[00:00:00] Speaker 01: The parties are flipped, but we're staying in the same place. [00:00:03] Speaker 01: And there's no cross-the-bill in this case, thankfully. [00:00:07] Speaker 01: Mr. Quinn. [00:00:09] Speaker 01: Thank you, Judge Fruz. [00:00:37] Speaker 02: may it please the court, the board's finding that the claims of the 668 patent are obvious is legally flawed and should be reversed. [00:00:45] Speaker 02: In a related proceeding over this patent, the ITC had found that Arista copied the patented features of Cisco's technology while willfully blinding itself to this patent, and it rejected challenges to the validity of the patent over the same Amara reference. [00:00:58] Speaker 02: Now, the board here reached a different outcome for three independent reasons, any one of which requires remand. [00:01:05] Speaker 02: First, it fundamentally misread the patent's requirement that packets destined to the control plane receive not just control plane port services, but also normal port services. [00:01:15] Speaker 02: Second, its analysis of secondary considerations of non-obviousness is legally flawed and misunderstands the legal significance of Arista's blatant copying of Cisco's feature names, command line expressions, all while having access to Cisco's technology itself. [00:01:32] Speaker 02: And third, the board failed to satisfy the APA's requirements for reasoned decision making, particularly in addressing the issue of long felt need and whether the core builder reference can satisfy the limitations the board found missing from AMARA, a point that the board seems to assume in its opinion but never actually addresses. [00:01:50] Speaker 02: Again, any one of these requires that the board's decision be vacated. [00:01:54] Speaker 04: I guess getting to copying, the board seemed to say just because the command line expressions and name of the product were the same, that's not enough. [00:02:08] Speaker 04: And then I guess they also considered a claim chart where you showed how there are various functionalities that correspond to functionalities of your client's product. [00:02:21] Speaker 04: Even if there's infringement, even if there's the same usage of various nomenclature, that's not enough to actually show copying as opposed to infringement. [00:02:35] Speaker 04: Sure. [00:02:36] Speaker 04: So that's a judgment call that they made based on the record in front of them. [00:02:42] Speaker 04: How do we necessarily reverse that, that the board was compelled [00:02:49] Speaker 04: in light of all the evidence that they had to have found that there was actual intentional copying going on of Cisco's product? [00:03:01] Speaker 02: Sure, Judge Gent. [00:03:01] Speaker 02: Well, first, I think what the board did here is much like what this court criticized recently in Institute Pasteur, where the board, quote, stopped its analysis of the evidence of copying prematurely. [00:03:12] Speaker 02: And you know that the board did this because the board's conclusion on this matter at Appendix 33 [00:03:17] Speaker 02: is to say that neither the feature names nor the command line expressions are claimed in the 668 patent, and thus dismisses the evidence of copying as being very weak. [00:03:26] Speaker 02: The fact that the feature names and the command line expressions are not in the patent is beside the point. [00:03:32] Speaker 02: Neither are the user manuals, which Arista copied, including down to the typos, and you can see that at appendix 13241. [00:03:39] Speaker 02: The point that the board doesn't seem to actually grapple with is that when you have the copying [00:03:46] Speaker 02: of the same feature names, the same command lines, of the same user experience as their CEO, Arista's CEO testified about, and you have the accused product giving you the same results, the only reasonable inference to draw is that the features were copied too. [00:04:06] Speaker 02: After all, if the user is using the same names and having the same look and feel and getting the same results, [00:04:15] Speaker 02: then the reasonable inference, this is classic circumstantial evidence. [00:04:19] Speaker 02: The reasonable inference is that it should be, that you could conclude that the features themselves are copied. [00:04:25] Speaker 02: And this is not a surprising conclusion for the board to draw, given that you have Arista's CEO