`Trials@uspto.gov
`571-272-7822
`
`
`
`
`
`Paper # 25
`Entered: August 7, 2023
`
`
`
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`__________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`_______
`
`NETFLIX, INC.,
`Petitioner
`v.
`CA, INC.,
`Patent Owner.
`__________
`
`IPR2022-00322
`Patent 8,656,419 B2
`__________
`
`Record of Oral Hearing
`Held: June 2, 2023
`__________
`
`Before JEAN R. HOMERE, THOMAS L. GIANNETTI, and
`NATHAN A. ENGELS, Administrative Patent Judges.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`IPR2022-00322
`Patent 8,656,419 B2
`
`
`
`APPEARANCES:
`
`ON BEHALF OF THE PETITIONER:
`
`HARPER BATTS, ESQ.
`CHRIS PONDER, ESQ.
`Sheppard, Mullin, Richter & Hampton LLP
`1540 El Camino Real, Suite 120
`Menlo Park, California 94025
`
`(650) 815-2673 (Batts)
`(650) 815-2676 (Ponder)
`hbatts@sheppardmullin.com
`cponder@sheppardmullin.com
`
`
`
`ON BEHALF OF THE PATENT OWNER:
`
`DANIEL S. YOUNG, ESQ.
`Quarles & Brady LLP
`8210 Southpark Terrace
`Littleton, Colorado 80120
`(303) 268-0077
`dyoung@quarles.com
`CHAD E. KING, ESQ.
`King IAM, LLC
`PO Box 630917
`Lone Tree, Colorado 80163
`(303) 482-1528
`chad@king-iam.com
`
`
`
`
`The above-entitled matter came on for hearing Monday,
`
`June 2, 2023, commencing at 10:00 a.m. EDT, via Video-conference.
`
`
`
`
`
`
`
`2
`
`
`
`
`IPR2022-00322
`Patent 8,656,419 B2
`
`
`P-R-O-C-E-E-D-I-N-G-S
`
`
`
`10:00 a.m.
`JUDGE HOMERE: Welcome to the Board. This is an oral argument
`in IPR2022-00322. The challenged patent is US Patent 8,656,419 B.
`Petitioner is Netflix, Inc. Patent Owner is CA, Inc.
`
`I am Judge Homere. Judge Giannetti and Judge Engels are also here
`with us.
`
`At this time we'd like Counsel to introduce yourselves, starting with
`Petitioner.
`
`MR. BATTS: Yes, Your Honor. Good morning. This is Harper Batts
`from Sheppard Mullin, on behalf of Petitioner. With me here today is Chris
`Ponder, also Sheppard Mullin.
`
`MR. YOUNG: Good morning, Your Honor. Dan Young from
`Quarles & Brady, representing CA, Inc. With me is Chad King from King
`Intellectual Asset Management.
`
`JUDGE HOMERE: Thank you.
`
`Before we begin, we remind the parties the guidance for this hearing
`was provided in an oral hearing order on May 2, 2023, Paper 21. As you
`know from the May 2nd oral hearing order, each side has been given up to
`60 minutes total time for oral argument.
`
`Petitioner will proceed first to present its case as to the challenged
`claims of the challenged patent. Petitioner may reserve time for rebuttal.
`Thereafter, Patent Owner will argue its position in this case. Patent Owner
`may reserve time for sur-rebuttal.
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`
`
`
`
`
`3
`
`
`
`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
`
`26
`
`
`IPR2022-00322
`Patent 8,656,419 B2
`
`Absent certain circumstances, Petitioner may reserve no more than
`
`half its time for rebuttal. Patent Owner may reserve no more than half its
`time for sur-rebuttal.
`
`We have a few reminders. First, our primary concern is your right to
`be heard. If at any time during this proceeding you encounter technical
`difficulties that you feel fundamentally undermine your ability to adequately
`represent your client, please let us know immediately. For example, by
`contacting the team members who provided you with connection
`information.
