` BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`INTEL CORPORATION and )
`XILINX, INC., )
` Petitioners, ) IPR2020-01449
` v. ) Patent No.
`FG SRC LLC, ) 7,149,867
` Patent Owner. )
`
` VIDEOTAPED DEPOSITION OF STANLEY SHANFIELD, Ph.D.
` September 23, 2021
`
` Remote videotaped deposition of
`STANLEY SHANFIELD, Ph.D., commencing at 11:03 a.m.,
`on the above date, before CORINNE T. MARUT, C.S.R.
`No. 84-1968, Registered Professional Reporter,
`Certified Realtime Reporter and Notary Public.
`
`Intel Exhibit 1043 - 1
`
`
`
`Page 2
`
` APPEARANCES
` All Parties Appearing Via Zoom Videoconference
`
`ON BEHALF OF THE PETITIONER INTEL CORPORATION and
`THE WITNESS:
` PILLSBURY WINTHROP SHAW PITTMAN LLP
` 401 Congress Avenue, Suite 1700
` Austin, Texas 78701
` 512-580-9600
` BY: BRIAN C. NASH, ESQ.
` brian.nash@pillsburylaw.com
`
` PILLSBURY WINTHROP SHAW PITTMAN LLP
` 2550 Handover Street
` Palo Alto, California 94304
` 650-233-4087
` BY: MATTHEW W. HINDMAN, ESQ.
` matthew.hindman@pillsburylaw.com
`
`ON BEHALF OF THE PATENT OWNER, FG SRC LLC:
` DIMUROGINSBERG, PC
` DGKeyIP GROUP
` 1101 King Street, Suite 610
` Alexandria, Virginia 22314
` 703-289-5118
` BY: HENNING SCHMIDT, ESQ.
` hschmidt@dimuro.com
`
`ALSO PRESENT:
` RHONDA POLVADO, Paralegal,
` Shore Chan, LLP
`
`REPORTED BY: CORINNE T. MARUT, C.S.R. No. 84-1968
`
`PohlmanUSA Court Reporting
`(877) 421-0099 PohlmanUSA.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
`
`Intel Exhibit 1043 - 2
`
`
`
` I N D E X
`STANLEY SHANFIELD, Ph.D. EXAMINATION
` BY MR. SCHMIDT..................... 4
`
`Page 3
`
` E X H I B I T S
` PREVIOUSLY MARKED EXHIBITS
`EXHIBIT REFERRED TO
`Exhibit 1001 U.S. Patent No. 5
` 7,149,867
`Exhibit 1003 Document, Proceedings, 26
` International
` Conference on Computer
` Design, VLSI In
` Computers and Processors"
`Exhibit 1006 Declaration of Stanley 6
` Shanfield, Ph.D.,
` concerning U.S. Patent
` No. 7,149,867
`Exhibit 1034 Declaration of Stanley 22
` Shanfield, Ph.D., in
` Support of Petitioner's
` Opposition to Patent
` Owner's Motion to Amend
`Exhibit 1037 U.S. Patent No. 64
` 5,737,631
`
`1
`2
`3
`
`4 5 6
`
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`PohlmanUSA Court Reporting
`(877) 421-0099 PohlmanUSA.com
`
`Intel Exhibit 1043 - 3
`
`
`
`Page 4
`
` THE REPORTER: Good morning. We are on the
`record for the deposition of Dr. Stanley Shanfield.
`The date is September 23, 2021, and the time is
`11:03 a.m. Central Standard Time.
` This deposition is taken in the matter
`of Intel Corporation and Xilinx, Inc., Petitioners
`vs. FG SRC, Patent Owner, before the Patent Trial
`and Appeal Board.
` (WHEREUPON, the witness was duly
` sworn.)
` MR. SCHMIDT: Counsel introduce themselves?
` THE REPORTER: I'm sorry. Yes.
` MR. SCHMIDT: This is Henning Schmidt with
`DiMuroGinsberg on behalf of FG SRC.
