`
`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
`
` R. HORST - 6/8/2018
` IN THE UNITED STATES DISTRICT COURT
` FOR THE EASTERN DISTRICT OF TEXAS
` MARSHALL DIVISION
`
`ALACRITECH, INC., )
` )
` Plaintiff, )
`VS ) Case No. 2:16-cv-693
` ) Case No. 2:16-cv-692
`CENTURY LINK, et al., ) Case No. 2:16-cv-695
` )
` Defendants.)
` )
`INTEL CORPORATION, )
` )
` Intervenor,)
`and )
` )
`CAVIUM, INC., )
` )
` Intervenor.)
`
` DEPOSITION OF ROBERT HORST
` JUNE 8, 2018
` HOUSTON, TEXAS
`
`REPORTED BY: Linda Russell, CSR
`JOB NO: 142731
`
`TSG Reporting - Worldwide 877-702-9580
`
`
`
` R. HORST - 6/8/2018
`
`Page 2
`
` Friday, June 8, 2018
` 9:37 a.m.
`
` DEPOSITION OF ROBERT HORST, produced as a
`witness at the instance of the Patent Owner, and
`duly sworn, was taken in the above-styled
`and numbered cause on June 8, 2018, from
`9:37 a.m. to 11:41 a.m., before Linda Russell,
`CSR, RPR, CLR in and for the State of Texas,
`reported by machine shorthand, at the offices of
`Weil, Gotshal & Manges, 700 Louisiana, Suite
`1700, Houston, Texas.
`
`TSG Reporting - Worldwide 877-702-9580
`
`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
`
`
`
`Page 3
`
` R. HORST - 6/8/2018
` A P P E A R A N C E S
`
`ATTORNEYS FOR THE PLAINTIFFS:
` ZIYONG "SEAN" LI, ESQ.
` Quinn Emanuel Urquhart & Sullivan
` 50 California Street
` San Francisco, California 94111
`
`ATTORNEYS FOR INTEL:
` Justin Constant, ESQ.
` Weil, Gotshal & Manges
` 700 Louisiana
` Houston, Texas 77002
`
`TSG Reporting - Worldwide 877-702-9580
`
`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
`
`
`
`Page 4
`
` I N D E X
` Page
`EXAMINATION:
` BY MR. LI 6
`REPORTER'S CERTIFICATION 62
`ERRATA SHEET 63
` EXHIBITS MARKED
` No. Description Page
` Exhibit 1 U.S. Patent Number 5,937,169, 9
` INTEL Ex. 1043.001 - 019
` Exhibit 2 Declaration of Robert Horst, 10
` Ph.D. in Support of
` Petitioner's Reply to Patent
` Owner's Response to Petition
` for Inter Partes Review of U.S.
` Patent No. 9,055,104, INTEL Ex.
` 1223.001 - 021
` Exhibit 3 Declaration of Robert Horst in 14
` Support of Petitioner's Reply
` to Patent Owner's Response to
` Petition for Inter Partes
` Review of U.S. Patent No.
` 7,673,072, INTEL Ex. 1223.001 -
` 024
` Exhibit 4 Declaration of Robert Horst, 15
` Ph.D. in Support of Petition
` for Inter Partes Review of U.S.
` Patent No. 7,673,072, INTEL Ex.
` 1003.001 - 153
` Exhibit 5 Declaration of Robert Horst in 25
` Support of Petitioner's Reply
` to Patent Owner's Response to
` Petition for Inter Partes
` Review of U.S. Patent No.
` 7,237,036, INTEL Ex. 1223.001 -
` 024
`
`TSG Reporting - Worldwide 877-702-9580
`
`1
`2
`
`3
`
`4
`
`56
`
`7
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`25
`
`
`
`1
`
`2
`
`3
`4
`
`5
`
`6
`
`7
`8
`
`9
`
`10
`
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`Page 5
`Exhibit 6 Declaration of Robert Horst, 26
` Ph.D. in Support of Petition
` for Inter Partes Review of U.S.
` Patent No. 7,237,036, INTEL Ex.
` 1003.001 - 137
`Exhibit 7 Declaration of Robert Horst in 33
` Support of Petitioner's Reply
` to Patent Owner's Response to
` Petition for Inter Partes
` Review of U.S. Patent No.