`
`Second, when you're not speaking, please mute yourself. Third,
`please identify yourself each time you speak. This helps the court reporter
`put together an accurate transcript.
`
`Fourth, we have the entire record including demonstratives. When
`you refer to demonstratives, please do so by paper number or exhibits. And
`specifically, please identify the slide or page number that you're referring to.
`
`Also, please pause a few seconds after identifying it to provide us time
`to find it. This helps with the preparation of an accurate transcript. And
`finally, be aware that members of the public may be listening.
`
`After the hearing, please stand by for the court reporter in case he has
`any questions regarding spelling. A full transcript of the entire hearing will
`become part of the record.
`
`Any questions before we begin?
`
`MR. BATTS: No, Your Honor.
`
`MR. YOUNG: No, Your Honor.
`
`JUDGE HOMERE: Okay. Any time you are ready, Counsel for
`Petitioner, you may proceed.
`
`
`
`
`
`
`4
`
`
`
`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
`
`26
`
`
`IPR2022-00322
`Patent 8,656,419 B2
`
`MR. BATTS: Thank you, Your Honor. I guess I'll plan on 40
`
`minutes for my opening time.
`
`So I'd like to start with Petitioner's slide 4. This is an IPR regarding --
`essentially, the grounds are about this Verbeke reference and a few
`combinations to the Verbeke reference.
`
`We've identified on slide 4 what we believe are the four remaining
`issues in dispute, starting with the question of whether the claims require this
`direct communication or indirect communication. It's actually an argument
`that permeates many of Patent Owner's arguments, but we've started with it
`as a separate argument as Patent Owner has outlined it in its POR and
`briefing.
`
`If we go to slide 5, as the panel may recall in the proffer, Patent
`Owner raised the same argument about whether Verbeke disclosed direct
`instructions between the job submitter and the nodes, and whether the claims
`require direct instructions or direct communications or indirect
`communications are permitted under the claims or part of the claims.
`
`And if we go to slide 6, the institution decision rejected Patent
`Owner's argument that the claims require direct communication, noting that
`there's nothing in the claim language that would have that sort of narrowing
`effect and that the claim language appears to encompass both direct and
`indirect communications.
`
`Going to slide 7, you can see that Patent Owner is essentially
`rehashing and re-raising that same argument in its POR post-institution,
`again arguing that you can't have this intermediated communication or
`indirect communication, that the challenged claims require direct
`communication.
`
`
`
`
`
`
`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
`
`25
`
`26
`
`
`IPR2022-00322
`Patent 8,656,419 B2
`
`So if we turn to slide 8, Patent Owner, in some of its arguments in the
`
`POR, you'll see that they've basically turned claim analysis on its head.
`They're essentially arguing that unless the claim explicitly recites a specific
`subset, it must necessarily be narrower.
`
`For example, not just for direct or indirect communication, but
`another example I could point out in this claim is the claimant refers to an
`interface. It doesn't say whether that interface is a high-speed interface or a
`low-speed interface. The claims are silent as to the type of interface. So this
`is a basic principle of claim construction that you don't narrow a claim when
`the claim doesn't have this narrowing language within it to start with.
`
`If we turn to slide 9, what we see is that Patent Owner is taking
`somewhat inconsistent positions. They take a lot of inconsistent positions
`I'll walk through today, but on this issue, at one point they're claiming that
`the claims do not require direct communication with all nodes. But they still
`contend that the claims do not contemplate indirect communications.
`
`In going back to the claim language, as noted in the institution
`decision if we look at slide 10, what you see really is Patent Owner simply
`attempting to rewrite the claims, to narrow the claims to require direct
`communication when the claims simply don't have that requirement as
`written.
`
`Beyond the claim language itself, the parties have argued about Figure
`1 of the 419 specification. That's shown on slide 11. As you can see on
`Figure 1, you have the nodes that are labeled from 46a through 46n. And
`then you have on the left-hand side there the communication network 12.