` MR. NASH: This is Brian Nash of Pillsbury
`Winthrop Shaw Pittman. I'm also joined by my
`colleague Matt Hindman of the same firm, and we are
`here on behalf of Petitioner, Intel Corporation,
`and the witness.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`PohlmanUSA Court Reporting
`(877) 421-0099 PohlmanUSA.com
`
`Intel Exhibit 1043 - 4
`
`
`
`Page 5
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
` STANLEY SHANFIELD, Ph.D.,
`called as a witness herein, having been first duly
`sworn, was examined and testified as follows:
` EXAMINATION
`BY MR. SCHMIDT:
` Q. Good day, Dr. Shanfield. How are you?
` A. Hi. Good, Henning.
` Q. I assume you recall the framework for
`depositions, same as last time?
` A. Yes, I do.
` Q. Great. Is there any reason that you
`cannot or may be impaired in testifying accurately
`today?
` A. No.
` Q. Great. Do you have handy Exhibit 1001
`in this IPR?
` A. 1001.
` Q. It's the patent itself, the '867 patent.
` A. Yes.
` Q. If not, I can drop it into the chat.
` A. I've got it.
` Q. Great. Would you please take a look at
`claim 13.
` A. Okay.
`
`PohlmanUSA Court Reporting
`(877) 421-0099 PohlmanUSA.com
`
`Intel Exhibit 1043 - 5
`
`
`
`Page 6
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
` Q. Do you see the second claim element
`regarding "transferring the data between a
`computational unit and the data access unit"?
` A. Yes, I do.
` Q. Okay. After the section I just read,
`there is -- the next phrase is, "wherein the
`computational unit and the data access unit and the
`data prefetch unit are configured to conform to the
`needs of an algorithm implemented on the
`computational unit and transfer only data necessary
`for computations by the computational unit."
` Do you see that?
` A. Yes.
` Q. What is -- you've looked at this claim
`element before, right, in this claim?
` A. Yes, I have.
` Q. Okay. Could you please give me your
`understanding of this claim element, the section
`that I read.
` MR. NASH: Objection; beyond the scope and
`form.
`BY THE WITNESS:
` A. So, I refer you back to my original
`declaration that's Exhibit 1006, and I'm at
`
`PohlmanUSA Court Reporting
`(877) 421-0099 PohlmanUSA.com
`
`Intel Exhibit 1043 - 6
`
`
`
`Page 7
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`page 98, "Transferring the data between a
`computational unit and a data access unit," that's
`item --
`BY MR. SCHMIDT:
` Q. Dr. Shanfield, I'm sorry to interrupt.
` MR. NASH: I'm sorry. He was answering your
`question.
`BY MR. SCHMIDT:
` Q. Yeah. I'm sorry. Give me one second to
`pull up the exhibit first. Still looking for it.
` Okay. I have the exhibit pulled up.
`Please go ahead. I'm sorry to interrupt.
` A. It's all right.
` So, on paragraph 183, I explain my
`understanding of this element and also mention in
`the paragraph that Zhang and Gupta disclose this
`limitation.
` So, I wrote here, "A computational unit
`is 'a functional unit that performs a
`computation.'" That's directly from the patent.
`And "'a data access unit is a functional unit that
`accesses a component of a memory hierarchy and
`delivers data directly to the computational
`logic.'" Also that's directly from the patent.
`
`PohlmanUSA Court Reporting
`(877) 421-0099 PohlmanUSA.com
`
`Intel Exhibit 1043 - 7
`
`
`
`Page 8
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
` "A functional unit is a set of logic
`that performs a specific operation and the
`operation may be arithmetic, logical, control or
`data movement."
` And I go on to discuss "Functional units
`are used as building blocks of reconfigurable
`logic."
` Q. So, the claim element we're looking at
`refers specifically to wherein the computational
`unit, the data access unit and the data prefetch
`unit, right?
` A. Yes. It refers to those terms, right.
` Q. And regarding those three terms, it says
`they're "configured to conform to the needs of an
`algorithm implemented on the computational unit."
