[00:00:02] Speaker 04: Alright, our next case for argument is 24-1082, CA Inc. [00:00:07] Speaker 04: vs. Netflix. [00:00:30] Speaker 02: Good morning, Your Honors. [00:00:31] Speaker 02: May it please the court? [00:00:32] Speaker 02: Dan Young with Coral's and Brady representing CA Inc. [00:00:36] Speaker 02: The board's finding that the Verbecki reference invalidates the independent claims of the 491 patent lacks substantial evidence because there's no way to map the Verbecki reference in a manner that discloses all the claim elements. [00:00:48] Speaker 02: And those issues are shown in those mistakes and errors are shown in the board's decision. [00:00:54] Speaker 02: With respect to claim one, it's important to understand what the actual elements are. [00:00:58] Speaker 02: It has a first node. [00:01:00] Speaker 02: and that node tells a plurality of nodes to perform an operation. [00:01:05] Speaker 02: That operation is a procedure of an application, so it's a subpart of an application. [00:01:11] Speaker 02: Instruct the plurality of nodes how to perform the operation using computer code. [00:01:16] Speaker 02: Tell the plurality of nodes what to do with the results. [00:01:20] Speaker 02: And do all of this wherein the first node does not know which one of the plurality of nodes will perform the operation. [00:01:31] Speaker 02: an instruction going to a plurality of nodes about a single operation, and the first node does not know which one of the plurality will perform the operation. [00:01:40] Speaker 02: This claim reflects the decentralized peer-to-peer system that's reflected in claim one. [00:01:46] Speaker 02: Decentralized meaning like it's in figure one of the 419 patent. [00:01:51] Speaker 02: You have a number of nodes in that particular embodiment. [00:01:53] Speaker 02: They have an evaluator, has logic. [00:01:57] Speaker 02: amongst the nodes themselves, it decides which one is going to perform the operation. [00:02:03] Speaker 04: Council, not to cut you off, but I don't think we need the technical tutorial. [00:02:06] Speaker 04: We've all read the patent. [00:02:07] Speaker 04: And so why don't you get to how you think the district court erred with regard to the claim limitations. [00:02:13] Speaker 02: Yes, Your Honor. [00:02:13] Speaker 02: So with respect to limitation 1B, the limitation requires the first node to tell a plurality of nodes to perform an operation. [00:02:21] Speaker 02: In the Verbecki references, undisputed that the job submitter [00:02:26] Speaker 02: does not have any communications with the working OVC. [00:02:28] Speaker 01: So just to come down to it in English, your dispute with respect to what the board concluded is the board said you claim that the limitations require it communicate directly. [00:02:42] Speaker 01: And in the prior art reference, there is this dispatcher thing. [00:02:46] Speaker 01: Is that what the rub is? [00:02:47] Speaker 01: Whether the board correctly construed it is not necessarily being directly. [00:02:52] Speaker 01: And you could have this dispatcher, and it could still be this rub. [00:02:55] Speaker 02: That is part of it, Your Honor. [00:02:56] Speaker 02: The board made three specific decisions or findings with respect to limitation 1B. [00:03:01] Speaker 02: The first one is exactly what you said, which is the board says there's no requirement of direct communication from the first note. [00:03:09] Speaker 01: And so how did the board hear in that regard? [00:03:11] Speaker 02: Well, the claim language requires that there is a telling from the first node to a plurality of nodes. [00:03:18] Speaker 02: In the Rebekah reference, it's undisputed that the job submitter, which they map as the first node, has no communications with the worker nodes, which they define as the plurality of nodes. [00:03:30] Speaker 02: And it's very important that they did not map the task dispatcher, which is the center of the hub of Rebekah, to any of the claim elements. [00:03:39] Speaker 04: But the direct versus indirect point, there's nothing in the claim that says it has to be direct point-to-point communication with no intermediary, right? [00:03:48] Speaker 02: Yes, Your Honor. [00:03:49] Speaker 04: And then the problem with that is reflected in the next two findings of the board, which is... But didn't you all argue when this case was before the district court that indirect communication satisfied this exact claim limitation? [00:04:01] Speaker 04: Didn't you all already argue this point at the district court and you took the opposite side? [00:04:06] Speaker 02: No, Your Honor. [00:04:08] Speaker 02: They point out certain infringement contentions