[00:00:00] Speaker 04: We will hear argument next in number 231980, Juniper Networks Against Swarm Technologies. [00:00:08] Speaker 04: Ms. [00:00:08] Speaker 02: Boyd. [00:00:10] Speaker 02: Good morning, Your Honors. [00:00:11] Speaker 02: May it please the court? [00:00:11] Speaker 02: My name's Karen Boyd. [00:00:12] Speaker 02: I'm representing the Appellate Juniper Networks, who is the petitioner on IPR for the board. [00:00:20] Speaker 02: Unless Your Honors have a different preference, I'll first address the construction of the negative limitation without any communication with the controller. [00:00:29] Speaker 02: why that communication, the communication that it prohibits is two-way information exchange between a coprocessor and a controller. [00:00:38] Speaker 02: Then I'll turn to why under the incorrect construction applied by the board that allowed for both one-way and two-way information exchange that there still is [00:00:53] Speaker 02: that the Young-Apple talkbook combination still invalidates the patents at issue. [00:01:00] Speaker 02: And finally, I'll explain why that Apple-Young obviousness argument is properly fully considered here, including the particular parts of the argument raised in the reply brief in response to the patent owner's arguments. [00:01:18] Speaker 02: So the claim term at the center of this appeal for both the 275 and the 004 patent [00:01:23] Speaker 02: is there's a system that needs to accept multiple co-processors, quote, on a plug and play basis without any communication with the controller. [00:01:34] Speaker 02: The board focused on, for claim construction, the parties originally said that plain and ordinary meaning applied. [00:01:43] Speaker 02: The board said that that's what it was doing, but then it said, well, what is the plain and ordinary meaning? [00:01:49] Speaker 02: and said, well, we look at any communication. [00:01:51] Speaker 02: Any is a broad term. [00:01:53] Speaker 02: And so any communication should include one-way information transfer and two-way information transfer. [00:02:01] Speaker 02: And I will submit that the board stopped one word too soon in looking at this claim language. [00:02:09] Speaker 02: What this negative claim limitation prohibits is any communication with the controller. [00:02:16] Speaker 02: And I think that with, [00:02:18] Speaker 02: word is very important. [00:02:23] Speaker 02: With is a normal English word. [00:02:28] Speaker 02: Prepositions matter. [00:02:29] Speaker 02: On something is different from under something. [00:02:33] Speaker 02: With is another preposition. [00:02:35] Speaker 02: And it implies and it means [00:02:38] Speaker 02: that there is communication back and forth. [00:02:41] Speaker 03: I don't see that. [00:02:43] Speaker 03: I feel as if I am communicating with you, even if you stay silent and give me no response whatsoever. [00:02:50] Speaker 03: Haven't I communicated with you once you receive it? [00:02:54] Speaker 02: I think that that's a good analogy. [00:02:58] Speaker 02: And something that I have thought about in preparing for this is [00:03:04] Speaker 02: talking at someone versus talking to someone versus talking with someone. [00:03:10] Speaker 02: And I think those are three different things. [00:03:13] Speaker 02: When you talk at someone, you don't even know if they're receiving. [00:03:17] Speaker 01: There's no way to... [00:03:20] Speaker 01: intrinsic evidence that supports that or even extrinsic evidence that supports that because I have the same view as Judge Stark that you still could communicate. [00:03:31] Speaker 01: It's whether the co-prosser is communicating with the controller and you've got two things there and one of them is communicating. [00:03:40] Speaker 02: One of them is sending information. [00:03:43] Speaker 01: But it's sent to the controller, right? [00:03:46] Speaker 02: It is sent broadly and [00:03:50] Speaker 02: This is actually a different part of the argument. [00:03:52] Speaker 01: But there is... Well, let's go back to the intrinsic evidence on one-way versus two-way communication. [00:03:59] Speaker 02: So every time that the term communicate or communication is used in the specification, it's either communicate with or it doesn't really specify. [00:04:14] Speaker 02: Or there is one section where it is clearly too [00:04:20] Speaker 02: two-way information transfer back and forth that is communicated, but that's being discussed in the specification. [00:04:26] Speaker 02: I think the more important intrinsic evidence that supports this beyond what's