` Do you see that?
` A. Yes, I do.
` Q. What's your understanding of that
`requirement, that limitation, that these
`components, these three listed components, are
`configured to conform to the needs of an algorithm
`implemented on the computational unit?
` MR. NASH: Objection; beyond the scope, form
`and asked and answered.
`
`PohlmanUSA Court Reporting
`(877) 421-0099 PohlmanUSA.com
`
`Intel Exhibit 1043 - 8
`
`
`
`Page 9
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`BY THE WITNESS:
` A. So, if you go to paragraph 185 in my
`original declaration, Exhibit 1006, a few sentences
`in, I talk about what a person skilled in the art
`would understand about Zhang and how that is
`consistent with the claim elements.
` It says, "A person skilled in the art
`would understand that Zhang's processor uses a set
`of logic to access components of the memory
`hierarchy to retrieve computational data."
`BY MR. SCHMIDT:
` Q. You're reading in paragraph 185?
` A. Yes.
` Q. About where?
` A. Well, I'm about to jump to the end of
`page 100 in paragraph 185.
` Q. Okay.
` A. And then I go on to explain, "A person
`of skill in the art would understand that, in order
`to perform matrix computations, the Zhang processor
`requires a functional unit that accesses components
`of the memory hierarchy," which is the L1 or L2
`cache memory, "and delivers prefetch data," which
`in the diagram in Zhang is labeled value 1,
`
`PohlmanUSA Court Reporting
`(877) 421-0099 PohlmanUSA.com
`
`Intel Exhibit 1043 - 9
`
`
`
`Page 10
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`value 2, value 3, et cetera, "directly to the
`computational logic on the processor to perform the
`matrix computations." That's shown in Figures 4
`and 5.
` So, if you look further down at the
`bottom of paragraph 185, on the right is Figure 5.
` And then I go on to explain that a
`person -- and this is now paragraph 186. "A person
`skilled in the art also understands Zhang to
`disclose a data access unit that uses one or more
`registers as a storage element to deliver data
`directly to the computational units."
` Q. So, in the context of the '867 patent,
`what do you understand the phrase "configured to
`conform to the needs of an algorithm implemented on
`the computational unit" to mean?
` MR. NASH: Objection; form, beyond the scope,
`asked and answered.
` I don't know that we can continue down
`this path. I feel like you continue to be asking
`him to provide you with a claim construction or
`something to that effect, which is not the scope of
`what this deposition is intended to cover.
` So, if we're going to keep asking the
`
`PohlmanUSA Court Reporting
`(877) 421-0099 PohlmanUSA.com
`
`Intel Exhibit 1043 - 10
`
`
`
`Page 11
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`same question over and over again, it might be that
`we need to take a pause and get a ruling from the
`Court on whether or not this is appropriate.
` MR. SCHMIDT: Brian, why do you say outside of
`the scope? This is one of the claims that was
`amended, is it not?
` MR. NASH: Yeah, and if you want to ask him
`about his testimony as to the amendment, that's
`fine. But you appear to be asking him to give you
`an on-the-fly claim construction as to an entire
`phrase.
` You're asking -- you're reading a phrase
`and then asking him what he understands it to mean.
`He has given you his analysis as to how he analyzed
`that claim. So, I think he's asked and answered
`that question.
` I don't know what else you're asking
`for. If you want him to give you a claim
`construction on the fly, I don't think he is going
`to be able to do it and I think we'll have to talk
`to the Court about getting a ruling on that.
` So, that's my objection. It's beyond
`the scope to ask him to provide you with a claim
`construction.
`
`PohlmanUSA Court Reporting
`(877) 421-0099 PohlmanUSA.com
`
`Intel Exhibit 1043 - 11
`
`
`
`Page 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
`
` MR. SCHMIDT: His declaration is all about the
`difference in scope between original and amended
`claim, correct?