acknowledging that there's, quote, a lot of Cisco DNA in Arista, and that they, this is in appendix 11802, and at 11809, they design their system where [00:04:44] Speaker 02: Cisco user quote would be able to use Arista right away because we have similar command line interfaces and operational look and feel where we don't have to invent we don't end quote now That's not the only error that the board This is creating an atmosphere of suggesting that they were copying but at the same time maybe [00:05:07] Speaker 04: What the board is trying to say is we need something a little bit more to show actual copying of the elements of Cisco's product that ultimately create the same result that Arista's product is designing and everything else that you're pointing to in the articles is really more about Arista's desire to have its products to be compatible with [00:05:36] Speaker 04: Cisco's products. [00:05:38] Speaker 02: Sure, Judge Shin. [00:05:39] Speaker 02: And to be clear, I'm not asking for a reversal on this issue. [00:05:41] Speaker 02: I'm asking for a reman in the sense that the board just simply didn't follow through on the analysis. [00:05:47] Speaker 02: We're here under APA review. [00:05:49] Speaker 02: And where the board's conclusion on all of this is to say, well, quote, neither the feature names nor the command line expressions are claimed in the patent, [00:05:57] Speaker 02: That's stunting the analysis. [00:05:58] Speaker 02: I understand they're not in the patent. [00:06:00] Speaker 02: That's not the question. [00:06:01] Speaker 02: The question is, does this provide circumstantial evidence of copying? [00:06:05] Speaker 02: And so I think the board got off the train at too early of a stop, just like what happened at Institut Pasteur. [00:06:10] Speaker 02: And that's a reason enough for this court to remand. [00:06:13] Speaker 02: There's a similar issue with respect to long-fell need. [00:06:16] Speaker 02: And if you have the board saying that Cisco, quote, never addressed whether the need was satisfied by another, that's appendix 28, [00:06:24] Speaker 02: and it's suggesting that Cisco, quote, never asserted its product satisfied the long-fell need, Appendix 30. [00:06:31] Speaker 02: You can't reconcile that with the material we presented to the board at Appendix 280, 283, 3637, 284, [00:06:39] Speaker 02: And 235, excuse me, and 335. [00:06:43] Speaker 02: And if the board. [00:06:44] Speaker 04: I think this issue comes down to how did you define what the Longfell need was, that there was this need that needed to be solved in the industry and that it hadn't been solved for a very long time. [00:06:55] Speaker 04: And so what was that need? [00:06:56] Speaker 04: And I thought your side had articulated as being trying to defend against these DOS denial of service attacks. [00:07:07] Speaker 04: And the claimed invention is really more about protecting the control plane against these sorts of flooding attacks. [00:07:16] Speaker 02: Well, and Judge Chen, I think the board took an inconsistent approach to this. [00:07:19] Speaker 02: For example, in looking at some of the evidence on this issue, the board said, well, the question was the sending off denial of service attacks for others. [00:07:32] Speaker 02: It said, are evidence that others had failed? [00:07:34] Speaker 02: wasn't sufficient because that didn't address whether or not they were solving the problem at the control plane. [00:07:40] Speaker 02: So the board's reasoning, and this is articulated in our brief, is internally inconsistent on this point, and that again is yet another reason that the board ought to be remanded. [00:07:49] Speaker 03: My concern is whether Cisco's arguments were internally inconsistent. [00:07:54] Speaker 02: I appreciate the question. [00:07:56] Speaker 02: I don't think that they were. [00:07:57] Speaker 02: I think Cisco's position was always that this was about [00:08:00] Speaker 02: mitigating denial of service attacks and doing so by protecting the control plane. [00:08:05] Speaker 02: And indeed, that actually sort of feeds back into the principle argument here, or the first argument here, which is how is it that this is done? [00:08:15] Speaker 02: The claims require the packets destined for the control plane receive both the control plane port services and the normal port services. [00:08:22] Speaker 02: And you have