in, beyond the claim language itself, is the purpose of the invention and the purpose of this element is the idea of plug and play without any communication with the controller. [00:04:46] Speaker 02: The whole purpose of it is to [00:04:49] Speaker 02: reduce the controller resources that are required for this system. [00:04:55] Speaker 01: Where is that stated in the specification? [00:04:59] Speaker 02: That is... Your Honor, it will take me a moment to find that particular citation and so I will... I'm happy to provide that during my rebuttal time if that's a more efficient way of... If that's what you prefer to do. [00:05:17] Speaker 01: I think it's a very important question, though, to your argument because I'm having a hard time seeing that in the specification. [00:05:24] Speaker 02: OK. [00:05:41] Speaker 04: Were you thinking about the general observation just at the very end of the background section about multi-processor architecture needed, which reduces CPO management overhead? [00:05:53] Speaker 04: Or is there something different? [00:05:55] Speaker 02: That is part of it. [00:05:56] Speaker 02: This is sort of the raison d'etre of the patent. [00:06:01] Speaker 02: So it is throughout the specification. [00:06:03] Speaker 02: But that certainly is an example, Your Honor. [00:06:05] Speaker 02: Thank you. [00:06:11] Speaker 02: So another piece of intrinsic evidence that supports this idea is that in the claim language in claims eight and nine of the 004 patent, that uses a different term. [00:06:32] Speaker 02: For one-way communication, those claim terms use the term notify. [00:06:37] Speaker 02: Now it's a different, it's not the same element. [00:06:41] Speaker 02: But it is a different way of, a different word that's used in those claims that means something different. [00:07:10] Speaker 02: The extrinsic evidence also supports the construction of without any communication with the controller being about two-way information transfer between the controller and the co-processor. [00:07:29] Speaker 02: The expert for Juniper gave testimony about that in his expert report. [00:07:40] Speaker 02: And in deposition, the expert for the patent owner also says, and sort of out of nowhere says, by the way, when there's only one way communication, data packets going out and not being responded to or nothing happening from the controller based on those data packets, that isn't communication. [00:08:08] Speaker 03: But the board found that not to be credible, kind of at their own sponsor's invitation. [00:08:16] Speaker 03: And we usually defer to a board credibility finding, don't we? [00:08:22] Speaker 02: Certainly a credibility finding is normally deferred to. [00:08:27] Speaker 02: But here, the reason for the determination that the board said that this was not credible was that it was inconsistent with [00:08:38] Speaker 02: the expert's other report to reject a position taken, sort of a voluntary position taken by an expert about what a term in this patent means because it is not the same as what he said before. [00:09:00] Speaker 02: It's kind of ignoring the point of how do people of ordinary skill in the art use this term. [00:09:05] Speaker 03: Is it an abuse of discretion for the [00:09:08] Speaker 03: or to conclude, we're just not going to put any weight on what he said in the deposition. [00:09:13] Speaker 02: I believe it is not an abuse of discretion. [00:09:19] Speaker 02: But I think that the fact that the extrinsic evidence separately supports the intrinsic evidence and the purpose of the patents, as well as the normal English language of the difference between [00:09:36] Speaker 02: notify and communicate with, it is another piece of evidence that it's not dispositive. [00:09:46] Speaker 02: And so I think that that's where the abuse of discretion is standard. [00:09:50] Speaker 02: It doesn't change the decision. [00:09:52] Speaker 02: It supports a different decision as well, though. [00:09:58] Speaker 02: Yes, Renner? [00:10:03] Speaker ?: No. [00:10:04] Speaker 02: Even if the term without any communication with the controller is construed to include both two-way information transfer and one-way information transfer, the Apple Talkbook-Liang combination should be found to invalidate these patents and to disclose this element. [00:10:30] Speaker 02: And that's because the combination does not necessarily and always result in even one way information transfer. [00:10:43] Speaker 01: Why is that? [00:10:46] Speaker 01: I mean, because the Apple Talkbook does say, doesn't it, that there's