. Can you remember anything else that you've done
`with respect to video decoding?
` A. If I had my -- my vitae, I could probably point
`out something. Those are the things that are on the top
`of my head, because they -- at least those patents came
`up at the last deposition.
` Q. Uh-huh. And so are you familiar with the MPEG
`standard?
` A. Quite.
` Q. Can you explain what that is briefly?
` A. I can explain it at various levels. Let me do
`it at a very high level. At a high level, the objective
`of the MPEG standard is to compress video by large
`
`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
`
`HTC and LG v. PUMA, IPR2015-01500
`Patent Owner - Exhibit 2013, p. 10
`
`

`
`HAROLD STONE
`
`Page 11
`
`factors, and typically, it's 100-to-1.
` It does it in a way that allows you to
`decompress, and the process of compressing and
`decompressing produces a pretty good image, but not a
`perfect representation of what you had originally, but
`it's good enough, and it's widely, widely used for
`broadcast and other purposes because of the advantage of
`the 100 to 1 compression.
` Q. And does MPEG include different number of
`standards?
` A. I'm not sure how to answer your question
`exactly. If you -- your question is vague to me, and so
`if you can sharpen it up, that would be --
` Q. Sure. I can rephrase.
` A. Okay.
` Q. Are there different versions of the MPEG
`standard?
` A. Yes, there are different versions of the MPEG
`standard.
` Q. What are the different versions?
` A. Well, the original version was called MPEG.
`Subsequently MPEG-2 was an enhanced version of it, and
`they went back and renamed MPEG to MPEG-1. Subsequently,
`there are other things that are called MPEG now with
`different numbers, but there -- primarily, it's MPEG-2
`
`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
`
`HTC and LG v. PUMA, IPR2015-01500
`Patent Owner - Exhibit 2013, p. 11
`
`

`
`HAROLD STONE
`
`Page 12
`
`and MPEG-1 that people are using.
` Q. Is there a version 3 of the MPEG standard?
` A. There is something called MPEG-3. It's not --
`not the same kind of thing as MPEG-2.
` Q. How is it different?
` A. I'm not -- I didn't prepare on that. I'm sorry.
`I can -- I can go through a few things that I'm aware of,
`but it -- without any information in front of me, I'd be
`speculating. All I can say is that it does not address
`the kind of compression and decompression that MPEG-2
`addresses.
` Q. Are you familiar with MPEG-4?
` A. A little bit. The same situation there. It --
`it -- it is an enhancement of MPEG-2, and it doesn't do
`the compression and decompression the same way, and it's
`not as widely used as MPEG-2.
` Q. So you're more familiar with MPEG-1 and 2; is
`that right?
` A. That's correct.
` Q. Are you familiar with something called Px64?
` A. Yes, I am.
` Q. And what is that?
` A. That was an earlier standard. It preceded MPEG,
`or MPEG-1 as they now call it. It was a low resolution,
`low repetition rate standard, and it was put in place
`
`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
`
`HTC and LG v. PUMA, IPR2015-01500
`Patent Owner - Exhibit 2013, p. 12
`
`

`
`HAROLD STONE
`
`Page 13
`
`primarily for video conferencing.
` Q. And how is Px64 different from MPEG-1 and 2?
` A. Well, it's not.
` Q. So is it the same as MPEG-1 and 2?
` A. Well, that -- it -- it's incorporated in MPEG-1
`and 2. So MPEG-1 -- if you have an MPEG-1 decoder, you
`can decode Px64 streams. Okay. So I would say that
`MPEG-1 is an enhancement. MPEG-2 is an enhancement, but
`the -- the mechanism for compression and decompression is
`the same for Px64 as it is for MPEG-1 and MPEG-2.
` Q. So if you were complying with the Px64 standard,
`would you also be compliant with, say, the MPEG-1
`standard?
` A. Yes, you would.
` Q. And if you were compliant with the Px64
`standard, would you also be compliant with the MPEG-2
`standard?
` A. I'm not -- I believe that -- yes. Yes, it is
`the case. You can't go the other way, though. The -- in
`other words, if you produced a MPEG-2 bit stream and you
`had a Px64 decoder, you would not be able to decode it.
` Q. I see.
` A. Unless it were that part of MPEG-2 that happened
`to be conformed to Px64.
` Q. And had if you had a MPEG-1 stream, would a
`
`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
`
`HTC and LG v. PUMA, IPR2015-01500
`Patent Owner - Exhibit 2013, p. 13
`
`