that have an accused functionality of the accused system. [00:04:18] Speaker 02: None of the actual underlying functionality of that system is before the board in this case. [00:04:25] Speaker 04: On page A22 of this opinion that's in front of us at footnote three, this opinion notes that you argued [00:04:37] Speaker 04: in your infringement contentions that indirect communication, and this is a quote, indirect communication between the first note highlighted in blue and the second note, Titus agent, meets the claim limitation. [00:04:49] Speaker 04: So you actually use the words in your district court proceeding that indirect communication was sufficient to meet infringement of this exact claim note limitation. [00:05:02] Speaker 02: So what do we do with that? [00:05:08] Speaker 04: You can't have it one way for validity and a different way for infringement. [00:05:11] Speaker 02: How do you think? [00:05:12] Speaker 02: But the claim language requires that the first node tell the plurality nodes perform an operation. [00:05:18] Speaker 02: So whether, now if you want to say indirect communication means that it goes through a router or another server, that communication has to go from the first node to the plurality of nodes. [00:05:29] Speaker 02: And in this situation with the Rebecca reference, that clearly doesn't happen. [00:05:34] Speaker 02: all the communications from the first node go to the task dispatcher. [00:05:38] Speaker 02: The task dispatcher then takes what was a job, which the board, I'm sorry, which Netflix mapped to, that's a job, to an application, which is not an operation, an application. [00:05:50] Speaker 02: The task dispatcher then takes that [00:05:53] Speaker 02: communication from the first node, separates it out and says to individual nodes, you do this one, you do this task, you do this task, you do this task, you do this task. [00:06:03] Speaker 02: So in the Verbecki reference, there is no communication, whether it's direct or indirect, that go from the first node to the actual plurality of nodes, which is what... Well, what about figure six in Verbecki? [00:06:18] Speaker 01: Yes, Your Honor. [00:06:20] Speaker 01: The first node goes to the task dispatcher, and then there's interaction between the task dispatcher and the other nodes. [00:06:29] Speaker 01: I can barely read these because the boards are so small, but am I missing something about this figure? [00:06:33] Speaker 02: Yes, Your Honor. [00:06:34] Speaker 02: What you're missing is that the job submitter is submitting a job. [00:06:39] Speaker 02: That is the application. [00:06:41] Speaker 02: It says run this application. [00:06:42] Speaker 02: What the task dispatcher does then is take that job [00:06:46] Speaker 02: break it up into tasks, and then dispatch those tasks to the various notes. [00:06:50] Speaker 02: So the communication that the job submitter, the first note, is giving to the task dispatcher is not the communication that the task dispatcher then sends to the individual notes. [00:06:59] Speaker 02: And if you see figure 14, which is another representation of what the task dispatcher does. [00:07:04] Speaker 04: I'm having a little trouble understanding. [00:07:05] Speaker 04: I think that maybe you've pivoted [00:07:07] Speaker 04: you're no longer disputing indirect communication. [00:07:12] Speaker 04: Am I right? [00:07:14] Speaker 04: Have you now pivoted? [00:07:15] Speaker 04: Have you now given in that even if it occurs by indirect communication, in this case, the accused reference doesn't disclose it. [00:07:25] Speaker 04: So have you now accepted that indirect communication is sufficient? [00:07:32] Speaker 02: Your Honor, if it's indirect communication of the operation, [00:07:38] Speaker 02: In other words, the individual operation. [00:07:43] Speaker 02: The whole crux of this patent is that a single operation has to be instructed to a plurality of nodes. [00:07:50] Speaker 02: And then one of those nodes has to perform the operation. [00:07:54] Speaker 02: And if you look at the other board's findings, this is where this ends up falling apart. [00:07:59] Speaker 02: It says, as an example, the board used figure five of Rebecca to show, oh, no, there's evidence here that the [00:08:08] Speaker 02: task dispatcher sends the same operation to multiple nodes. [00:08:13] Speaker 02: And it said exactly on Appendix 23, Figure 5, Verbecky depicts the task dispatcher, upon receiving a job from the job submitter, dispatches partitions of the job to a polarity of worker nodes as a polarity of task. [00:08:29] Speaker 04: So is your problem now that when the job submitter communicates [00:08:34] Speaker 