communication? [00:10:54] Speaker 02: Liang says that there is data packet transfer, that there's a data packet that's broadcast. [00:11:01] Speaker 02: And the question is, does the controller ever receive that information? [00:11:07] Speaker 02: And so the evidence that was provided to the board was that if an Apple Talk packet, I'm sorry, I apologize your honor, I think I had the references mixed up there, it's the Apple Talk packet [00:11:25] Speaker 02: Apple Talk sends data packets and if it is sent to a Leon controller in a combination, Juniper presented evidence to the board that the controller would ignore it or discard it. [00:11:39] Speaker 02: It doesn't receive the information. [00:11:41] Speaker 02: It's sort of like when you ask your teenagers to clean their rooms and they just [00:11:48] Speaker 02: You know, it falls on deaf ears. [00:11:49] Speaker 01: But don't you still communicate with them, even if they don't listen? [00:11:55] Speaker 04: You get to say the next day, I asked you to do that, and you know that, right? [00:12:01] Speaker 02: But if they haven't done anything, and they haven't cleaned their room, and they haven't said, I can't clean my room... So they can say, you never told me to clean my room? [00:12:10] Speaker 01: Can they say that? [00:12:12] Speaker 02: I've certainly had them say that. [00:12:13] Speaker 01: But is it accurate? [00:12:15] Speaker 02: Communication though, is there a communication that's effective communications that has happened there is the question. [00:12:23] Speaker 01: Does the claim say effective communication? [00:12:28] Speaker 02: It says communication with. [00:12:29] Speaker 01: It says any communication. [00:12:31] Speaker 02: It says any communication with the controller and the purpose [00:12:34] Speaker 02: again, is to limit what the controller has to do. [00:12:38] Speaker 03: And you're saying the proper construction requires at least receipt by the controller of the community. [00:12:44] Speaker 02: Otherwise, it's not the communication. [00:12:46] Speaker 02: And that is not disputed by Swarm, to my knowledge. [00:12:51] Speaker 02: The way that they have talked about one-way communication is that the communication does need to be received. [00:12:59] Speaker 02: The information does need to be understood. [00:13:02] Speaker 02: And Juniper presented evidence that because of the way that Apple Talk and Yong would be combined to use the plug and play components of the Apple Talk book, that you would not, that the controller wouldn't necessarily receive any of that information, even if it was broadcast by [00:13:32] Speaker 02: Apple, Apple Talk. [00:13:34] Speaker 04: You are well into your rebuttal time. [00:13:36] Speaker 04: Maybe you want to reserve the rest of this time for rebuttal. [00:13:42] Speaker 04: We'll hear from the other side and then you can. [00:13:44] Speaker 02: Yes, I will do that. [00:13:45] Speaker 02: Thank you, Your Honor. [00:14:00] Speaker 00: Good morning, your honors. [00:14:01] Speaker 00: May it please the court. [00:14:03] Speaker 00: I'd like to pick up right where Miss Boyd left off with regard to receiving the communication. [00:14:11] Speaker 00: So have a talk. [00:14:14] Speaker 00: Appendix 802, quoting here, quote, a broadcast packet doesn't travel to one particular node, but is sent to every node on the network in turn. [00:14:25] Speaker 00: and this is the key part, every node receives a broadcast packet, and if appropriate, responds to it. [00:14:33] Speaker 00: So I don't think there's any question in Apple Talk whether the packet is received. [00:14:39] Speaker 03: You concede, though, that it must be received by the controller in order to be one-way communication. [00:14:46] Speaker 00: Yes, yes. [00:14:47] Speaker 00: You're just saying it is received. [00:14:49] Speaker 00: Yes, absolutely. [00:14:51] Speaker 04: So I think as I heard Ms. [00:14:53] Speaker 04: Boyd, I'm not sure whether I heard this right, that she was talking about whether that would be true in the combination [00:15:03] Speaker 04: of Apple Talk and Leon. [00:15:09] Speaker 04: I don't understand exactly whether I heard that right or what it would mean. [00:15:15] Speaker 04: Is it clear that if it's in Apple Talk, then it would also be in the combination? [00:15:23] Speaker 04: Was there something about combining with Leon that might undo that piece, that little in turn is received piece of Apple Talk? [00:15:33] Speaker 00: Right. [00:15:34] Speaker 00: So I'll try to characterize as best as I can Juniper's argument on that and respond. [00:15:39] Speaker 