to read B and D2, the elements, element B and element D2 of the claim [00:08:28] Speaker 02: of the claim together. [00:08:31] Speaker 02: And specifically, if you look at element B of claim 19, it provides that one of the steps is, quote, executing port services on packets entering and exiting the physical network interface ports. [00:08:44] Speaker 02: That's not limited. [00:08:45] Speaker 02: It's not limited to some of the packets. [00:08:47] Speaker 02: That language is unqualified. [00:08:50] Speaker 02: And there is no path depicted in the specification. [00:08:54] Speaker 02: There is no example where a packet reaches the control plane [00:08:57] Speaker 02: without having both normal port services applied and control plane port services applied. [00:09:04] Speaker 02: I think that means that the decision here is inconsistent with this court's reasoning in cases like Henry Smith and Vernetix and Symed, which require that you look at the claims, of course, in light of the specification itself. [00:09:17] Speaker 02: And indeed, under ordinary rules of grammar, when a noun is left unqualified, it applies without qualification. [00:09:23] Speaker 02: Now, another point here, if you look at the argument that Arista is making, [00:09:27] Speaker 02: And their argument is that, well, it's enough that some packets might receive normal port services. [00:09:33] Speaker 02: Well, under that rationale, you have to read B and D2 together. [00:09:38] Speaker 02: D2 says, quote, the control plane port services operate on packets. [00:09:44] Speaker 02: It also isn't modified. [00:09:46] Speaker 02: And so under their rationale, it would be enough that control plane port services were applied to only some of the packets. [00:09:53] Speaker 02: Number one, that makes no sense. [00:09:55] Speaker 02: That would be self-defeating. [00:09:56] Speaker 02: The whole point of this [00:09:57] Speaker 02: is to have enhanced protection for the packets that are going to the control plane. [00:10:03] Speaker 02: And if some of the packets going to the control plane are not going to get control plane port services, then the invention doesn't work. [00:10:10] Speaker 02: Second, it's inconsistent with the specification itself. [00:10:13] Speaker 02: If you look at column three, line 53 to 57, it contemplates that all of the packets are going to receive control plane port services. [00:10:23] Speaker 02: Packets always have to pass through the control plane port entity [00:10:27] Speaker 02: to ensure that those services are applied. [00:10:29] Speaker 02: That's what it says at column 3, line 53. [00:10:32] Speaker 02: And there's a similar point at column 4, line 9. [00:10:35] Speaker 04: What about column 3, line 35, 41, where it says port services are then typically applied to operate on packets entering into or exiting from each individual physical port? [00:10:48] Speaker 02: So Judge Chen, as we explained in our brief, I think what's being talked about there as being typical [00:10:53] Speaker 02: is what's typical is that packets are going to enter through the physical ports. [00:10:57] Speaker 02: There are obviously ways that packets can enter into a device. [00:11:00] Speaker 02: They could come by Wi-Fi and enter through an antenna. [00:11:03] Speaker 02: They're not always going to come through a physical port. [00:11:06] Speaker 02: And when it says typically there, it's talking about typical operation of the system. [00:11:12] Speaker 02: And that is what typically is referring to there. [00:11:17] Speaker 02: It's not suggesting that there are going to be circumstances in which packets are not going to receive normal port services. [00:11:23] Speaker 02: And that would defeat one of the purposes of the invention. [00:11:25] Speaker 02: You have an administrability issue. [00:11:27] Speaker 02: In addition to having enhanced protection for the control plane, the invention is about doing it in a way that, number one, does not sacrifice the effectiveness of transporting transit packets quickly. [00:11:42] Speaker 02: And number two, that it can be done in an administrable way, as opposed to having to go and tinker with all the different physical ports on a device. [00:11:51] Speaker 02: If you had a system where [00:11:53] Speaker 02: the packets headed to the control plane are not going to receive normal port services, then you'd have