` MR. NASH: Are you deposing me? I'm not
`answering your questions. If you want to ask him
`questions about his declaration, you can ask him
`questions about his declaration. But --
` MR. SCHMIDT: Then let me --
` MR. NASH: -- you're asking him to give you an
`opinion about what the claim term means.
` MR. SCHMIDT: Then let me set out my position
`before I continue. And then once you understand my
`position, you can make your decision if you want to
`go take this to the Court or not.
` His declaration obviously is about the
`difference in scope between original and amended
`claim. Therefore, establishing the scope of the
`original claim and establishing the scope of the
`amended claim is relevant to determining the
`difference thereto. And that's what I'm asking
`about right now.
` And so far every time I've asked about
`the '867 patent, his question -- his answer begins
`with Zhang discloses blank. He is not answering
`
`PohlmanUSA Court Reporting
`(877) 421-0099 PohlmanUSA.com
`
`Intel Exhibit 1043 - 12
`
`
`
`Page 13
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`the questions. The questions are squarely within
`the scope of the declaration regarding scope of
`these claims, the amended claim, the difference,
`and I will continue asking until I either get an
`answer or you take it to the Court.
` MR. NASH: Well, I don't know that there is
`going to be a resolution on that issue because I
`think he actually has answered your question. He
`gave you the areas where there are constructions on
`those terms, which he pointed to -- I believe it
`was a paragraph from Exhibit 1006 that sets forth
`any constructions that he's applying. Otherwise he
`is using the plain and ordinary meaning of what
`those words are.
` And then to give you an illustration of
`that, he's pointed to how he's analyzed that in the
`context of Zhang.
` So, other than for him to repeat the
`claim language again, what else are you expecting
`him to say?
` If you ask me what does the color blue
`mean, I would tell you that the color blue is blue.
`I don't have any other further explanation of that.
` And you're not going to require him to
`
`PohlmanUSA Court Reporting
`(877) 421-0099 PohlmanUSA.com
`
`Intel Exhibit 1043 - 13
`
`
`
`Page 14
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`do a claim construction right now.
`BY MR. SCHMIDT:
` Q. Dr. Zhang -- I mean, sorry,
`Dr. Shanfield, is your understanding of the phrase
`"is configured to conform to the needs of an
`algorithm implemented on the computational unit"
`simply plain and ordinary meaning?
` A. Well, I actually talk about claim
`construction in my original declaration, and let's
`go there.
` Give me a moment.
` Q. Sure.
` A. So, I'm at paragraph 95 on page 35 of my
`original declaration, Exhibit 1006, and I have
`explained what I used as my understanding of the
`terms listed here. So, computational unit, for
`example, or functional unit.
` Computational unit is paragraph 100.
`And like I explain in that paragraph, the term
`"computational unit" ought to be construed to mean
`a functional unit that performs a computation,
`which is exactly what's written in the patent
`specification.
` And as I wrote, the patent specification
`
`PohlmanUSA Court Reporting
`(877) 421-0099 PohlmanUSA.com
`
`Intel Exhibit 1043 - 14
`
`
`
`Page 15
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`"describes computational units as the functional
`units that perform a specific type of computatio,
`(such as add, subtract and/or multiply)."
` Q. What does it mean to conform to the
`needs of an algorithm in the context of the '867
`patent?
` MR. NASH: Objection; beyond the scope.
`BY THE WITNESS:
` A. So, I had been showing or discussing how
`Zhang provides that or meets that, that claim
`limitation, as an example. And if you will permit
`me to go back there, I can talk further about it.
`If that's not answering your question, that's the
`extent of my consideration.
`BY MR. SCHMIDT:
` Q. You're right inasfar as addressing Zhang
`here does not appear to answer my question.
` My question just is what does it mean to
`conform to the needs of an algorithm in the context
`of the '867 patent?
` MR. NASH: Objection; asked and answered,
`beyond the scope.
`BY MR. SCHMIDT:
` Q. Maybe let me rephrase the question.