00: But before I do that, I'd like to highlight that these particular arguments were also rejected by the board for being raised late in the IPR proceeding. [00:15:49] Speaker 00: So while they rejected them for that reason, they did actually go on to say, but even if we did consider them, they still don't [00:15:58] Speaker 00: get us there. [00:15:59] Speaker 00: The argument I think that Juniper raised on this issue was that if you further modified this combination of Apple Talkbook, Leon, to place the co-processors and the one particular co-processor that they set apart as the controller, it's called the surveying unit in Leon, if you put them on different zones [00:16:23] Speaker 00: and then you set up some bridges and gateways, you could block the transmission of a broadcast packet to a particular node. [00:16:33] Speaker 00: The reason that argument fails on the merits, in addition to being late raised, Apple Talkbook describes setting up zones to limit inter-network traffic, but the [00:16:47] Speaker 00: Controller has a need to communicate with the task pool. [00:16:52] Speaker 00: So do the co-processors. [00:16:53] Speaker 01: Can I make sure I understand something? [00:16:55] Speaker 01: Are you saying the only basis for saying that the combination wouldn't have the controller receive the broadcast would be if the zones were used? [00:17:06] Speaker 00: Correct. [00:17:07] Speaker 01: Is there any other argument being made that the combination, like modifying Leon in view of Apple Talk, there would be no need for the controller to receive the broadcast message? [00:17:24] Speaker 00: Not as I understand it. [00:17:26] Speaker 00: The Apple Talk book is very clear. [00:17:28] Speaker 00: If you're on the network, if you're in that network, and there's a broadcast packet sent, [00:17:32] Speaker 00: all of the nodes receive the broadcast packet. [00:17:36] Speaker 00: The example provided in Apple Talkbook, which is the example that Juniper relies upon in both of its petitions, looking at appendix 803, and I think it's just a little bit illustrative because it's kind of a concrete example here. [00:17:54] Speaker 00: You can see in the figure 411, it says use of broadcast packets, and we see exactly how this works. [00:18:00] Speaker 00: One of the nodes reaches out, says, any laser writer's out there. [00:18:05] Speaker 00: And you can see the arrows pointing to each and every node on that network that they're getting that broadcast as a broadcast packet. [00:18:13] Speaker 00: One of them, who is a laser writer, might respond by saying, yeah, here I am. [00:18:18] Speaker 00: But that doesn't mean the others didn't receive it. [00:18:20] Speaker 00: and that it didn't create some burden or some communication disruption to have to receive it, read it, choose, I guess, not to respond to it if it's not a broadcast packet you're going to respond to. [00:18:39] Speaker 00: Returning, if there's no other questions, returning to the earlier portion of the discussion on [00:18:51] Speaker 00: the board's construction. [00:18:54] Speaker 00: The construction is consistent with the specification. [00:18:59] Speaker 00: I think one of the key portions of the specification relevant here. [00:19:06] Speaker 00: I'm looking at appendix 89, that's the 04 patent, column 10. [00:19:12] Speaker 00: Lines 29 through 36 and then later lines 45 through 48. [00:19:16] Speaker 00: What's described here [00:19:18] Speaker 00: is how the coprocessors are admitted into the system. [00:19:23] Speaker 00: And it describes that the CPU or the controller is going to delegate all of this responsibility over to the task pool. [00:19:32] Speaker 00: It's going to give it the protocols, the authentication requirements. [00:19:35] Speaker 00: It's going to place it in charge of whether it's going to admit or exclude particular coprocessors. [00:19:41] Speaker 00: So it's now the task pool that's being entrusted with letting coprocessors in. [00:19:47] Speaker 00: and then further down in that same column that there's the embodiment where all of the co-processors are even wired directly to the task pool. [00:19:56] Speaker 00: In fact, it goes so far as to say you could remove the switching fabric, which means there's only one communication path left for these co-processors to communicate with any node on its particular network, and that's the task pool. [00:20:10] Speaker 00: So this particular example is consistent with [00:20:14] Speaker 00: the board's interpretation that there's no communication without any communication, whether