to reverse engineer and come up with a way for the control plane entity to provide the normal port services as well, which would vary potentially from port to port. [00:12:09] Speaker 02: If the court has further questions, I'm happy to take them, but perhaps if I could reserve the balance of my time for rebuttal, I can respond to... I have a little question. [00:12:16] Speaker 02: Please. [00:12:18] Speaker 04: D1 talks about the control plane port entity. [00:12:22] Speaker 04: so that a set of control plane port services can be applied there too. [00:12:28] Speaker 04: So control plane port services are being applied to a control plane port entity. [00:12:36] Speaker 04: What does that mean? [00:12:39] Speaker 04: What does it mean to have control plane port services applied to the control plane port entity? [00:12:49] Speaker 02: Well, Judge Chen, just two points. [00:12:50] Speaker 02: First, so we're clear, I don't think anything on the arguments that I'm making today turns on this. [00:12:54] Speaker 02: No, I'm just trying to understand what is really going on in this claim. [00:12:58] Speaker 02: I appreciate the question. [00:13:00] Speaker 02: And I think that what's going on here is a reference to the packets that are coming through the control plane port [00:13:12] Speaker 02: entity, because that's going to be what's in transit. [00:13:16] Speaker 02: But even if it's referring back to the control plane port entity, that's another way of saying, well, that's where the control plane port services are going to be applied. [00:13:29] Speaker 02: Thank you. [00:13:30] Speaker 02: Thank you, Chief Judge Prost. [00:13:58] Speaker 00: I'd like to begin with the secondary considerations issue, if that's acceptable to the court. [00:14:02] Speaker 01: Yeah, because the copying is a little, the analysis is a little thin, is it not, given the evidence? [00:14:09] Speaker 00: With respect, I think there's two issues there. [00:14:12] Speaker 00: One is, is there substantial evidence to support the board's conclusion? [00:14:17] Speaker 00: It's a factual question. [00:14:18] Speaker 00: They made a factual determination. [00:14:20] Speaker 00: That is something that has given deference in the question of substantial evidence. [00:14:25] Speaker 00: the record created by Cisco on this, they chose to create this attempt to create a fog of the so-called culture of copying and all of that without really tying it to the issue that they have to prove. [00:14:37] Speaker 00: And the board stuck them with that. [00:14:40] Speaker 00: The board said, we're not going to pay attention to all of this general things that have no nexus at all to the claims. [00:14:46] Speaker 00: We want to see evidence that actually has a nexus to the claim, which is what's legally required. [00:14:52] Speaker 00: And that was not provided. [00:14:53] Speaker 00: They chose to not provide. [00:14:54] Speaker 04: Are you saying if the board had gone the other way and found that there was copying, that they would have lacked substantial evidence to make that finding in the face of all the evidence? [00:15:05] Speaker 00: I'm not saying that. [00:15:06] Speaker 00: I'm saying their determination of the lack of substantial evidence based on this record is supported by substantial evidence, that their conclusion of a lack of copying. [00:15:15] Speaker 00: But there's a second question, too. [00:15:18] Speaker 00: And that is, let's assume the board [00:15:21] Speaker 00: had found evidence of copying. [00:15:24] Speaker 00: I believe under this record, that would not, and if the board had found that that evidence of copying had defeated the substantial evidence of obviousness, I believe that would be error. [00:15:38] Speaker 00: Because the only purpose of these so-called secondary considerations is really to prevent the hindsight combination of core builder and a mark. [00:15:48] Speaker 00: But what's important is below, they never ever challenged the motivation to combine CoreBuilder and Amara themselves. [00:15:55] Speaker 00: They challenged to handle, which is not an issue here, but CoreBuilder and Amara, they did not challenge the motivation to combine that. [00:16:02] Speaker 00: And for good reason. [00:16:03] Speaker 00: Amara is a 3Com patent and CoreBuilder is a 3Com router. [00:16:08] Speaker 00: And so they did not challenge that. [00:16:10] Speaker 