`
`You can see that the figure as drawn shows that the communications,
`the transmissions go from the first node, 46a, through the communication
`
`
`
`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
`
`
`IPR2022-00322
`Patent 8,656,419 B2
`
`network to the next node. And I don't think there's any dispute between the
`parties that that is how that figure is showing how the nodes communicate.
`Column 2 discussion of Figure 1 makes clear that the transmissions occur
`over the communication network.
`
`At the bottom of this slide, you see that our expert provided testimony
`regarding that there's no indication on this figure that they communicate
`directly with each other. They have to communicate through this
`communication network.
`
`On slide 12, we asked Patent Owner's expert about this figure and
`about what the figure shows. I'd like to walk through these answers here
`because I think they're quite telling.
`
`His first and question and answer, we asked him when there's a
`message that's going from node 20a to node 20n through the communication
`network, it may pass through other nodes in the communication network
`such as a switch or a router.
`
`He answered yes, the communication network is. So his first answer
`admits that communication network 12 has this complicated internal
`structure and that a message going from node 20a to 20n would pass through
`other nodes. That's indirect communication.
`
`In the second set of questions that we have at the bottom of this slide,
`we asked him what is his opinion about whether the nodes must be directly
`connected to every other node in the plurality of nodes. He admits that the
`claims are silent and do not require direct communication. So we have not
`only our own expert but Patent Owner's expert agreeing with our position
`about what Figure 1 shows.
`
`
`
`
`
`
`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
`
`26
`
`
`IPR2022-00322
`Patent 8,656,419 B2
`
`And then I guess if there's any questions on it, we also included on
`
`slide 13 additional testimony from our expert in both of his declarations as to
`why the claims do not require a direct connection.
`
`And then also, on slide 14 we have further testimony from Patent
`Owner's expert regarding whether claim 1 precludes having an intermediary
`between the first node and then the other nodes that are being communicated
`with. He said there's nothing that precludes this.
`
`Now, the sur-reply asserts that the claim wasn't about the scope of
`claim 1; it was about a hypothetical situation. But I'd ask the judges on the
`panel to look at the question and the answer because I just don't think that's
`correct. It's clear that we are asking about claim 1 specifically, and his
`answer repeatedly refers to claim 1 and the claim. So I just don't think it's a
`meritful position to say that he was not discussing claim 1 in this question
`and answer.
`
`And then furthermore, on slide 15 we've laid out from the
`specification -- they had an original argument about whether there was direct
`communications. They pointed to column 4 and this example in column 4.
`
`We explained in our reply that this example was talking about
`different operations, not the same operation. So it talks about operations X,
`Y, and Z. I don't think Patent Owner is disputing that anymore, that that
`example does not support their position for direct versus indirect
`communication.
`
`And then lastly, on slide 16, the other example from the specification
`that I believe was pointed to is simply a passive voice example. It just says
`one or more of the nodes that received instruction. It doesn't say directly
`receive. It just says they receive an instruction.
`
`
`
`8
`
`
`
`
`
`
`IPR2022-00322
`Patent 8,656,419 B2
`
`And so if we go to slide 18, I guess the remaining issue was that
`
`Patent Owner alleges in their sur-reply that we've changed our position from
`the petition, that originally we said the claims required anonymity, and that
`now we have a new theory that does not require anonymity but permits
`anonymity.
`
`If you look at the petition, I think it's a pretty straight-forward analysis
`here. We've laid it out and done the underlining on this slide. We're simply
`saying that it's an elementary aspect of the patent law. A claim can be broad
`enough to cover something, but doesn't necessarily require it. So we said
`that the claim was inclusive of communications through interposed nodes
`and would include within their scope. We didn't say that it had to acquire it.
`
`And then I'd like to touch on last for this issue the infringement
`contentions because I know Judge Engels and Judge Giannetti have been in
`a few different of these proceedings now with Broadcom and Netflix. And
`yet again, their positions are completely at odds with their infringement
`contentions. It really puts the nail in the coffin for this issue.