one-way or two-way. [00:20:21] Speaker 00: This distinction of one-way or two-way communication is something that Juniper gends up to create this controversy, but it's not something you would expect to find in the patent specification described because that's not how the patent looks at communication. [00:20:34] Speaker 00: It's trying to take the controller out of that process altogether. [00:20:38] Speaker 00: We don't want one-way communications. [00:20:40] Speaker 00: We don't want two-way communications for an England controller. [00:20:43] Speaker 01: Your best argument here, as I understand it in column 10, is you're saying there's an embodiment. [00:20:48] Speaker 01: And in that embodiment, there's no communication whatsoever. [00:20:51] Speaker 01: So that supports the idea when it says without any communication, it means any. [00:20:58] Speaker 01: Right. [00:21:00] Speaker 01: Is there anything to your knowledge in the specification that even uses the phrase one-way communication or two-way communication? [00:21:07] Speaker 00: No, not as I understand it. [00:21:09] Speaker 00: The examples provided by Juniper, which were reviewed by the board, Juniper had its opportunity before the board to put forward something that would discuss this. [00:21:18] Speaker 00: Juniper put forward a few examples. [00:21:20] Speaker 00: They're not in the context of adding new co-processors into the system, but rather the passages cited by Juniper are in the context of [00:21:31] Speaker 00: co-processors receiving tasks, processing them. [00:21:34] Speaker 00: But even one of those examples, the board found the communication channel phase example was an example that supported the board's construction because it described a form of one-way communication where a task was retrieved from the task pool to the co-processor. [00:21:50] Speaker 00: And that was just a one-way communication in the communication channel phase. [00:21:56] Speaker 03: Do you agree that the purpose of the patent is to reduce the use of controller resources? [00:22:02] Speaker 03: And if so, does that affect the claim construction? [00:22:06] Speaker 00: Yes, I agree with that, absolutely. [00:22:08] Speaker 00: One of the main goals of the patent is to reduce the burden on the controller of processing, responding, being interrupted. [00:22:16] Speaker 00: Any of these examples are examples of the controller having its own resources taken away [00:22:23] Speaker 00: and we're trying to save its resources for more important things. [00:22:29] Speaker 00: The task pool is there as an intermediate device, and that's who the co-processors communicate directly with, particularly in these limitations. [00:22:37] Speaker 00: I don't think it changes the claim construction. [00:22:41] Speaker 03: Can I ask you, I'm sure you're aware the next appeal involves, I think, the same patents. [00:22:46] Speaker 03: Yes. [00:22:48] Speaker 03: My only question is, in that appeal, I believe, [00:22:54] Speaker 03: The petitioner persuaded the board that all of claims 1 to 12 of the 004 are unpatentable. [00:23:01] Speaker 03: If we were, and I'm not saying we would, but if we were to affirm that they're unpatentable, 1 to 12 in that appeal, would we have to decide anything about the 004 in this appeal? [00:23:17] Speaker 00: Well, for the 275 patent, yes. [00:23:19] Speaker 00: For the 275, I'm just talking about the 004. [00:23:20] Speaker 00: Right. [00:23:21] Speaker 00: If the 004 patent is declared invalid on the one hand, then... Right. [00:23:27] Speaker 03: Even though you prevailed on the IPR in this case on some of those claims, if we affirm their non-survival in the later appeal, then anything about the 004 would be moot in this appeal. [00:23:41] Speaker 03: Is that right? [00:23:42] Speaker 03: I would understand it that way. [00:23:43] Speaker 03: Okay. [00:23:43] Speaker 03: Thank you. [00:23:48] Speaker 00: One last point just to throw out there and I'll conclude. [00:23:52] Speaker 00: I think it's important to take a look in this case because this is the root cause for a lot of the issues that we're here discussing today in the 004 and 275 petitions, particularly the section that we're speaking about today. [00:24:07] Speaker 00: Juniper calls it element 1.5. [00:24:09] Speaker 00: That section in both of the petitions describes [00:24:16] Speaker 00: the motivation to combine, it describes the likelihood of success it describes. [00:24:21] Speaker 00: the plug and play, the dynamic acceptance, but you won't find its description of how the Leon Appletalk book meets this