`
`PohlmanUSA Court Reporting
`(877) 421-0099 PohlmanUSA.com
`
`Intel Exhibit 1043 - 15
`
`
`
`Page 16
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
` Does that mean that these components
`implemented in reconfigurable logic? Is that what
`it means to be conformed to the needs of an
`algorithm?
` MR. NASH: Objection; form.
`BY THE WITNESS:
` A. So, if you'll allow me, on paragraph 189
`in my -- in Exhibit 1006, I again -- I refer to
`Zhang as an example.
` "Zhang teaches that its key components
`are implemented in programmable logic to enable
`them to be customized to match the algorithms
`instantiated in the computational logic on the
`reconfigurable processor."
` So, in addressing your question,
`Zhang -- in referring to Zhang, there is
`reconfigurable logic that enables it. But that is
`an example.
` Q. So, that would be an example of
`conforming to the needs of an algorithm?
` A. Correct.
` Q. Okay. And that applies to the three
`referenced elements here, right, computational
`units, the access unit and data prefetch unit, is
`
`PohlmanUSA Court Reporting
`(877) 421-0099 PohlmanUSA.com
`
`Intel Exhibit 1043 - 16
`
`
`
`Page 17
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`that right?
` MR. NASH: Objection; form.
`BY THE WITNESS:
` A. I don't understand your question. What
`do you mean "that applies"?
`BY MR. SCHMIDT:
` Q. Well, the claim --
` A. I'm saying Zhang is an example where
`programmable logic like reprogrammable logic such
`as an FPGA is used.
` Q. Okay. So, the claim element here says
`"wherein," and then it lists the three components,
`the computational unit, the data access unit and
`the data prefetch unit, "are configured to conform
`to the needs of an algorithm."
` So, the example you gave applies to
`those three listed components in the claim element,
`right?
` MR. NASH: Objection; beyond the scope, asked
`and answered.
`BY THE WITNESS:
` A. Yeah, I think I've explained that the
`combination of Zhang and Gupta discloses this
`limitation, and then I go on in paragraph 189 to
`
`PohlmanUSA Court Reporting
`(877) 421-0099 PohlmanUSA.com
`
`Intel Exhibit 1043 - 17
`
`
`
`Page 18
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`explain why.
` Zhang teaches a reconfigurable
`processor. It has computational units, a data
`access unit and a data prefetch unit implemented in
`reprogrammable logic.
`BY MR. SCHMIDT:
` Q. And that in your opinion meets the
`requirement of this claim element, right?
` MR. NASH: Objection; beyond the scope. I
`mean, we're not talking about the motion to amend
`or his declaration related to the motion to amend.
` So, I guess we'll need to call the Court
`and have a ruling on my objection. We are wasting
`the witness' time and everybody else's time.
`BY MR. SCHMIDT:
` Q. Are you going to answer the question?
` MR. NASH: Let's get the Court on the phone.
` MR. SCHMIDT: You're welcome to instruct the
`witness not to answer, Brian, and then call the
`Court.
` MR. NASH: I'm not instructing the witness not
`to answer. I'm asking for a ruling on my motion if
`you're going to go beyond the scope.
` MR. SCHMIDT: If you're not instructing the
`
`PohlmanUSA Court Reporting
`(877) 421-0099 PohlmanUSA.com
`
`Intel Exhibit 1043 - 18
`
`
`
`Page 19
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`witness not to answer, then please be quiet and let
`him answer.
` MR. NASH: I need a ruling on my objection,
`and under the rules I'm permitted to get one.
` MR. SCHMIDT: Well, are you shutting down the
`deposition for now or not?
` MR. NASH: I'm not shutting it down. I'm
`asking that we get the PTAB on the phone and we get
`a ruling on the objection.
` MR. SCHMIDT: So, can I continue with my
`questions?
` MR. NASH: No, we need a ruling on this.
` MR. SCHMIDT: Okay. So you want to hold off
`for a minute.
` Are you going to call the Board?
` MR. NASH: I think you're supposed to be the
`one to call the Board.