`
`And so we have the figures from the contentions themselves that they
`marked up. This is their highlighting. And then even more telling is, if we
`go to slide 20, their expert's testimony about the contentions.
`
`We asked their expert specifically about what is being shown in the
`contentions about this Titus Master. He agreed with us that what the
`contentions are contending as infringing is a Titus Master that receives jobs,
`jobs are then submitted to this master, and then the master schedules out the
`tasks to various agents. He said, yes, I see that.
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`
`
`
`
`
`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
`
`26
`
`
`IPR2022-00322
`Patent 8,656,419 B2
`
`And then we asked him, the jobs aren't submitted directly to the
`
`agents, are they? And he said, based on following what I see here, that's
`correct.
`
`And I think if we turn -- I'd actually like to go to that exhibit. That's
`Exhibit 1019. If we go to slide -- because I see it in Patent Owner's slides
`for this hearing. They have a lot of slides that start out about this. Verbeke
`is a completely different architecture, it's a hierarchy architecture, and our
`architecture for the 419 is a flat architecture, a completely different
`architecture. That seems to be a thematic issue that they start a lot of their
`slides with.
`
`If we go to Exhibit 1019 and page 7, this is infringement contentions
`on page 7. You can see for yourself what I'm referring to and what the
`expert was answering questions about. What they're accusing is a system
`where you have a master, similar to Verbeke's task dispatcher. You have
`this Titus Master, which is receiving these job requests and then scheduling
`these tasks or job requests. And it's sending them to particular agents to
`have the available resources. So this is yet again showing an indirect
`communication allegation.
`
`According to Broadcom in District Court, the claims can cover
`indirect communications. And similarly on page 20, at the top of page 20,
`you again see where what they're pointing to is where job specifications are
`submitted to a master. They consist of tasks that are then sent. The master
`schedules tasks onto Titus agents that launch the containers based on the
`task job specification.
`
`I don't know that there could be a clearer example of the Patent Owner
`taking contradictory positions here. They've had no answer. We raised this
`
`
`
`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
`
`
`IPR2022-00322
`Patent 8,656,419 B2
`
`in our briefing, in the initial petition. We raised this again in the reply.
`They've had no answer to this other than to say, you haven't shown how
`Netflix's system actually works. You're just telling us here's some
`documents about how it might work.
`
`That's irrelevant to the question of whether Patent Owner is taking
`inconsistent positions about the claims scope. They're alleging that's how it
`works. They're alleging infringement based upon an indirect
`communication. And therefore, I think that really kills their argument. It's
`directly contradictory to their arguments here and they have no answer for it.
`
`Unless there's any other questions about that, I'd like to move to slide
`21 and address the next argument. Okay. The next argument is, I think, the
`second main argument that they have, which is this question of whether the
`job submitter is telling or instructing a plurality of worker nodes about a
`single task or a single operation.
`
`Slide 22, I think, lays out their original argument, which was they
`claimed that the claims require that there be the same communication to each
`of the plurality of nodes about a single task or single operation. That was in
`their POR and citations we've provided.
`
`If we go to slide 23, what we see is that in our deposition of Dr.
`Rosenblum, Patent Owner's expert, he admitted that the instructions or the
`telling limitations do not require that the same communication be sent to
`each of the nodes.
`
`And in slide 24, I believe they've now conceded this. It sounds like in
`their sur-reply they're saying it doesn't need to be the same communication,
`but the operation still needs to be the same. So that seems to be where we've
`
`
`
`
`
`
`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
`
`26
`
`
`IPR2022-00322
`Patent 8,656,419 B2
`
`boiled down to on this argument of what Patent Owner has left on its
`argument.
`
`If we turn to slide 25, Patent Owner's argument is that Verbeke never
`teaches multiple workers are told to perform the same operation. And let's
`keep in mind the claim language. I always have my claim slide up there so I
`can go back up to the beginning, slide 3. The language that we're talking
`about here is in 1b, which is this operation comprising a procedure of an
`application. So that is the language that's in dispute on this second issue.