04: indirectly to the worker node through the task dispatcher, is your problem that the operation is being broken up and distributed in the prior art to a number of different nodes, which are each performing a part of the operation, but not any one of the plurality of nodes is performing the entire operation? [00:08:54] Speaker 04: Is that your argument? [00:08:56] Speaker 02: Yes, Your Honor. [00:08:56] Speaker 04: So that's not, then you're not having a problem with indirect communication. [00:08:59] Speaker 04: So even though I understood the bulk of your brief to go to this indirect communication, that isn't, it seems to me that you seem to, if the job submitter sent the entire task to the task dispatcher and then the task dispatcher sent that entire task to a single worker node and it was all performed like that, you're telling me that would be okay. [00:09:20] Speaker 02: So your honor, in your example, if the communication was like through a conduit and it says the task dispatcher is just a conduit. [00:09:28] Speaker 04: Yeah, but the task dispatcher chooses, for example, which of the plurality of notes has free time to perform this function, this operation. [00:09:34] Speaker 04: So the task dispatcher chooses which note to give it to. [00:09:37] Speaker 04: You're saying that would meet the claim limitations. [00:09:41] Speaker 02: No, Your Honor, because the claim limitations require, this is very important, this is the crux of this, that when the first node has to tell a plurality of nodes to perform without knowing which one is going to perform the operation, which one. [00:09:56] Speaker 02: Right, it doesn't care. [00:09:57] Speaker 04: It doesn't care which one. [00:09:59] Speaker 04: It doesn't care which one. [00:09:59] Speaker 02: It just wants the operation to get done. [00:10:01] Speaker 02: So in the conduit example, [00:10:03] Speaker 02: The task dispatcher does not just send the request through the job submitter to the plurality of nodes. [00:10:09] Speaker 02: It doesn't do that. [00:10:10] Speaker 02: It says the job submitter submits an application, the operations are part of that, to the task dispatcher. [00:10:18] Speaker 02: The task dispatcher then says for each task, one individual node is going to perform that operation. [00:10:27] Speaker 02: It's not, the task dispatcher is not saying you three nodes perform it and then you nodes decide amongst yourselves which one is going to perform. [00:10:35] Speaker 02: This is the problem. [00:10:36] Speaker 01: If you use the task dispatcher, the reason they don't map- Are you referring to any- Are you referring to figure five? [00:10:45] Speaker 01: Are we talking about figure five here? [00:10:47] Speaker 02: Figure five is the only evidence that the board relied on to buttress their finding that [00:10:55] Speaker 02: the task dispatcher assigns a single operation to multiple nodes. [00:11:00] Speaker 02: And that's clearly not what's going on there. [00:11:03] Speaker 02: Figure five actually shows in the annotated yellow boxes, each one of those is a task. [00:11:09] Speaker 02: But even if you were to say that's all one task and each node is doing the task, then you're not satisfying element one E because it says [00:11:20] Speaker 02: to know which one of the tasks are going to perform the operation. [00:11:25] Speaker 02: This is the difference between the centralized system of Rebecki and the decentralized system of the claimed invention. [00:11:32] Speaker 02: The claimed invention allows the workers in Rebecki don't do exactly what they're told. [00:11:37] Speaker 04: By the way, you said the only thing that the board relied on was Figure 5 from Rebecki. [00:11:41] Speaker 04: But didn't they also rely on expert testimony? [00:11:44] Speaker 04: I seem to remember them expressly citing expert testimony as well. [00:11:48] Speaker 02: They relied on expert testimony with respect to figure five issue. [00:11:51] Speaker 04: Right. [00:11:51] Speaker 04: OK. [00:11:51] Speaker 04: So it wasn't just the figure itself. [00:11:53] Speaker 04: It's not like I, Kimberly Moore, can look at that figure the same way the board did. [00:11:57] Speaker 04: They looked at it through the lens of expert testimony that they found credible. [00:12:01] Speaker 04: Correct? [00:12:03] Speaker 02: They absolutely relied on expert testimony. [00:12:06] Speaker 02: But again, the expert testimony can't contradict what was clearly on the face of the figure. [00:12:11] Speaker 02: Figure five clearly shows that you have individual tasks in this particular case [00:12:17] Speaker 02: task electrons 1 through 1,200, 401 through 600, 201 through 1,400. [00:12:23] Speaker 