` 7,337,241, INTEL Ex. 1223.001 -
` 031
`Exhibit 8 U.S. Patent Number 5,768,618, 34
` INTEL Ex. 1005.001 - 013
`
`Exhibit 9 "Gigabit Ethernet Technical 38
` Brief, Achieving End-to-End
` Performance, Alteon Networks,
` INTEL Ex. 1033.001 - 026
`
`TSG Reporting - Worldwide 877-702-9580
`
`
`
`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
`25
`
` R. HORST - 6/8/2018
` ROBERT HORST,
`having been first duly sworn, testified as
`follows:
` EXAMINATION
`BY MR. LI:
` Q. Good morning, Dr. Horst.
` A. Good morning.
` Q. I guess you have been deposed before?
` A. Yes.
` Q. Just to make sure everyone is on the
`same page, let me just repeat some ground rules.
` In this deposition, the court
`reporter is going to record what we say and it
`will be very hard for her to do her job if we
`talk over each other, so we should avoid that.
` And during the deposition your
`attorney is going to make a series of objections,
`but if he doesn't instruct you not to answer, you
`still have to answer my question. You understand
`that?
` A. Yes.
` Q. We can take a break every hour or,
`you know, hour and a half, it's flexible, but if
`there is a pending question, before the break,
`
`TSG Reporting - Worldwide 877-702-9580
`
`
`
`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
`25
`
` R. HORST - 6/8/2018
`you will have to answer the question before we
`take the break. Do you understand that?
` A. Yes.
` Q. Is there any reason that you cannot
`testify under oath today?
` A. No.
` Q. Any reason preventing you from
`telling the truth?
` A. No.
` Q. Did you prepare for this deposition?
` A. Yes.
` Q. How did you prepare for it.
` MR. CONSTANT: Objection.
`Privileged.
` You can answer to the extent it
`doesn't reveal any communications you may have
`had with attorneys as part of your preparation.
` A. I read over background material, my
`reports, some of the exhibits and so on, and met
`with attorneys.
` Q. When did you read those materials?
` A. Over the last week.
` Q. How long did you spend on them?
` A. In reading materials, probably less
`
`TSG Reporting - Worldwide 877-702-9580
`
`
`
`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
`25
`
` R. HORST - 6/8/2018
`than eight hours total.
` Q. When did you meet your attorney?
` A. Yesterday.
` Q. Who did you meet with?
` A. I met with three attorneys here at
`Weil.
` Q. Who are they?
` A. Garland, Stevens, and Melissa -- I
`can't remember her last name -- and Justin.
` Q. How long was that meeting?
` A. It was around eight hours.
` Q. Did you review any materials to
`refresh your memory over the meeting?
` A. During the meeting I went over my
`reports and some other materials.
` Q. Did you review any prior art
`references?
` A. I'm sorry?
` Q. Did you review any prior art
`references?
` A. Yes, I looked at some of the
`references.
` Q. Which ones did you look at?
` A. Primarily Erickson, but I can't
`
`TSG Reporting - Worldwide 877-702-9580
`
`
`
`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
`25
`
` R. HORST - 6/8/2018
`remember which other ones I looked at.
` MR. LI: Okay. I'm going to
`introduce Exhibit Number 1.
` (Exhibit 1 marked for identification.)
` Q. (BY MR. LI) Exhibit Number 1, the
`Bates number is INTEL Exhibit 1043001. Do you
`recognize what that is?
` A. Yes. This is the Connery reference.
` Q. Are you familiar with this reference?
` A. Yes.
` Q. Did you read this reference
`yesterday?
` A. I reviewed some of it, yes.
` Q. Let's go to page 4 of this reference,
`Sheet 3 of 6, Figure 4. Do you see Figure 4?
` A. Yes.
` Q. Do you see number 103 is "TCP"?
` A. Yes, I see that. Can you tell me
`what part of my expert report this is in
`reference to?
` Q. Did you cite this reference against
`any of the asserted patents?
` A. Yes.
` Q. Do you know which patent you asserted
`
`TSG Reporting - Worldwide 877-702-9580
`
`
`
`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
`25
`
` R. HORST - 6/8/2018
`this reference against?
` A. That would be the '104 patent, I
`believe.
` Q. So let's ask questions about this
`reference.
` A. I'd like to see the -- my report on
`the '104 patent.