` MR. SCHMIDT: I'm not going to do that. I'm
`going to continue my deposition until and unless
`you shut it down.
` MR. NASH: Well, we're going to need a get a
`ruling on the objection if you want to go down this
`line of questioning. If you want to ask about
`other questions and leave this to the end, I am
`
`PohlmanUSA Court Reporting
`(877) 421-0099 PohlmanUSA.com
`
`Intel Exhibit 1043 - 19
`
`
`
`Page 20
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`happy to do that or we can get a ruling now if this
`is the area that you want to focus on right now.
` MR. SCHMIDT: You have two options, either --
`I told you what I'm going to do. I'm going to ask
`my questions, and I'm going to ask the questions
`that I want to ask until I am finished within the
`allowed time. And if you want to instruct the
`witness not to answer my questions, then you do
`that. But I told you what I'm going to do. That's
`it.
` MR. NASH: Then I suggest we take a break and
`we call the PTAB and get them on the phone to rule
`on my objection.
` MR. SCHMIDT: I expect you to do that. I have
`told you what I'm going to do. So...
` If you are willing to do that, then we
`can take a break until you're, you know, ready to
`continue.
` MR. NASH: That works.
` MR. SCHMIDT: Okay. Sounds good. Let's go
`off the record, then, if Brian agrees.
` And, Brian, do you want to call the
`Court on the record or?
` MR. NASH: Yeah, I'm going to have to look
`
`PohlmanUSA Court Reporting
`(877) 421-0099 PohlmanUSA.com
`
`Intel Exhibit 1043 - 20
`
`
`
`Page 21
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`things up on how to do that.
` MR. SCHMIDT: Okay. We can take a ten-minute
`break for you to do that and then get back on when
`you're ready.
` MS. POLVADO: Going off the record at 11:29
`a.m.
` MR. SCHMIDT: Hold on a second. Brian hasn't
`agreed yet. We need all parties to agree.
` It looks like Brian is disconnected for
`now.
` MS. POLVADO: So, do you want me to stop
`recording?
` MR. SCHMIDT: Yeah, I guess so. I will be
`online, you know. I suppose Brian -- I don't know.
`I said ten minutes. He just left. I don't know if
`he means to be back in ten minutes or not.
` I guess let's go offline for 10 minutes
`or maybe let's say 15 minutes, until 11:45 a.m.,
`and then I'll be back online and we'll see what
`happens next.
` MS. POLVADO: Okay. Going off the record now
`at 11:30 a.m.
` MR. SCHMIDT: Thank you.
` (WHEREUPON, a recess was had
`
`PohlmanUSA Court Reporting
`(877) 421-0099 PohlmanUSA.com
`
`Intel Exhibit 1043 - 21
`
`
`
`Page 22
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
` from 11:30 to 12:08 a.m.)
` THE REPORTER: We are back on the record for
`the second part of the deposition of Dr. Stanley
`Shanfield. The date is September 23, 2021, and the
`time is 11:08 p.m. -- 12:08 p.m. I'm sorry.
`BY MR. SCHMIDT:
` Q. Sorry for the extended break,
`Dr. Shanfield. We will shift gears a little bit.
` MR. NASH: Henning, just real quick, I just
`want to state we conferred over the break and I
`just -- my understanding of what we are going to do
`is try and reframe the discussion so that we are
`asking questions that are directed to the scope of
`this testimony, which is Exhibit 1034.
` So, my understanding is he has offered
`no testimony on claim 13 or the limitations that we
`have been discussing to this point. But if you
`want to try and establish questions that are
`related to his testimony from Exhibit 1034, we can
`continue.
` MR. SCHMIDT: Okay. We'll shift gears a
`little bit.
`BY MR. SCHMIDT:
` Q. Let's take a look at Exhibit 1034, which
`
`PohlmanUSA Court Reporting
`(877) 421-0099 PohlmanUSA.com
`
`Intel Exhibit 1043 - 22
`
`
`
`Page 23
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`is your second declaration. Let me know when you
`have it.