02: Those are each a task given to a single node. [00:12:27] Speaker 02: And that, again, you have to have the instruction needs to be to a plurality of nodes where only one node performs that operation. [00:12:34] Speaker 00: That sounds like a claim construction argument to me, what you're saying, because the expert testimony was [00:12:40] Speaker 00: You look at this limitation, tele-clarality of nodes, and that a person of skill in the art would understand that what for Becky does is telling the plurality of nodes. [00:12:49] Speaker 00: So are you making a claim construction argument? [00:12:51] Speaker 02: I know, Your Honor. [00:12:52] Speaker 02: The claims are plain on their face. [00:12:55] Speaker 02: If you take that description, which you just said, with respect to element 1b, then you have not satisfied element 1b, which says one node has to perform the operation. [00:13:09] Speaker 02: And I'm kind of running out of time, but just very quickly, the findings just don't go with respect to limitation 1b. [00:13:15] Speaker 02: 1c, it says that you have to identify with computer code how to perform the job. [00:13:23] Speaker 02: Again, there's no instructions from the first node to the worker nodes how to perform the operation using computer code. [00:13:29] Speaker 02: What happens is the first node will put an identifier saying this is the type of operation it is. [00:13:36] Speaker 02: The task dispatcher will figure out from the repository whether or not the code exists. [00:13:40] Speaker 02: If the code exists, it will then send it to the worker nodes to perform the operation. [00:13:45] Speaker 02: But again, there's no direction from the job submitter to a plurality of worker nodes to perform the operation using code. [00:13:53] Speaker 02: And then similarly, there's no discussion in limitation 1D as to what to do with the results. [00:13:58] Speaker 02: All that happens is the job submitter submits the job. [00:14:02] Speaker 02: The task dispatcher assigns the job. [00:14:04] Speaker 02: The worker notes then return the results to the task dispatcher. [00:14:09] Speaker 02: And that's clear in many of the figures, including step 1526 of figure 9 as an example, 2026 of figure 14. [00:14:19] Speaker 02: And then finally, with respect to the dependent claims, [00:14:23] Speaker 02: The dependent claims 7, 8, and 14 require that the code requires the code to remove itself upon completion of the process. [00:14:33] Speaker 02: There's no disclosure whatsoever in Rebecca that removal occurs from the worker nodes. [00:14:39] Speaker 02: The worker nodes are where that code resides. [00:14:41] Speaker 02: That's where it has to execute. [00:14:43] Speaker 02: It can't execute in a repository. [00:14:45] Speaker 02: And all Verbecky says is that once the job is completed, the tasks are then removed from the repository. [00:14:50] Speaker 02: There's no discussion that the nodes are what actually remove the code upon completion. [00:14:56] Speaker 02: And even further, with respect to claim eight, it says the first node must instruct the second node to remove the code upon completion. [00:15:04] Speaker 02: That second node is part of the plurality nodes. [00:15:07] Speaker 02: And obviously, there's no interaction between individual working nodes. [00:15:11] Speaker 02: with respect to the code. [00:15:12] Speaker 00: That also sounds like a claim construction argument to me, that the code has to be removed from the worker nodes. [00:15:20] Speaker 02: Well, the claim specifically says that it has to be removed upon completion. [00:15:24] Speaker 02: The code, the entity basis for the code refers to the first claim where it's telling the worker nodes what to do with the codes. [00:15:32] Speaker 02: The plain language of the claim clearly shows [00:15:36] Speaker 02: where the code is supposed to be when executed and then again, that would have to be removed. [00:15:42] Speaker 04: Okay, thank you, Mr. Young. [00:15:44] Speaker 04: Let's hear from the closing counsel. [00:15:48] Speaker 04: Mr. Batts. [00:16:03] Speaker 03: Thank you, Your Honors. [00:16:04] Speaker 03: Good morning. [00:16:04] Speaker 03: Harper Batts on behalf of Pelley Netflix. [00:16:07] Speaker 03: I think I'd like to start off just initially briefly on the indirect communication. [00:16:13] Speaker 03: I think it was pretty clear. [00:16:16] Speaker 03: Judge Moore, you asked whether there was anything in the claim that required direct communication. [00:16:20] Speaker 03: You said no. [00:16:21] Speaker 03: I think that forecloses it. [00:16:23] Speaker 03: That's consistent with the admission that we pointed to from the hearing