` MR. LI: Okay. I'm going to
`introduce the next exhibit, Exhibit Number 2.
` (Exhibit 2 marked for identification.)
` Q. (BY MR. LI) Do you recognize this
`document?
` A. Yes.
` Q. Is this the report of the '104
`Patent, your declaration regarding the '104
`Patent?
` A. This is my reply declaration, not the
`original, but it's the patent owner reply.
` Q. If you'd go to page 14. Do you see
`Petersen and Connery in section number 2?
` A. Yes.
` Q. Okay. Then let's talk about Connery.
` So this the same Connery you cited
`there; is that correct?
`
`TSG Reporting - Worldwide 877-702-9580
`
`
`
`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
`25
`
` R. HORST - 6/8/2018
` A. Yes, that's correct.
` Q. If you look at Figure 4, do you
`recognize 112 of Figure 4?
` MR. CONSTANT: Objection. Form.
` Q. (BY MR. LI) It's on the lower right
`corner of the page.
` A. Box 112 is labeled, "Data Offset."
` Q. Or number 2 -- no. I think 112 refer
`to the entire block, but do you recognize the big
`block on the lower side of the sheet, including
`Source Port, Destination Port, Sequence Number,
`Acknowledge Number, Data Offset, Reserved, and a
`series of fields, and Window, and Checksum, and
`Urgent Pointer, Options, Padding, and Data? Do
`you recognize that box?
` A. This entire box is the TCP header.
` Q. So what sequence number is TCP
`header?
` MR. CONSTANT: Objection, form.
`Lacks foundation.
` A. What part of my report are you
`referring to about sequence number? I've been
`deposed on sequence numbers before, and TCP in
`general before. So what part of my new report
`
`TSG Reporting - Worldwide 877-702-9580
`
`
`
`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
`25
`
` R. HORST - 6/8/2018
`are you referring to here?
` Q. So I'm just asking this because you
`used this reference as a prior reference against
`the patent. And I'm just asking those general
`questions about this reference and the background
`so we can understand the reference better.
` Sitting here, do you have any
`knowledge about the term "sequence number" in
`this figure?
` MR. CONSTANT: Objection. Form.
` A. The sequence number is a field in the
`TCP header that is used to determine the sequence
`of the TCP packets.
` Q. And do you see "Window" there?
` A. Yes.
` Q. What is the purpose of this Window
`field?
` A. In my original report I have a pretty
`long section about how -- what the Window field
`means and how it's used. I can -- I'll refer to
`that if -- as explanation of the Window field.
` Q. Can you give a high-level summary and
`tell us what Window field is?
` MR. CONSTANT: Objection. Form.
`
`TSG Reporting - Worldwide 877-702-9580
`
`
`
`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
`25
`
` R. HORST - 6/8/2018
`Asked and answered.
` A. If you give me my original report on
`this, I will refer to that section.
` Q. Sitting here without looking at your
`original report, do you have any understanding of
`this Window field?
` MR. CONSTANT: Objection. Form.
`Asked and answered.
` A. I have an understanding of how the
`Window field is used in general. But if I am
`asked about the details, I want to be sure to be
`correct and I would refer to my report for that.
` Q. Can you describe the Window field in
`general?
` MR. CONSTANT: Objection. Form.
` A. As I said, I've -- I have described
`the Window field in my report and I will refer to
`that report to tell you how the Window field
`works.
` Q. So what does it take to determine
`this Window field for a TCP header?
` MR. CONSTANT: Objection. Form.
` A. I don't really understand the
`question. The determination of how the Window
`
`TSG Reporting - Worldwide 877-702-9580
`
`
`
`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
`25
`
` R. HORST - 6/8/2018
`field works is a fairly complex process,
`described in my report, described in Tanenbaum.
`There's not a short answer to that question.
` Q. Does the sender need to determine the
`Window field before it send out the TCP packet?
` MR. CONSTANT: Objection. Form.
`Asked and answered.
` A. The Window field is a part of the TCP
`header, so the Window field does have to be
`determined before the packet is sent out.
` MR. LI: I'm going to introduce the
`next exhibit, which is Exhibit Number 3.
` (Exhibit 3 marked for identification.)
` Q. (BY MR. LI) Exhibit -- the Bates
`number of Exhibit Number 3 is INTEL Exhibit
`1223.001. Do you recognize this document?