00: So the relevance of secondary considerations at this point is minimal to nil. [00:16:15] Speaker 00: Its importance, its relevance, is to prevent the hindsight combination, which is not challenged. [00:16:22] Speaker 00: And therefore, it could not defeat a prima facie finding of obviousness, which the board properly made. [00:16:30] Speaker 01: But how do we know necessarily that the board's conclusion, whether or not it was a legal conclusion or a factual conclusion, that the board would not have, if they had construed the copying as being significant and real, would not have concluded that that was sufficient? [00:16:48] Speaker 00: You're right. [00:16:49] Speaker 00: We can't know what the board would have said. [00:16:50] Speaker 00: I think this Court, in that case, if the board had found prima facie case of obviousness, had found copying, [00:16:58] Speaker 00: and trumped obviousness based on copying, I believe that would be reversible error in this Court, as a matter of law. [00:17:06] Speaker 00: Because the relevance of copying is, as I say, to defeat this combination, which was not challenged. [00:17:12] Speaker 00: And the Court has been very clear that these secondary considerations, although they must be considered, [00:17:17] Speaker 00: Do not. [00:17:18] Speaker 04: Have we ever had a case that said that? [00:17:20] Speaker 04: That if you don't, I mean, let's just say for a moment. [00:17:23] Speaker 04: I'm not aware of one. [00:17:25] Speaker 04: Let's say we're using the phrase prima facie case loosely here, because maybe we're not supposed to be. [00:17:30] Speaker 04: Understood. [00:17:30] Speaker 04: But in any event, right, there's no case. [00:17:33] Speaker 00: I'm not aware of one. [00:17:34] Speaker 04: You said if you don't challenge the underlying grounds of the prima facie case, then you are barred from relying on secondary considerations. [00:17:43] Speaker 00: It's not barred. [00:17:44] Speaker 04: Well, the court does say that... You have no claim to rely on secondary considerations. [00:17:50] Speaker 00: No, it's that the secondary considerations can't trump it. [00:17:53] Speaker 00: This court has said that many times. [00:17:55] Speaker 00: That even evidence of secondary considerations won't trump a strong case of obviousness. [00:17:59] Speaker 00: The court has said that for sure. [00:18:01] Speaker 00: And the court has said that the purpose of this copying of the secondary consideration is to try to negate a hindsight bias to combine references. [00:18:12] Speaker 00: And while the [00:18:14] Speaker 00: The third step in that logical syllogism, I don't think has ever been reached before, has ever been raised. [00:18:19] Speaker 00: It certainly hasn't been rejected by the court. [00:18:22] Speaker 00: But the third step to that syllogism is pretty straightforward. [00:18:25] Speaker 00: If the only relevance is to rebut the combination where the combination is not challenged, then I don't think it would be sufficient under this court's precedent to rebut a strong case of obviousness. [00:18:38] Speaker 00: The court, the board found a strong case of obviousness and said this weak evidence is not sufficient. [00:18:44] Speaker 00: That should be upheld, A, because the finding of weakness was supported by substantial evidence. [00:18:49] Speaker 00: But B, even if the court disagreed, it would not be sufficient to trump the strong case of obviousness anyway under this course precedent. [00:18:58] Speaker 04: What if we disagree with both those arguments? [00:19:01] Speaker 04: And we say, OK, you know what? [00:19:03] Speaker 04: There was copying going on there. [00:19:06] Speaker 04: And secondly, we're not going to make a rule that says you can never [00:19:13] Speaker 04: attack a prima facie case through secondary considerations if you don't attack the prima facie case itself. [00:19:21] Speaker 00: That's not the rule I'm suggesting. [00:19:23] Speaker 00: I'm suggesting that the evidence of copying and other secondary considerations would not be sufficient to overcome the strong case of obviousness. [00:19:31] Speaker 04: Well, that's a different statement. [00:19:33] Speaker 04: That's what I'm saying. [00:19:34] Speaker 04: OK. [00:19:35] Speaker 00: That's what I'm saying. [00:19:36] Speaker 04: Now you're talking in terms of leapfrog, where if it's a very strong case of