`
`If we go back to slide -- let's see -- I'll go to slide 26. The parties
`spent quite a bit of their reply and sur-reply briefing on this question in the
`context of Figure 5 of Verbeke. We've laid out Figure 5 here. You've seen
`it, I'm sure, a few times in the briefing and the different markups.
`
`What we've shown here in Figure 5 is that you have the job. It's
`identified in the purple text. It's this Monte-Carlo X-ray generation. And
`then we identified in our petition that what is the operation comprising a
`procedure of an application are the X-ray generation in aluminum, the X-ray
`generation in copper, and then the X-ray generation in silicon that are
`identified as 1104a, b, and c.
`
`So if we go to slide 27, I think that's clearly laid out on the top of the
`slide. I don't think this is a question of is it clear what the Petitioner is
`contending is the operation and is there some switch in positions.
`
`We've consistently pointed out that the worker nodes perform as
`operation comprising a procedure of an application. And then we gave as
`examples, for example, the three tasks that I just mentioned. Patent Owner's
`expert did not dispute this mapping. That's what's shown on the bottom of
`slide 27.
`
`
`
`
`
`
`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
`
`26
`
`
`IPR2022-00322
`Patent 8,656,419 B2
`
`If we go to slide 28, what we see is that the task of -- let's say for the
`
`task 1104a here, the task of X-ray generation in aluminum, that operation is
`then performed below in the three yellow boxes that are shown below.
`That's explained by our expert, Dr. Jagadish. So that task is being performed
`on three separate sets of electrons.
`
`We have this tension, I think. What you may have seen in Patent
`Owner's briefing is the tension between the Patent Owner and its expert.
`
`If we go to slide 29, not only does our expert, Dr. Jagadish, say that
`that's showing the operation being performed across three nodes, but Patent
`Owner's expert agrees that Figure 5 in this 1104a example is an example of
`where a single task is being executed across more than one -- here three --
`worker nodes.
`
`JUDGE HOMERE: Counsel?
`
`MR. BATTS: Yes.
`
`JUDGE HOMERE: This is Judge Homere. I'm confused about how
`Verbeke is intended to operate here because it seems to me what's going on
`in Verbeke is a distributing system where, essentially, a job can be done.
`The job submitter submits it in the task manager. The task manager
`partitions it into tasks, and then assigns a different task to different workers
`so it can move parallel as a way to expedite the processing of the job.
`
`So how would that comport with the overall objective of Verbeke if it
`were to assign the same task to multiple working nodes?
`
`MR. BATTS: So the claim language requires comprising a procedure
`of an application. The procedure that we were identifying here, you have the
`X-ray generation in aluminum, and then it's being performed across three
`different nodes with the nodes performing it on three different sets of data.
`
`
`
`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
`
`26
`
`
`IPR2022-00322
`Patent 8,656,419 B2
`
`If you look at slide 30, that's consistent with both our expert's
`
`explanation but also Patent Owner's expert's explanation. We asked
`specifically on multiple occasions, do you believe that there's three tasks or
`that there's a single task for 1104a? You see his answer is, there's a single
`task that's being given to a plurality of workers.
`
`And then we asked him again, so the task that's under 1104a, is it
`being spread across three different workers? He said yes. He said, this task
`is being executed with Worker 1 and it's also being executed with Worker 2,
`and then it's also being executed with Worker 3.
`
`And then lastly, we asked him, is the code being used to perform that
`task of the three the same? He said yes.
`
`So the operation is a portion of an application that is different from
`the data on which the application operates. And I want to get to that point
`because I think that's where in sur-reply Patent Owner has now gone, given
`the testimony that's been given by both experts.
`
`JUDGE HOMERE: But Counsel -- this is Judge Homere -- I don't
`think you answered my question. I understand that you're pointing to Patent
`Owner's expert testimony here that presents his view regarding the teachings
`of Verbeke.