`
`HAROLD STONE
`
`Page 14
`
`Px64 decoder be able to handle that?
` A. If it were a Px64 compatible stream, yes, and
`if -- if it were something that was beyond the Px64, it
`would not be able to handle it.
` Q. Have you ever designed an MPEG-compliant device?
` A. That's a good question, but I'm not sure if I
`can answer it "yes" or "no." I pointed to a patent at
`the last deposition where I described a scheme for doing
`an inverse discrete cosine transform, and that would be
`to decode images, and particularly, it would decode MPEG
`images that are called iframes or ipictures, and so to
`the extent that I can de- -- I designed something novel,
`I have a patent on it, that decodes some aspects of MPEG,
`then I have designed an MPEG device. I have not designed
`a complete MPEG decoder.
` Q. And this may be the same answer, but have you
`ever physically built an MPEG-compliant device?
` A. Physically built? Well, the answer is yes, to
`the extent that this MPEG -- this IDCT device that I
`designed, I implemented and it ran. And it's software
`that you put it in a hardware, and it works, so that --
`that hardware now becomes a MPEG-compliant device, at
`least to the extent that it can decode iframes.
` Q. And for this patent that you're talking about,
`was that ever implemented in a commercial product?
`
`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
`
`HTC and LG v. PUMA, IPR2015-01500
`Patent Owner - Exhibit 2013, p. 14
`
`

`
`HAROLD STONE
`
`Page 15
` A. I am not aware of it being implemented in a
`commercial product, no.
` (Exhibit 7, Reply Declaration of Harold S.
` Stone, Ph.D, 7.15.16, marked for
` identification.)
` Q. BY MR. CHEN: Okay. You've been handed what's
`been marked as Stone Exhibit 7. Now, this is the reply
`declaration that you submitted in the three IPR
`proceedings listed on the first page; is that correct?
` A. That's correct.
` Q. And if you take a look at the very last page, if
`you just flip it over, and this is your signature here;
`is that correct?
` A. That's correct. I signed in red, so this is a
`copy.
` Q. Yes. And you read this declaration carefully
`before you signed it; is that right?
` A. I did.
` Q. Okay. And did you write everything in your
`declaration?
` A. There were some edits that were added by other
`parties, but the answer is I wrote this from start to
`finish, and subject to a few changes here and there, I --
`I accept all the changes, and I'm the author.
` Q. And you had a chance to review and revise 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
`
`HTC and LG v. PUMA, IPR2015-01500
`Patent Owner - Exhibit 2013, p. 15
`
`

`
`HAROLD STONE
`
`Page 16
`
`declaration before you signed it?
` A. I did.
` Q. And is everything in here accurate, as far as
`you know?
` A. As far as I know, it's accurate. If there are
`typos, I haven't found them.
` Q. Is there anything that needs to be corrected to
`make it accurate?
` A. I'm aware of nothing.
` Q. And how long did it take you to prepare this
`declaration?
` A. My belief is under two days. I'm speculating
`now. I could be wrong, but I -- it's a reasonable
`estimate under two days.
` Q. Did you discuss this declaration with anyone?
` A. With counsel.
` Q. And counsel for which parties?
` A. That's a hard question, because I'm representing
`several parties. Certainly with LG and with HTC. I
`believe that's all.
` Q. You didn't speak to counsel for Apple, for
`example?
` A. I have spoken to counsel for Apple, but I have
`not discussed this with counsel for Apple.
` Q. Did anyone tell you what should not be in your
`
`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
`
`HTC and LG v. PUMA, IPR2015-01500
`Patent Owner - Exhibit 2013, p. 16
`
`

`
`HAROLD STONE
`
`Page 17
`
`declaration?
` A. The answer probably is no, but I'm -- I'm
`confused, because when I write a declaration, I put in
`what I have to put in, and everything else I don't put
`in. So I -- you understand that -- what -- to be
`instructed not to put in certain things, I mean, there's
`a whole universe out there. I received no instructions
`on how to write this. Maybe that's a better --
` Q. Yes.
` A. -- better answer.
` Q. And did you rely on anything for this
`declaration that's not cited in the declaration?
` A. I was careful to keep track of everything that I
`relied on, and I -- and it's in an exhibit of materials
`relied on. That's accurate.
` Q. Okay. If we can take a look at paragraph 2.
` And in paragraph 2 you cite to a paragraph from
`Dr. Thornton's declaration; right?
` A. That's correct.
` Q. And look at paragraph 3. The first sentence
`says, "There's no support for this opinion."
` Do you see that?
` A. I do.
` Q. So does that mean you disagree with the
`paragraph cited in paragraph 2?
`
`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
`
`HTC and LG v. PUMA, IPR2015-01500
`Patent Owner - Exhibit 2013, p. 17
`
`