` A. Yes. This is my reply declaration in
`the '072 patent.
` Q. Can you turn to page 21.
` A. Page 21?
` Q. Yes.
` A. Yes.
` Q. If you'll look at the first
`paragraph, the first full sentence is, "In two of
`
`TSG Reporting - Worldwide 877-702-9580
`
`
`
`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
`25
`
` R. HORST - 6/8/2018
`the proposed TCP scripts, the data transfer is
`performed by multiple segment-sized DMA
`operations." Do you see that?
` A. Yes.
` Q. And if you'll flip back to page 20,
`Section VI on page 20 is about, "Dividing, by the
`Interface Device, the Data Into Segments." Do
`you see that?
` A. Yes.
` Q. And that is one of the element of the
`'072 patent; is that correct?
` MR. CONSTANT: Objection. Form.
` A. Yes, dividing is part of some of the
`claims in the '072.
` MR. LI: Okay. And I'm going to
`introduce the next exhibit, which is Exhibit
`Number 4.
` (Exhibit 4 marked for identification.)
` Q. (BY MR. LI) Do you recognize this
`exhibit?
` A. Yes. This is my original declaration
`in the '072 patent.
` Q. And just for the record, the Bates
`number on this exhibit is INTEL Exhibit 1003.001.
`
`TSG Reporting - Worldwide 877-702-9580
`
`
`
`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
`25
`
` R. HORST - 6/8/2018
` Let's look back to your reply report,
`page 21, to the sentence that I just read to you.
`"In two of the proposed TCP scripts, the data
`transfer is performed by multiple segment-sized
`DMA operations." How does that relate to this
`claim element, "Dividing, by the Interface
`Device, the Data Into Segments"?
` MR. CONSTANT: Objection. Form.
`Outside of the scope.
` A. In my original report, I describe TCP
`scripts that could be run in the Erickson
`adapter. And those TCP scripts would perform the
`segmentation, which is the operation that relates
`to the dividing.
` Q. Can you look at your expert report,
`the original expert report, and find out where
`you discussed that?
` A. Dividing Limitation, which is
`described in Claim 1.4. And the pages of the
`report are INTEL 1003.100 through 1003.104.
` Q. Can you find the -- the TCP scripts
`discussed in your reply report in your opening
`report?
` A. The first script is on page 1003.102.
`
`TSG Reporting - Worldwide 877-702-9580
`
`
`
`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
`25
`
` R. HORST - 6/8/2018
` Q. Uh-huh.
` A. And then the other scripts are on the
`following pages.
` Q. Okay. Let's look at the first
`script. How does it divide the data into
`segments?
` A. Reading from the report, "The,
`'dividing, by the interface device, into
`segments' limitation is met by the foregoing
`obvious TCP script for Erickson. Erickson's
`senduserdatagram user process triggers the I/O
`adapter to send each segment by spanking the GO
`register. The adapter calculates the TCP
`checksum, transmits the packet, updates the
`shared state in Hardware Register 504, and waits
`for GO to be set again for the next segment.
`Repeated invocations of the TCP script, by
`repeatedly spanking the GO register, with
`pointers to data for consecutive segments result
`in the interface device dividing the user data
`stream into TCP segments."
` Q. And why would repeated locations of
`the TCP script with pointers to data for
`consecutive segments result in the interface
`
`TSG Reporting - Worldwide 877-702-9580
`
`
`
`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
`25
`
` R. HORST - 6/8/2018
`device dividing the user data stream into TCP
`segments?
` MR. CONSTANT: Objection. Form.
` A. The user data stream is longer than a
`single segment. So in this script, one segment
`at a time is sent each time the GO register is
`touched, or they call it spanking here.
` Q. So what would the adapter do after
`the GO register is spanked?
` MR. CONSTANT: Objection. Form.
` A. The adapter would send that one TCP
`segment.
` Q. To the area interface?
` MR. CONSTANT: Objection. Form.
` A. I didn't understand.
` Q. So you said the -- you said that the
`adapter would send that one TCP segment. And I
`guess the question is where does it send the
`segment to?
` A. The adapter is sending the segment to
`the destination. It's sending it across the
`Ethernet.
` Q. Yeah. But that's sending the
`segment. Why it divides the data into segments?