` A. Yep, I've got it.
` Q. Okay. Will you please take a look at
`paragraph 7. Have you seen it?
` A. Yes.
` Q. Are you looking at it?
` A. I'm looking at it, yes.
` Q. Okay. Let me know when you're ready to
`continue.
` A. Okay. I'm ready.
` Q. Okay. So, in here you're pointing out
`that you believe that claim amendments alter the
`scope of the claim somewhat, is that right?
` MR. NASH: Objection; form.
`BY THE WITNESS:
` A. Yeah, I'm arguing that the amendment
`changes the scope of the claim because it no longer
`requires the data prefetch unit itself to do
`prefetching from the second memory.
`BY MR. SCHMIDT:
` Q. How would the data arrive at the data
`prefetch unit?
` MR. NASH: Objection; form, calls for
`
`PohlmanUSA Court Reporting
`(877) 421-0099 PohlmanUSA.com
`
`Intel Exhibit 1043 - 23
`
`
`
`Page 24
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`speculation.
`BY THE WITNESS:
` A. I think the point is that the amended
`claim language means that a unit other than the
`data prefetch unit could retrieve computational
`data from the second memory, and that -- that's the
`problem.
`BY MR. SCHMIDT:
` Q. Looking at paragraph 8 in your
`declaration, you state here that the '867 patent
`"specifically identifies the data prefetch unit as
`the component that is responsible for prefetching
`the computational data." Right?
` A. Yes.
` Q. Do you -- can you point to any
`disclosure in the '867 patent that points to any
`other component performing that function instead?
` MR. NASH: Objection; form, beyond the scope.
`BY THE WITNESS:
` A. So, I think what I point out in
`paragraph 8 is -- or, rather, in paragraph 7 is the
`main point, which is the fact that another unit
`altogether could retrieve the computational data
`from a second memory. So, that's -- that's an
`
`PohlmanUSA Court Reporting
`(877) 421-0099 PohlmanUSA.com
`
`Intel Exhibit 1043 - 24
`
`
`
`Page 25
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`increase in scope. And that's the point, not --
`not what I can hypothesize might be true.
`BY MR. SCHMIDT:
` Q. In your opinion, does the '867 patent
`support the scope that you think would be disclosed
`here, which is that some other component retrieves
`the computational data from the second memory?
` A. No.
` Q. Okay. Can you take a look at paragraph
`14, please. At the very end, the last sentence of
`paragraph 14, you refer to Zhang here.
` Do you see that?
` A. Give me a moment.
` Q. It's on page 8.
` A. Yes. I see that.
` Q. Okay. So, you state here that "Zhang
`shows integrating programmable logic within the CPU
`itself, in addition to the cache, network
`interface, and memory." Right?
` A. Yes, if you look at Figure 2, that's
`clearly shown.
` Q. What's your understanding -- what do you
`mean by saying, "Zhang shows integrating
`programmable logic within the CPU itself"?
`
`PohlmanUSA Court Reporting
`(877) 421-0099 PohlmanUSA.com
`
`Intel Exhibit 1043 - 25
`
`
`
`Page 26
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
` A. Well, if you go to Figure 2 in Zhang,
`what you see is, and that's page 152, Exhibit 1003,
`I guess that's page 14. Sorry. He shows CPU with
`programmable logic in it, in the figure.
` Q. Okay. Can you point me to any other
`support in Zhang for your statement that the CPU --
`I'm sorry -- for your statement that Zhang shows
`integrating programmable logic within the CPU
`itself?
` A. Sure. At the bottom of page 13 in Zhang
`in 1003, he starts by saying, "We propose an
`architecture that integrates small blocks of
`programmable logic into key elements of a baseline
`architecture, including processing elements."
` So, the CPU is a processing element.
` Q. Okay.
` A. So, I think it says it pretty clearly.
` Q. Okay. Is there any other support in
`Zhang for your statement?
` A. His examples, particularly --
`specifically the sparse matrix multiplication
`example.