below, where the judge specifically asked them about indirect communication. [00:16:30] Speaker 03: And there was admission that there was indirect communication. [00:16:33] Speaker 03: So what I do want to turn to is I want to address the plane limitations 1B and 1E, and specifically this part. [00:16:40] Speaker 04: Well, I'm not sure. [00:16:41] Speaker 04: I found it a little confusing, but I think that I understood him to be saying even if indirect communication is satisfied, it's not the operation because it's being split off and performed by different nodes. [00:16:52] Speaker 04: So I'm not sure I understand whether that is distinct [00:16:56] Speaker 04: from the direct and indirect communication argument or part of it, or if it's part of the operation argument. [00:17:01] Speaker 04: But can you address that part? [00:17:02] Speaker 04: Wherever it falls in the claims, can you address it? [00:17:06] Speaker 03: Yes. [00:17:06] Speaker 03: So I think what I look at is the claim language that they were disputing below. [00:17:10] Speaker 03: I guess what they have as their second argument in their brief was to tell a plurality of nodes to perform an operation. [00:17:18] Speaker 03: And so on that issue, they were arguing that Figure 5 does not show that. [00:17:23] Speaker 03: So the board credited multiple different portions of Rebecca for this issue beyond Figure 5. [00:17:28] Speaker 03: But I do want to address Figure 5 because they went into detail on that. [00:17:31] Speaker 04: It's tell a plurality of nodes to perform an operation. [00:17:34] Speaker 04: Instruct them to perform the operation. [00:17:37] Speaker 04: Tell the plurality of nodes what to do with the result of the operation. [00:17:43] Speaker 04: I guess the question, and I'm not positive I understood this to be the argument, but I think the question is, do you just have an intermediary where it's picking a node and sending it and the whole operation is being done by that intermediary, or is the operation being broken up into parts and distributed among multiple nodes? [00:18:01] Speaker 03: So yeah, so I'd like to walk through the mapping of what we did for Rebecca to answer your question. [00:18:06] Speaker 03: OK. [00:18:06] Speaker 03: So the mapping of Rebeki. [00:18:07] Speaker 03: And we could start with figure. [00:18:09] Speaker 04: What page? [00:18:09] Speaker 03: Oh, I'll start with appendix page 83, just briefly back to the petition, just the starting point. [00:18:15] Speaker 03: And this is limitation 1b. [00:18:17] Speaker 03: And at the bottom there, we explain how Rebeki renders obvious 18b. [00:18:21] Speaker 03: And we say, each operation of the plurality of operations comprising a procedure of the application. [00:18:28] Speaker 03: And then we put in parentheses as an example, eg, [00:18:31] Speaker 03: a task 1104A through 1104C. [00:18:35] Speaker 03: So what we identified in Rebecca is satisfying an operation which is a comprising procedure of the application was one of these, any of these tasks 1104A, B, or C. And so I'd like to go to that figure five because I do think walking through that makes this clear. [00:18:57] Speaker 03: Let's look at figure five. [00:18:59] Speaker 03: And I'm looking on the red brief at page 43, because that's the annotated version that the parties were discussing about figure five. [00:19:08] Speaker 03: You see there on figure five, there is on the third row, there is something called in the middle, the job repository for Monte Carlo X-ray generation code 1102B. [00:19:18] Speaker 03: That was identified as the job. [00:19:22] Speaker 03: The application is the job, so that job. [00:19:24] Speaker 03: That job then branches out into three different boxes. [00:19:28] Speaker 03: And those are the boxes that we just identified. [00:19:30] Speaker 04: Let me ask a stupid question. [00:19:33] Speaker 04: What is synonymous to an operation? [00:19:35] Speaker 04: A task, or a job, or something, a subset? [00:19:39] Speaker 04: So that's what I'm getting to. [00:19:41] Speaker 03: So then we start with job. [00:19:43] Speaker 04: Very excited about that, weren't you? [00:19:45] Speaker 03: Sorry. [00:19:46] Speaker 03: Sorry. [00:19:46] Speaker 03: I like this one. [00:19:47] Speaker 03: So 1102B is a job, right? [00:19:50] Speaker 03: That's the application overall. [00:19:52] Speaker 03: Then part of the application gets broken up into what are called, the application gets broken up into operations. [00:19:59] Speaker 03: And that, for Becky, is a task. [00:20:02] Speaker 03: So X-ray generation gets split up into 1104A is X-ray generation in aluminum. [00:20:08] Speaker 03: That's one