`
`HAROLD STONE
`
`Page 18
` A. That says that Dr. Thornton hasn't offered a
`reason for his -- his opinion, so when I say there's no
`support for his opinion, that by itself doesn't say I
`disagree, but the next sentence says, "On the contrary,"
`and it offers contrary evidence, so the bottom line is I
`do disagree, but not in the first sentence that you
`cited.
` Q. So the first sentence of Dr. Thornton's
`paragraph that you cited says, "Typically, a decoder
`requires its own dedicated memory."
` Do you see that?
` A. I do.
` Q. And is that something you disagree with?
` A. Well, yes.
` Q. And so it's your opinion that at the time of the
`invention, a typical decoder would not have a dedicated
`memory; is that true?
` A. The problem is what the word "typically" means
`in this case. I mean, Dr. Thornton might be looking at a
`certain restricted set, and in that set, typically it has
`something. This set is larger, and from all the decoders
`that were out there, some have dedicated memory, some do
`not. And I list a certain number, I think it looks like
`half a dozen or so, prior art references that do operate
`without a dedicated memory. They have shared memory and
`
`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
`
`HTC and LG v. PUMA, IPR2015-01500
`Patent Owner - Exhibit 2013, p. 18
`
`

`
`HAROLD STONE
`
`Page 19
`
`not dedicated.
` And so I'm not claiming that what I've described
`is typical, nor what Dr. Thornton described is typical.
`What I'm saying is there are lots of decoders out there.
`Some have dedicated memory. Some have shared memory.
` Q. So are you able to opine on whether a typical
`decoder at the time of the invention had a dedicated
`memory?
` A. No, because the word "typical" is -- is what's
`difficult here. I'm able to opine that there are
`decoders out there that had dedicated memory and that
`there are decoders out there that had shared memory, but
`I'm not ready to say what's typical.
` Q. How about what was conventional at the time of
`the invention?
` A. It's the same problem, conventional, typical.
`It -- it depends what universe you're looking at, you
`know.
` Q. If you can turn to paragraph 12, and the first
`sentence here says, "Professor Thornton also opines,
`incorrectly, that the MPEG format am incompatible with
`floating-point and that a fixed-point DSP is necessary
`for MPEG decoding."
` And so are you saying that a floating-point DSP
`is -- would be compatible with the MPEG standard?
`
`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
`
`HTC and LG v. PUMA, IPR2015-01500
`Patent Owner - Exhibit 2013, p. 19
`
`

`
`HAROLD STONE
`
`Page 20
`
` A. Yes, I am.
` Q. And if a floating-point DSP were performing MPEG
`decoding, would pixels be represented by a floating-point
`value?
` A. It -- you know, the answer is possibly, possibly
`not. You have many different ways of implementing a
`floating-point decoder, and if you were really adamant in
`not representing a pixel as a floating-point number, you
`could do that.
` Q. And for a floating-point DSP performing MPEG
`decoding, would color components be represented by a
`floating-point value?
` A. I'm confused. I'm not sure exactly what you
`mean by "color components." Do you mean the chroma? Do
`you mean the pixels?
` Q. Well, we could take RGB, for example. For a
`pixel there would be, I think, color components for the
`red, green and blue.
` A. Those are the pixels. I would -- when you asked
`about pixels, I -- so you asked a question earlier, and I
`think I answered it.
` Q. So are you saying that both the pixels and the
`color components could be represented by floating-point
`values?
` A. I said that I don't distinguish pixels from the
`
`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
`
`HTC and LG v. PUMA, IPR2015-01500
`Patent Owner - Exhibit 2013, p. 20
`
`