`
`The thing that I do not understand about this explanation is if the
`intent is to expedite processing of a job by providing the different workers
`so that the job can be done and processed in parallel, why would the same
`job be assigned to different workers? That would be counter-intuitive to the
`overarching intent of Verbeke. How do you explain that?
`
`MR. BATTS: Well, Verbeke is -- so I think this goes back to what is
`an operation. I think your question is asking what is an operation.
`
`
`
`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
`
`26
`
`
`IPR2022-00322
`Patent 8,656,419 B2
`
`I have to go to column 3. I'll start with this language. If we go to
`
`column 1 of the 419 patent, the 419 patent says an operation may be -- this is
`lines 14 to 16. It says the operation may be pre-compiled and both the
`clients remain for what the operation of the server is capable of performing.
`So an operation is a procedure of an application. Here it's making it clear it
`can be pre-compiled code.
`
`The 419 patent also explains -- for example, on column 3, it gives
`examples of operations. It says on column 3, line 6, operations may be any
`specific actions, tasks, procedures, and/or functions of an application. And
`then it gives some examples of those.
`
`So what the claim requires, if we're going back to the claim language
`of the 419 patent, is that you have an operation which is basically
`comprising a procedure of an application. It's the actual software. It's the
`calculations that are being performed.
`
`It has, for example, on column 3, the word processing example at line
`8 where it talks about a word processing application, to include operations
`for handling text such as copying, pasting, formatting, spell checking,
`printing, et cetera.
`
`So what has to be performed on the nodes is this application. This
`operation needs to be being performed. And what we have in Verbeke is
`Verbeke is saying you have this distributed computer framework where you
`have multiple nodes that are available to perform these operations.
`
`And we've identified operations, for example, on slide 28 where you
`have the operation in 1104a for X-ray generation in aluminum. Verbeke
`allows you the flexibility of then splitting off that operation. You have that
`operation being performed in four different sets.
`
`
`
`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
`
`
`IPR2022-00322
`Patent 8,656,419 B2
`
`So the operation is same. The data set it's being performed on is
`
`different. There's nothing in the claim about the term operation that requires
`the data set to be the same. That would be contrary to what I just explained
`from column 1 and column 3. You can't have a pre-compiled data set. That
`simply doesn't make sense.
`
`So Verbeke matches what the claim requires in having multiple nodes
`being told to perform an operation. And then we have both experts agreeing
`that that operation, that single operation is being performed by the nodes.
`
`JUDGE HOMERE: Do you draw a distinction between an operation
`and a task as referred to in Verbeke? It seems that column 1 refers to an
`operation as including a text. When you look at Figures 5 and 6, it's
`confusing as to how Verbeke actually operates.
`
`It looks like we have different embodiments here in Verbeke, but look
`at Figure 6. If you were to follow Figure 6, you have the job submitter
`submitting a job to the task dispatcher, and the task dispatcher partitions the
`job into a variety of tasks. The tasks themselves are assigned to different
`workers.
`
`Is that your understanding of Figure 6 in Verbeke?
`
`MR. BATTS: I want to be careful about the term task as used by
`Verbeke because we specifically said what we identified. And that goes
`back to slide 28.
`
`We identified that what we are contending is the operation comprising
`a procedure of an application, for example, for a, b, and c. Those are the
`tasks in Verbeke that we contend are the operation comprising procedure of
`an application.
`
`
`
`
`
`
`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
`
`
`IPR2022-00322
`Patent 8,656,419 B2
`
`Now, on Figure 6, you do see that you have the job submitter. And
`
`then you have the task dispatcher that's then showing a representation of
`how the task dispatcher makes a decision on which worker to send out a
`particular calculation to.
`
`(Simultaneous speaking.)
`
`JUDGE HOMERE: I understand that. First of all, you said earlier
`that there's a distinction between function and task. When you go back to
`column 1 of Verbeke, it does say that a function includes a task. So let's be
`sure that we have got the right definition for those terms here.
`
`So is that correct? Is a function the same thing as a task in Verbeke?