task. [00:20:09] Speaker 03: A second task is shown, 1104B, X-ray generation in copper. [00:20:14] Speaker 03: And lastly, there's X-ray generation in silicon. [00:20:17] Speaker 03: Then what you see below that, and highlighted in the yellow boxes, is that task, so for example, 1104A for X-ray generation in aluminum, that task, which is an operation, is being performed on different sets of electrons. [00:20:32] Speaker 03: So it's kind of, I guess, analogy would be the task is multiplication, subtraction, addition. [00:20:38] Speaker 03: And then you're performing that task on a data set. [00:20:41] Speaker 03: The data sets there are the yellow boxes. [00:20:44] Speaker 03: And I think the key issue here is beyond the figure five, [00:20:48] Speaker 03: And Dr. Jagdish's explanations of that were credited by the board. [00:20:52] Speaker 03: We also have their expert, Dr. Rosenblum. [00:20:55] Speaker 03: And I'd like to go to 1720, appendix 1720. [00:20:59] Speaker 03: This is exactly what blows up their argument that they don't have support even from their own expert. [00:21:05] Speaker 03: So if we go to page 1720 of the appendix, this is the deposition of Dr. Rosenblum. [00:21:12] Speaker 03: We asked him on deposition page 186, line 19, [00:21:18] Speaker 03: Do you believe, this is asking about figure five, do you believe that there are three tasks under 1104A, or is there a single task under 1104A? [00:21:27] Speaker 03: And his answer, I believe in this example that there's a single task, and it's the example of the situation described in the specification of Rebecca, where a single task is given to a plurality of workers. [00:21:38] Speaker 03: So again, going back to the claim language, tell a plurality of nodes to perform an operation. [00:21:43] Speaker 03: What we said operation was, was a task. [00:21:45] Speaker 03: He's answering. [00:21:46] Speaker 03: that a single task is being given to a plurality of nodes. [00:21:50] Speaker 03: And then we continue that, you know, any confusion, again, another question following that immediately. [00:21:56] Speaker 03: So for this task that's under 11.04a, do you believe that different boxes would be spread across the three workers? [00:22:01] Speaker 03: And you said, yes, I believe that. [00:22:03] Speaker 03: You're executing the task on these different sets of electrons. [00:22:06] Speaker 03: So the same task is being [00:22:08] Speaker 03: that they're telling the porality nodes to perform an operation or a task. [00:22:12] Speaker 03: And that task can be 1104A, 1104B, 1104C, as well as other examples. [00:22:18] Speaker 03: And it's not limited to figure five, but that's the figure that the board they raised in their patented response, the experts discussed in the deposition, and the board then discussed in the final decision in rejecting their argument. [00:22:31] Speaker 03: So I think that addresses one B. I'd just like to briefly address the one E arguments about the one note. [00:22:38] Speaker 03: And the reason for that, I think, is if we look at the final written decision on pages, page 19 and 20. [00:22:46] Speaker 01: Inside what page? [00:22:47] Speaker 03: 19 and 20. [00:22:48] Speaker 03: The reason I'm pointing to these pages is that [00:23:04] Speaker 03: Limitation 1E was not contested in any way by the patent owner below at the PTAB. [00:23:10] Speaker 03: So the board summarized our explanation of how Rebecca teaches 1E on page 19 and continues on page 120 to say, having reviewed the contentions, we find the evidence persuasive. [00:23:24] Speaker 03: And if we go to even the patent owner response on appendix 1334 to 1335, the patent owner response below at the PTAB, there's no mention of 1E. [00:23:33] Speaker 03: This is literally an argument that has to be forfeited. [00:23:37] Speaker 03: There's no mention in their appellate brief. [00:23:39] Speaker 03: There's no section in their appellate brief about 1E. [00:23:42] Speaker 03: So these arguments today about the which one node and 1E are nowhere to be found in their briefing here or in the whole PTAT proceedings. [00:23:52] Speaker 03: We never had an argument about 1E being satisfied. [00:23:56] Speaker 03: We still have our obligation to show it. [00:23:58] Speaker 03: We showed it. [00:23:58] Speaker 03: The board credited it. [00:23:59] Speaker 03: But they had no counter arguments to us. [00:24:01] Speaker 03: And I'd like to direct. [00:24:03] Speaker 03: your attention also to appendix pages 548 to 550 and I'll go to those briefly. [00:24:16] Speaker 03: And these appendix pages are from our experts declaration and from paragraph 156 [00:24:23] Speaker 03: through paragraph 160, he