limitation of doing it all without any communication with the controller. [00:24:33] Speaker 00: It's absent. [00:24:34] Speaker 00: So as an example, appendix 238 to 244. [00:24:38] Speaker 00: I think that's a good place to look. [00:24:39] Speaker 00: It's not there, and as a result of that, these issues bubbled up and resulted the way it did. [00:24:46] Speaker 00: With that, I'll conclude. [00:24:47] Speaker 00: If there's no more questions. [00:24:49] Speaker 00: Thank you. [00:25:13] Speaker 02: To start, I want to confirm that we agree that if the O4 patent is invalidated separately, that it's moved here. [00:25:24] Speaker 02: To follow up on the question about is the point of this patent to reduce burdens on the CPU or the controller, two different terms for the same thing, [00:25:39] Speaker 02: We have an acknowledgment that that is indeed the purpose. [00:25:43] Speaker 02: Citations including at column one paragraph, line 61 through column two at line seven. [00:25:56] Speaker 02: I'd like to address briefly the concern that some of these arguments were raised on reply before the board. [00:26:07] Speaker 04: In particular, the one about the zones. [00:26:09] Speaker 02: In particular, the one about the zones and how exactly, when you combine that Apple Talk reference and the Leon reference, what happens to communication, what happens to [00:26:22] Speaker 02: these data packets that are sent and are they sent across zone boundaries? [00:26:28] Speaker 02: So there are two different things to consider about what's properly raised in reply. [00:26:35] Speaker 02: One is, is it a new theory of obviousness? [00:26:39] Speaker 02: And that is a question that's reviewed de novo under inovative. [00:26:45] Speaker 02: Here, it is the same combination. [00:26:48] Speaker 02: It is the same combination using the same elements [00:26:52] Speaker 02: tracking to the two different pieces of prior art. [00:26:55] Speaker 02: So we're not pointing to a new embodiment. [00:27:00] Speaker 01: We're not pointing to a new... Isn't the addition of zones, we don't think that's something new. [00:27:07] Speaker 01: I mean, the petition doesn't mention zones, but then the reply argument is that you should separate different parts of the same network using zones, and that would then cut off communication. [00:27:18] Speaker 01: Do I understand that right? [00:27:19] Speaker 02: It would cut off communication, but it is not a different embodiment because the Apple Talkbook is a single reference about a single way to do, it's about how to network these, at this point, very old Apple products. [00:27:34] Speaker 02: So although it's a long reference, it's all about the same reference. [00:27:37] Speaker 03: But you could have one zone or multiple zones in Apple Talkbook, could you not? [00:27:42] Speaker 03: Not once you combine it with Leon, because Leon doesn't have its... How was the board or the patent owner to know that was your view from reading your petition? [00:27:54] Speaker 02: That there would be... Until you have this issue of two-way communication versus one-way communication that has been raised sort of in response, and that's the other part of is it too late to raise it on replies, is it responding to [00:28:12] Speaker 02: something in the opposition brief that's filed by the pad member. [00:28:17] Speaker 02: And that's what this argument is doing. [00:28:19] Speaker 03: And that we review for abuse of discretion, right? [00:28:21] Speaker 02: And that is reviewed for abuse of discretion, yes, Your Honor. [00:28:25] Speaker 02: And I would just want to point quickly to the case, Apple versus Andrea Electronics, 949 Fed 3rd, 697. [00:28:39] Speaker 02: This is actually cited in the [00:28:41] Speaker 02: appellants briefs. [00:28:46] Speaker 02: This court held that the board improperly concluded that Apple's reply brief raised a new theory of patentability. [00:28:56] Speaker 02: The comment is that Apple's legal ground did not change in its reply. [00:28:59] Speaker 02: Its reply still asserted that claims six through nine would have been obvious over two references. [00:29:04] Speaker 02: Here, the same Leon plus talkbook combination is relied on in reply. [00:29:09] Speaker 02: And the zones argument is simply responding to the previously relied on combination in response to appellee's arguments in response to the petition. [00:29:20] Speaker 04: Your time has run out. [00:29:23] Speaker 04: We should say this case is submitted. [00:29:26] Speaker 04: Thank you, Your Honor. [00:29:27] Speaker 04: And move to the next.