`
`TSG Reporting - Worldwide 877-702-9580
`
`
`
`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
`25
`
` R. HORST - 6/8/2018
` MR. CONSTANT: Objection. Form.
` A. The contiguous buffer on the host is
`not sent all at once. It has to be divided into
`multiple segments.
` So this -- assuming that the term
`"dividing" is construed in that way, dividing
`would mean taking just a piece of that large
`buffer, or dividing that large buffer into
`individual smaller buffers, each the size of a
`TCP segment.
` Q. But why that is done by the interface
`device?
` MR. CONSTANT: Objection. Form.
` A. The interface device is -- in this
`script is transferring only one segment's worth
`at a time. So that -- it's dividing the big
`buffer by extracting only one segment's worth of
`data.
` Q. And if you'll look at the second
`script, how does the second script divide data
`into segments?
` A. Reading from my original report:
`"The second script would transfer, via DMA, and
`transmit one maximum-segment-size, MSS, segment
`
`TSG Reporting - Worldwide 877-702-9580
`
`
`
`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
`25
`
` R. HORST - 6/8/2018
`of data, 'in practice, usually about 1500 bytes,
`at a time from the block identified by the user
`data address pointer and length passed to the
`script, until all of the data from the block was
`sent."
` Q. Why that would divide the data into
`segments?
` A. The second script is similar to the
`first, except that the GO register is spanked
`only once, so that each time the adapter is ready
`to send a new TCP segment, it copies the data for
`that segment from the host to the adapter. And
`that copying of data is not the entire buffer,
`but just that segment's worth, so that it is
`dividing that big buffer into smaller chunks.
` Q. And you also have a third script. Do
`you see that?
` A. Yes.
` Q. And how would a third script divide
`data into segments?
` MR. CONSTANT: Objection. Form.
` A. Again, reading from the original
`report: "The 'dividing the data into multiple
`segments' limitation is also met by a third
`
`TSG Reporting - Worldwide 877-702-9580
`
`
`
`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
`25
`
` R. HORST - 6/8/2018
`obvious TCP script for Erickson. Like the second
`alternative script, this third alternative script
`builds on the single segment script, but requires
`only one spank of the GO register for a
`multi-segment send. Instead of transferring one
`MSS-sized segment of user data at a time, the
`adapter would DMA all of the user data identified
`by the user data address pointer and length in
`one large transfer. The adapter would then
`repeatedly extract one segment of data at a time
`from the transferred block, encapsulate it in a
`packet, and transmit."
` In this case, the dividing is where I
`say "repeatedly extract one segment of data at a
`time." So the full buffer is residing on the
`adapter, but it is dividing by extracting one
`segment at a time.
` Q. What is the size of the buffer on the
`adapter here?
` MR. CONSTANT: Objection. Form.
`Outside of the scope.
` A. The buffer size could be -- is a
`design choice that the designer would make.
`Typically it would be larger than one segment,
`
`TSG Reporting - Worldwide 877-702-9580
`
`
`
`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
`25
`
` R. HORST - 6/8/2018
`but it could -- wouldn't have to be more than one
`segment's worth.
` Q. How would the adapter know when to
`stop transferring data?
` MR. CONSTANT: Objection. Form.
` A. I'm not sure what you mean by
`"transferring data," if you're talking about
`moving it to the adapter or moving it from the
`adapter to the Ethernet.
` Q. Moving from the adapter to the
`Ethernet. So let me clarify my question.
` Under your third script, how would
`the adapter know when to stop transferring data
`from the adapter to Ethernet?
` MR. CONSTANT: Objection. Form.
` A. I don't give the details in this
`brief description of what the script would do,
`but, in general, if it had no more data to send,
`it would stop sending. So when the buffer became
`empty or didn't have a full segment anymore, it
`would stop sending.
` Q. Is there any other reason the buffer
`would stop sending?
` MR. CONSTANT: Objection. Form.
`
`TSG Reporting - Worldwide 877-702-9580
`
`
`
`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
`25
`
` R. HORST - 6/8/2018
`Outside the scope.
` A. There -- there could be other
`termination conditions. I don't specify those
`here.
` Q. Sitting here, do you know any other
`stop conditions?
` MR. CONSTANT: Objection. Form.
`Outside the scope.