`
`HAROLD STONE
`
`Page 21
`color components. And my answer was for pixels that they
`may or may not be represented by floating-point values.
`It depends on who -- who wrote the code.
` Q. Do you know what a typical number of bits used
`for a floating-point value would be in a DSP?
` A. Please don't use the word "typical."
` Q. Okay. What word would you like me to use?
` A. Okay. You can say what are ranges of numbers
`that are used in floating point for DSPs.
` Q. Sure. What's the answer to that question?
` A. Okay. I've seen DSPs that have 32 bits of -- 32
`bits of precision, and that would be including mantissa
`and exponent, and they -- there are -- at the time, 1995,
`there were floating-point implementations that had up to
`64 bits of precision, which would include mantissa and
`exponent.
` Now, things are even more strong, but I'm
`answering for the 1995 time frame. So you -- today, you
`would go to 128 bits, but that -- I'm not -- I'm not
`claiming that 128 was available in -- in 1995.
` Q. So say around 1995, would a pixel be -- strike
`that.
` Back in 1995, are you aware of any systems that
`would use a 32-bit floating-point value to represent a
`pixel?
`
`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
`
`HTC and LG v. PUMA, IPR2015-01500
`Patent Owner - Exhibit 2013, p. 21
`
`

`
`HAROLD STONE
`
`Page 22
`
` A. Yes.
` Q. And what systems are they?
` A. Well, the -- I cited several. We have the
`systems that use the DSP3210 cited by Bowes. I also
`cited to a floating-point implementation, and I
`believe -- let me find that.
` I'd like to point to my paragraph number 21,
`subsection I, that appears on page 21. It cites to a
`paper by Ramaswamy and Miller. The title of the paper is
`"Efficient Implementation of the Two-Dimensional Discrete
`Cosine Transform For Image Coding Applications on the DSP
`96002 Processor." And it has a date of 1993.
` That's another DSP floating point, and it was
`used to implement the discrete cosine transform portion
`of MPEG. So those are two. And it -- I'm not limiting
`my -- my testimony to those two, but that's sufficient.
` Q. And in this Ramaswamy paper that you mentioned,
`do you recall if they used a 32-bit floating-point value
`to represent pixels?
` A. I don't have it in front of me, and the best I
`can do is from memory and speculate, and at the risk of
`being wrong, I don't think I'll speculate.
` Q. Okay. If we could look at paragraph 13.
` And in the second sentence here it says, "The
`MPEG data stream encodes data as signed magnitude
`
`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
`
`HTC and LG v. PUMA, IPR2015-01500
`Patent Owner - Exhibit 2013, p. 22
`
`

`
`HAROLD STONE
`
`Page 23
`
`integers of variable precision."
` Do you see that?
` A. I do.
` Q. Now, is it faster to convert signed magnitude
`integers of variable precise to floating-point format or
`fixed-point format?
` A. It's the same instruction, basically. It would
`be the same -- same time. Let's just say it's one
`instruction. It's a table lookup. You say, "I want to
`convert the number that's in position 16 to floating
`point," and in the table, there's the floating-point
`number. You pull it out. It goes in the register. Or
`you say, "I want to convert the number that's the same
`location to fixed point," then you store a fixed-point
`number.
` THE REPORTER: Then you would?
` MR. EVERETT: You'd store a fixed-point number.
` THE WITNESS: Store a fixed-point number. And
`then the instruction would load that fixed-point number
`and put it in a register. So you would preload your
`table with floating-point numbers or fixed-point numbers,
`depending how you would like to do the conversion.
` Q. BY MR. CHEN: And so it would take about the
`same speed?
` A. About the same time.
`
`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
`
`HTC and LG v. PUMA, IPR2015-01500
`Patent Owner - Exhibit 2013, p. 23
`
`

`
`HAROLD STONE
`
`Page 24
` Q. Okay. And the next sentence in paragraph 13
`says, "These data have to be converted to a scaled two's
`complement format for fixed-point calculations or to a
`floating-point format for floating-point calculations so
`that a format conversion is required whether the DSP uses
`floating-point or fixed-point representations."
` Do you see that?
` A. That's correct.
` Q. So is it your opinion that signed magnitude
`integers of variable precision must be converted to
`either fixed-points or floating-point form in order for
`the integers to be decoded using a DSP?
` A. Yes. I mean, let me -- let me try to say that
`maybe, maybe somewhere, you could find an arithmetic unit
`that operates on sign magnitude. That would be rare.
`And in that rare instance, perhaps you'd be able to load
`some sign magnitude integers from the MPEG data stream
`without conversion. However, they are of variable
`scaling, so no matter what happens, you would still have
`to rescale them, and you would do that by this table
`lookup.
` So the answer is I -- I cannot conceive of a way
`of taking data from an MPEG stream and putting it into
`fixed-point format without doing a conversion.
` Q. But is it possible to perform decoding 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
`
`HTC and LG v. PUMA, IPR2015-01500
`Patent Owner - Exhibit 2013, p. 24
`
`