`
`MR. BATTS: I apologize, Your Honor. Can you tell me where
`you're referring to in, I believe, column 1 of Verbeke?
`
`JUDGE HOMERE: Yes. You pointed to it earlier, right, regarding
`task and function?
`
`MR. BATTS: No, Your Honor. I was pointing actually to the 419
`patent.
`
`JUDGE HOMERE: Okay.
`
`MR. BATTS: I was saying that the 419 patent, if we're going back to
`what the claim requires, the 419 patent column 1, lines 13 to 15, says the
`operation has to be pre-compiled. I'm saying that Patent Owner's argument
`seems to have morphed by the sur-reply to saying it can't be the same
`operation because it's being performed on different sets of electrons,
`different data.
`
`And my point being that that argument can't possibly be correct when
`looking at the 419 patent specification. The claim language itself is saying
`
`
`
`
`
`
`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
`
`
`IPR2022-00322
`Patent 8,656,419 B2
`
`an operation comprises a procedure of an application, which would not be
`the data set that you're performing it on.
`
`Not just the claim language itself but even column 1, column 3, and I
`think there's other locations in 419 that make it clear that what's being
`referred to for an operation is the software. It's the calculation that's being
`performed. It's not whether the same underlying data is being used or
`having that calculation performed on it. How would you have pre-compiled
`data sets?
`
`I'm sorry if I was unclear, Your Honor. Clearly the example on the
`419 patent in column 3, it's saying that operations may be any specific
`actions, tasks, procedures, and/or functions of the applications.
`
`So given the admissions by Patent Owner's expert on what Figure 5
`shows consistent with our expert, it is the same operation. And that
`operation is X-ray generation in aluminum. It's just being performed on
`different data sets, but there's nothing in the claim language. And it certainly
`would be inconsistent with the specification of the 419 to have to require
`that the data sets be the same for the operation to be the same.
`
`JUDGE HOMERE: Okay.
`
`MR. BATTS: And then I do think similar to this, if we go to -- I want
`to address this on a few slides. I think there's a lot of misrepresented
`citations in the sur-reply. If we go to slide 31, the Patent Owner says Dr.
`Rosenblum didn't agree with this.
`
`Their main argument is our expert didn't actually say what he said.
`Then they say, instead, Dr. Rosenblum correctly described that the task
`dispatcher will provide three distinct instructions to the workers. We put the
`
`
`
`
`
`
`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
`
`26
`
`
`IPR2022-00322
`Patent 8,656,419 B2
`
`exact testimony from that citation below. There is no reference to distinct
`instructions.
`
`And if you actually look at the testimony, it supports Petitioner's
`position that 1104a is the task and that task is performed by three workers.
`He says, for that task, yes. I believe for that task, the task dispatcher will tell
`the workers. Worker 1 will execute this task, then Worker 2 will execute
`this task, and Worker 3 will execute this task.
`
`JUDGE HOMERE: Okay. It seems to me what you are saying here is
`that the three workers are performing the same task but on different data
`sets.
`MR. BATTS: Exactly, Your Honor. And it seems like from their sur-
`
`reply, they're now contending you have to pull a data set into the term
`operation and say it has to be the same data set. Otherwise, it's not the same
`operation.
`
`That's why I tried to walk through the claim language. Clearly that
`makes no sense comprising a procedure and application, but it's also
`inconsistent with both experts' testimony. And it's also inconsistent with the
`citations I provided from the 419 specification about what an operation
`would be.
`
`And then if we go to slide 33, this is again where the Patent Owner is
`misrepresenting its own expert's testimony in the sur-reply because then they
`say, Dr. Rosenblum testified that each of those yellow boxes is a separate
`task.
`And if you look at the language that we bolded below, we did the
`
`citation. He was asked, so you would expect that there is supposed to be
`three tasks performed using the three different data sets?
`
`
`
`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
`
`
`IPR2022-00322
`Patent 8,656,419 B2
`
`And he said, I would say what's going on here is there is a single task,
`
`X-ray generation in alum