` A. There is part of the TCP protocol
`that has control bits -- a finish bit that says
`that the TCP segment is -- or that TCP send is
`ending. So it would have to account for that as
`well.
` Q. Any other stop conditions?
` MR. CONSTANT: Objection. Form.
`Outside the scope.
` A. There could possibly be other
`conditions. I haven't detailed those. It
`would -- the full conditions are described in the
`code for TCP, in particular the BSD code that's
`described in the -- the Stevenson references.
` Q. Other than those stop conditions
`described in the Stevenson references, sitting
`here, can you think of any other stop conditions?
`
`TSG Reporting - Worldwide 877-702-9580
`
`
`
`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
`25
`
` R. HORST - 6/8/2018
` MR. CONSTANT: Objection. Form.
`Outside of the scope.
` A. Those references describe the full
`requirements of the TCP protocol, so that's
`the -- they would describe the other conditions.
` Q. Would stop conditions include the
`consideration of the window size?
` MR. CONSTANT: Objection. Form.
` A. The window size is part of the TCP
`protocol described in the Stevenson references.
`So that's -- that code would account for window
`size.
` Q. And would window size be part of the
`stop conditions?
` MR. CONSTANT: Objection. Form.
` A. Yes, the window size is another one
`of the conditions that causes it to stop, if --
`because the window size indicates that the
`receive side no longer has room to accept the
`writes.
` Q. So before all the packets are
`transferred out to the Ethernet, how would the
`adapter know if a window size has been reached?
` MR. CONSTANT: Objection. Form.
`
`TSG Reporting - Worldwide 877-702-9580
`
`
`
`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
`25
`
` R. HORST - 6/8/2018
`Outside of the scope.
` A. The code to handle window size is
`described in Stevenson. And so that -- that code
`would just be ported and be part of this script.
` MR. LI: So let's look at another
`exhibit, which is Exhibit Number 5.
` (Exhibit 5 marked for identification.)
` Q. Do you recognize what this exhibit
`is?
` A. This is the reply declaration for the
`'036 patent.
` Q. For the record, the Bates number of
`this exhibit is INTEL Exhibit 1223.001.
` Did you review this reply report
`yesterday?
` A. Yes.
` Q. If you'd go to page 10, paragraph 27.
`Do you see that?
` A. Yes.
` Q. It says, "As I discussed in my
`original declaration, the Alacritech 1997
`Provisional Application states that the header
`prediction code or fast path in the FreeBSD
`release forms the basis of Alacritech's receive
`
`TSG Reporting - Worldwide 877-702-9580
`
`
`
`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
`25
`
` R. HORST - 6/8/2018
`offload to its INIC." Do you see that?
` A. Yes.
` Q. Do you see the term "head
`prediction"?
` A. Yes.
` Q. "Header prediction."
` What does that mean?
` A. Header prediction code is code that
`was developed by Van Jacobson, and it divides the
`implementation of TCP into a fast-path and a
`slow-path, where the fast-path handles the --
`more of the normal conditions and the slower path
`handles the exception conditions.
` Q. Did you discuss this header
`prediction in your original report -- in your
`opening report?
` A. I believe I did, yes.
` MR. LI: Let me introduce another
`exhibit, which is Exhibit Number 6.
` (Exhibit 6 marked for identification.)
` Q. Do you recognize this exhibit?
` A. Yes. This is my original declaration
`in the '036.
` Q. For the record, the Bates number of
`
`TSG Reporting - Worldwide 877-702-9580
`
`
`
`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
`25
`
` R. HORST - 6/8/2018
`this exhibit is INTEL Exhibit 1003.001.
` Can you go to page with Bates number
`1003.112. On this page it discusses claim
`element 2.1. Do you see that, on top of the
`page?
` A. I must be on the wrong page. What
`page are you on?
` Q. It's Bates Number 112.
` A. Oh, 112.
` Yes, I see that.
` Q. And then the last paragraph on this
`page, starting with the word, "First." Do you
`see that?
` A. Yes.
` Q. In the middle of this paragraph on
`the left side, there's a sentence starting with,
`"Including this hardware in Erickson, as
`Tanenbaum teaches below, would perform the bypass
`test by, for example, checking the sequence
`number as part of its header prediction, checking
`the connection record, and classifying the packet
`according to fast or slow path." Do you see
`that?