` Q. What part of that example specifically,
`in your opinion, supports that Zhang shows
`
`PohlmanUSA Court Reporting
`(877) 421-0099 PohlmanUSA.com
`
`Intel Exhibit 1043 - 26
`
`
`
`Page 27
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`integrating programmable logic within the CPU
`itself?
` A. So, if you head back to my Exhibit 1034,
`page 10, I write that, and this is the middle of
`paragraph 14, "Zhang confirms these calculations
`are performed using computational elements
`implemented in programmable logic, which includes
`an FPGA."
` And that's because what Zhang says is
`that "by adding a small amount of programmable
`logic to the memory units, we can yield some
`benefits of having computational elements within
`the memory."
` Q. Okay. This refers specifically to
`memory and memory units. Does this support your
`statement that Zhang shows integrating programmable
`logic within the CPU itself?
` MR. NASH: Objection; form, lacks foundation.
`BY THE WITNESS:
` A. Well, I think Figure 2 is probably the
`best indication, the fact that the programmable
`logic is indicated in Figure 2 right in the box
`that's labeled "CPU."
`BY MR. SCHMIDT:
`
`PohlmanUSA Court Reporting
`(877) 421-0099 PohlmanUSA.com
`
`Intel Exhibit 1043 - 27
`
`
`
`Page 28
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
` Q. Okay. And we have already discussed
`that. Okay.
` So, is there any other support in Zhang
`that we have not yet discussed?
` MR. NASH: Objection; form.
`BY THE WITNESS:
` A. I think a person of skill in the art
`reading Zhang would have been well aware of how to
`use FPGAs to perform sparse matrix computations. I
`think I cited five IEEE articles.
` And, in fact, I myself had experience at
`that time with sparse matrix computations using
`FPGAs. This is for an application at Draper Labs
`where we did multiple Fourier transforms to get a
`very accurate GPS location fix. And that's what
`made it easy for me to be able to pull out many
`articles like the ones I've showed in my report.
` So, a person of skill in the art, to
`them it would be obvious to use an FPGA to perform
`these computations required by the algorithm.
`BY MR. SCHMIDT:
` Q. Is there any further support you can
`point to for the statement that Zhang shows
`integrating programmable logic within the CPU
`
`PohlmanUSA Court Reporting
`(877) 421-0099 PohlmanUSA.com
`
`Intel Exhibit 1043 - 28
`
`
`
`Page 29
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`itself?
` A. Well, another thing a person of skill in
`the art would understand is that there really
`weren't a lot of options. There are essentially a
`few choices that an engineer could make in
`implementing computational units in programmable
`logic. And Zhang identifies, like I said, the
`FPGAs is one of two options along with LSI logic.
` So, I think, like I listed, I think it's
`paragraph 16 that the alternatives were PLDs,
`programmable logic arrays, or PALs (sic), or PROMs
`or EPROMs.
` So, this was a relatively limited number
`of options for a person of skill in the art, and
`that would again make it likely that they would
`think of FPGAs.
` Also, I -- as I explain at the bottom of
`that paragraph 16, that because in referencing the
`combination of Zhang and Gupta, Gupta is teaching
`reconfigurable logic blocks in FPGA for
`application-specific cache organization policies.
` So, it would have been clear to a person
`of skill in the art as a way to implement Zhang's
`reconfigurable data prefetch architecture.
`
`PohlmanUSA Court Reporting
`(877) 421-0099 PohlmanUSA.com
`
`Intel Exhibit 1043 - 29
`
`
`
`Page 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
`
` Q. Would that include making it clear that
`these programmable components could be integrated
`within the CPU itself?
` A. Well, at the time, I mean -- I don't
`understand -- I guess I don't understand your
`question since at the time FPGAs had CPUs in them,
`and that was routine.
` A CPU -- a small, you know, relatively
`low-power or not very powerful relative to today's
`CPUs, but they were incorporated into the FPGAs of
`the day. I remember them being included in the
`ones I used at the time.
` So, I don't understand your que