`
`HAROLD STONE
`
`Page 25
`on the sign magnitude integers of variable precision?
` A. You would have to have a very unusual arithmetic
`unit to do that. And if you say it's possible, hey,
`somebody might design one, but doubtful. Doubtful. For
`the arithmetic units used in practice in 1995, you'd have
`to have conversion.
` Q. Have you ever seen an unusual arithmetic unit
`like that?
` A. Not to my knowledge. Oh, I'd like to add to it.
`The MPEG standard shows in the conversion from signed
`magnitude to two's complement fixed point. It's -- it's
`in the exemplary code in the standard, and so if somebody
`skilled in the art would want to implement the standard,
`they would look at that exemplary code and do what it
`says.
` Q. Is it your opinion that a scaled two's
`complement format and floating-point format are the only
`two possible formats that could be used to perform MPEG
`decoding calculations?
` A. I don't have an opinion on what are the possible
`formats. Those two are possible.
` Q. If you look at the last sentence of this
`paragraph 13, it says, "Such extended precision or
`additional scaling adds computational overhead for
`fixed-point calculations but is absent in floating-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
`
`212-279-9424
`
`Veritext Legal Solutions
`www.veritext.com
`
`212-490-3430
`
`HTC and LG v. PUMA, IPR2015-01500
`Patent Owner - Exhibit 2013, p. 25
`
`

`
`HAROLD STONE
`
`Page 26
`
`calculations."
` Do you see that?
` A. I do.
` Q. Okay. So for MPEG decoding, is it your opinion
`that floating-point calculations are preferable compared
`to fixed-point calculations?
` A. They have advantages. Whether they're
`preferable, I'd have to look at the implementations.
`There are some aspects of floating point which may be
`slower or less desirable. There may be some aspects that
`are desirable. At least with respect to scaling,
`floating point is superior to fixed point.
` Q. So could there be scenarios where floating-point
`calculations are superior for MPEG decoding?
` A. There could be.
` Q. And similarly, are there situations where
`fixed-point calculations could be preferable for MPEG
`decoding?
` A. There could be.
` Q. Are you aware of any systems that use
`floating-point calculations for MPEG decoding?
` A. Well, I have given the -- the citation to
`Ramaswamy and -- I need to talk to counsel for a moment.
` MR. CHEN: Sure. We could take a break.
` THE WITNESS: Can we take a break on 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
`
`HTC and LG v. PUMA, IPR2015-01500
`Patent Owner - Exhibit 2013, p. 26
`
`

`
`HAROLD STONE
`
`Page 27
`
` MR. CHEN: Yeah, we can take a break.
` (Discussion off the record.)
` Q. BY MR. CHEN:

This document is available on Docket Alarm but you must sign up to view it.


Or .

Accessing this document will incur an additional charge of $.

After purchase, you can access this document again without charge.

Accept $ Charge
throbber

Still Working On It

This document is taking longer than usual to download. This can happen if we need to contact the court directly to obtain the document and their servers are running slowly.

Give it another minute or two to complete, and then try the refresh button.

throbber

A few More Minutes ... Still Working

It can take up to 5 minutes for us to download a document if the court servers are running slowly.

Thank you for your continued patience.

This document could not be displayed.

We could not find this document within its docket. Please go back to the docket page and check the link. If that does not work, go back to the docket and refresh it to pull the newest information.

Your account does not support viewing this document.

You need a Paid Account to view this document. Click here to change your account type.

Your account does not support viewing this document.

Set your membership status to view this document.

With a Docket Alarm membership, you'll get a whole lot more, including:

  • Up-to-date information for this case.
  • Email alerts whenever there is an update.
  • Full text search for other cases.
  • Get email alerts whenever a new case matches your search.

Become a Member

One Moment Please

The filing “” is large (MB) and is being downloaded.

Please refresh this page in a few minutes to see if the filing has been downloaded. The filing will also be emailed to you when the download completes.

Your document is on its way!

If you do not receive the document in five minutes, contact support at support@docketalarm.com.

Sealed Document

We are unable to display this document, it may be under a court ordered seal.

If you have proper credentials to access the file, you may proceed directly to the court's system using your government issued username and password.


Access Government Site

We are redirecting you
to a mobile optimized page.





Document Unreadable or Corrupt

Refresh this Document
Go to the Docket

We are unable to display this document.

Refresh this Document
Go to the Docket