` A. Yes.
`
`TSG Reporting - Worldwide 877-702-9580
`
`
`
`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
`25
`
` R. HORST - 6/8/2018
` Q. Why checking sequence number is part
`of header prediction?
` MR. CONSTANT: Objection. Form.
`Outside of the scope.
` A. To answer that, I'd really have to
`see that section of Tanenbaum or Stevens2 that
`gives the details on how the sequence number is
`used in header prediction.
` Q. Sitting here without looking at
`Tanenbaum or Stevenson, you don't know how
`sequence number is used as -- as part of header
`prediction?
` MR. CONSTANT: Objection. Form.
`Outside of the scope.
` A. In general, it's used to make sure
`that the receive packets are in order. And if --
`if they aren't received in order, then the
`fast-path check fails and is handled by the
`slow-path.
` Q. So what if it passed this bypass
`test, would the adapter remove the TCP header and
`IP header and pass the data into the host?
` MR. CONSTANT: Objection. Form.
`Outside of the scope. Incomplete hypothetical.
`
`TSG Reporting - Worldwide 877-702-9580
`
`
`
`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
`25
`
` R. HORST - 6/8/2018
` A. In the combination of Erickson and
`Tanenbaum, when the test is passed, it can use
`the fast mechanism for the fast application and
`do the transfer into memory directly without any
`intervention from the host software stack.
` Q. If fast-path mechanism is used, would
`the adapter remove TCP header?
` MR. CONSTANT: Objection. Form.
`Incomplete hypothetical.
` A. I'm not exactly sure what you mean by
`"remove." It would process the header and only
`transfer the data into user space.
` Q. What if the bypass test fails, would
`the adapter process the TCP header --
` MR. CONSTANT: Objection. Form.
` Q. -- or would the host process the TCP
`header?
` MR. CONSTANT: Objection. Form.
`Outside the scope. Incomplete hypothetical.
` A. In general, if the bypass path fails,
`the host would process that packet instead.
` MR. LI: Can we take a short break?
` MR. CONSTANT: Sure.
` (A break was taken from 10:18 a.m. to
`
`TSG Reporting - Worldwide 877-702-9580
`
`
`
`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
`25
`
` R. HORST - 6/8/2018
` 10:25 a.m.)
` MR. LI: Back on the record.
` Q. (BY MR. LI) Are you familiar with
`the concept interrupt?
` MR. CONSTANT: Objection. Form.
`Lacks foundation.
` A. Yes.
` Q. What does interrupt mean?
` MR. CONSTANT: Objection. Form.
` A. Are you referring to one of my
`reports?
` Q. Did you ever mention the term
`"interrupt" in any of your reports?
` MR. CONSTANT: Objection. Form.
` A. Yes, "interrupt" is at least in the
`'241 report.
` Q. And what does that mean in the
`context of that report?
` MR. CONSTANT: Objection. Form.
` A. I -- in my original '241 report, in
`it I have a pretty long discussion of interrupts.
`And I would refer to that.
` Q. Sitting here without looking at
`original report, do you have any general
`
`TSG Reporting - Worldwide 877-702-9580
`
`
`
`Page 31
`
`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
`
` R. HORST - 6/8/2018
`understanding of what "interrupt" means?
` MR. CONSTANT: Objection. Form.
` A. I have a general understanding, but I
`can't give you a precise definition that meets
`the claim limitations of "interrupt."
` Q. Can you tell me what your general
`understanding is?
` MR. CONSTANT: Objection. Form.
`Outside the scope.
` A. Again, in my original report I
`describe what interrupts are. I can refer you to
`that.
` Q. But you just told me you don't --
`without looking at the report, you have a general
`understanding. So I'm just asking what is that
`general understanding of "interrupt"?
` MR. CONSTANT: Objection. Form.
` A. I don't recall if it's part of the
`claim construction or not. So I don't know what
`you want in terms of a general understanding of
`"interrupt."
` Q. So your general understanding should
`not depend on the claim construction. Like
`before this case, do you understand the term
`
`TSG Reporting - Worldwide 877-702-9580
`
`
`
`Page 32
`
`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
`
` R. HORST - 6/8/2018
`"interrupt"?
` MR. CONSTANT: Objection. Form.
` A. Yes.
` Q. What does that mean?
` MR. CONSTA