provided various explanations of how Rebeki's overall system teaches which worker or which node. [00:24:33] Speaker 03: So for example, in paragraph 156 on page 549, he talks about which worker will perform the task. [00:24:42] Speaker 03: He then goes on in paragraph 160 on appendix page 550. [00:24:48] Speaker 03: He explains also that it would have been obvious. [00:24:50] Speaker 03: So beyond the actual teachings of Rebeki, [00:24:53] Speaker 03: an explanation of obviousness based upon four other prior references. [00:24:58] Speaker 03: The patent owner chose in the PTAP proceedings not to contest 1E. [00:25:02] Speaker 03: They chose not to even address these other references, the obviousness argument. [00:25:05] Speaker 04: OK, we have this. [00:25:06] Speaker 04: Why don't you move to 1D and then hit upon the dependent claim, and then take your seat. [00:25:10] Speaker ?: Sure. [00:25:14] Speaker 03: I think the removing code, so dependent claim 17, roomy code, I think Judge Hall correctly identified. [00:25:19] Speaker 03: It is essentially that they're asking for a claim construction issue. [00:25:22] Speaker 03: The evidence is clear. [00:25:25] Speaker 03: I don't think it's disputed that code is removed. [00:25:29] Speaker 03: And they are now arguing that, well, the code has to be removed from a worker node. [00:25:32] Speaker 03: And the board did not agree with that. [00:25:34] Speaker 03: They did not propose a claim construction for that. [00:25:36] Speaker 03: So substantial evidence supports that position. [00:25:40] Speaker 03: 1, D. [00:25:44] Speaker 03: I think you asked about one other one. [00:25:51] Speaker 03: I think those were the ones that I wanted to address, Your Honor, unless there's any other questions. [00:25:55] Speaker 04: Okay, thank you very much. [00:25:57] Speaker 03: Thank you. [00:26:05] Speaker 04: Mr. Young, I gave you two extra minutes. [00:26:30] Speaker 02: Thank you, Your Honor. [00:26:31] Speaker 02: Just very quickly, with respect to figure five and that testimony with respect to Dr. Rosenblum, on appendix 1720 on page 185, the question is, is your understanding that each one of those yellow boxes is a separate task? [00:26:44] Speaker 02: Answer, yes. [00:26:46] Speaker 02: That is my understanding. [00:26:47] Speaker 02: It's the data for three separate tasks. [00:26:50] Speaker 02: So what you have here is a situation where you have each node [00:26:54] Speaker 02: is being assigned a specific task from the task dispatcher. [00:26:58] Speaker 02: The situation is not where the task dispatcher says to three nodes, perform an operation, and those nodes decide which one will do the operation. [00:27:08] Speaker 02: This also dovetails them with respect to claim one E, where they say we did not raise it, or we did not argue that claim one E was not satisfied. [00:27:17] Speaker 02: The reason for that is they decided to map, to not include the task dispatcher as part of their mapping. [00:27:23] Speaker 02: So under that mapping, [00:27:25] Speaker 02: then yes, the first node doesn't know which one of the nodes is performing the operation, because all the first node does is send in the application, which again, is mapped to the, application is mapped to job, operation is mapped to task. [00:27:41] Speaker 02: So all the job dispatcher, all the job submitter does is submit the operation or the job to the task dispatcher. [00:27:49] Speaker 02: Task dispatcher makes that determination. [00:27:51] Speaker 02: So under that situation, yes, one E is satisfied. [00:27:54] Speaker 02: But you can't satisfy one E and still satisfy one B. That's the point. [00:27:59] Speaker 02: Because there is nothing that goes from the job submitter that here's an app, here's an operation. [00:28:04] Speaker 02: You three or four or five nodes, your plurality of nodes perform the operation. [00:28:09] Speaker 02: And then we don't know which one of you performs the operation. [00:28:13] Speaker 02: So with respect to figure five, even if you say, yes, this is the same [00:28:19] Speaker 02: task or same operation that each of the three nodes is performing, that's not what the back end of the claim requires. [00:28:25] Speaker 02: It says only one. [00:28:26] Speaker 02: And that's because you go back to the centralized versus decentralized nature of the asserted claim, which is decentralized, to the centralized nature of the prior art. [00:28:36] Speaker 02: They're vastly different. [00:28:38] Speaker 04: OK. [00:28:38] Speaker 04: Thank you, counsel. [00:28:39] Speaker 04: Your time is up. [00:28:39] Speaker 04: I thank all counsel. [00:28:40] Speaker 04: Those cases take under submission.