a prima facie, then [00:19:43] Speaker 00: And that's what the board found. [00:19:47] Speaker 04: Well, the board found here that first it found that the copying was very weak evidence. [00:19:52] Speaker 00: True. [00:19:52] Speaker 04: And then it said, no, that's not right. [00:19:55] Speaker 04: It was actually pretty strong evidence. [00:19:57] Speaker 04: And the question is, would we have to remand that? [00:20:00] Speaker 04: Or would we still say, I guess in your view, nevertheless, in the face of this very strong case of obviousness, that even [00:20:12] Speaker 04: strong evidence of copying cannot prevail in light of the strong case. [00:20:17] Speaker 00: That's my point. [00:20:17] Speaker 00: This Court's precedents are clear that we have a very strong case of obviousness, and I would submit that precedent is even stronger where the relevance predicate of copying isn't there, that the evidence of the secondary considerations cannot then trump the strong case of obviousness. [00:20:35] Speaker 00: And this Court has held that in many other contexts, and that holding would be sufficient to affirm [00:20:40] Speaker 00: even if you disagreed on copying. [00:20:43] Speaker 00: But I don't think you can, or should, because Cisco's record on copying is this sort of amorphous stuff. [00:20:52] Speaker 00: And what the board did was hold quite properly that the CLI, well, that was a copyright case that they lost, that that doesn't mean that the product and the specific functionality as claimed was copied, which is true. [00:21:08] Speaker 00: It's their burden to prove nexus [00:21:11] Speaker 00: between this amorphous stuff about copying and the claims. [00:21:15] Speaker 00: And they failed to do it. [00:21:18] Speaker 00: And they're stuck with their failure to do it. [00:21:19] Speaker 00: They chose to try to rely on that amorphous type of evidence. [00:21:24] Speaker 00: And the board properly held them to that and said that's not sufficient. [00:21:28] Speaker 00: The point on long-felt need, there's only two issues I want to talk about there. [00:21:34] Speaker 00: One is, and on copying, Cisco tries to suggest that the board's discussion on this [00:21:41] Speaker 00: is insufficient. [00:21:43] Speaker 00: But the board took the step here, which this Court has blessed in the in the Pace case of February 1, of saying, we adopt petitioners' reasoning as our own. [00:21:56] Speaker 00: That's the quote from the board, as our own. [00:21:59] Speaker 00: And so the mere fact that they didn't copy and paste the petitioners' arguments which they cited, they cited the pages that they were adopting, quote, as our own, [00:22:10] Speaker 00: The fact that they didn't copy and paste it into their opinion is irrelevant under PACE. [00:22:16] Speaker 00: And so Cisco's complaint that the words used in the actual board opinion were not as effusive as they would like falls squarely under PACE, which says it is permissible if the board does say we adopt it as our own. [00:22:31] Speaker 00: They did so here. [00:22:32] Speaker 00: So all of their complaints about so-called inadequate discussion [00:22:39] Speaker 00: fall by the wayside in that case. [00:22:40] Speaker 04: What page did they say we adopted? [00:22:44] Speaker 00: I believe it was 24, but look. [00:22:47] Speaker 00: 22. [00:23:03] Speaker 00: Appendix 22. [00:23:06] Speaker 00: And that is the bulk of Cisco's criticism [00:23:09] Speaker 00: of the board's reasoning. [00:23:10] Speaker 00: And that wipes away the bulk of their criticism about that issue. [00:23:15] Speaker 00: With regard to Longfellow's need, there is only one other issue. [00:23:17] Speaker 04: This is about the combination of Amara and Colbert. [00:23:23] Speaker 00: No, it's about the whole issue of obviousness. [00:23:31] Speaker 00: I don't think there's any question about that. [00:23:35] Speaker 00: And the board cites to various portions [00:23:38] Speaker 00: of the petition in supporting various of the arguments throughout the discussion. [00:23:44] Speaker 00: And so if you look at page 22, we have reviewed the petition, patent owner's response, and petitioner's reply, as well as relevant evidence that sufficiently establishes petitioner's contentions for claims and lists the claims. [00:24:06] Speaker 00: And we adopt Petitioner's contentions discussed below as our own. [00:24:10] Speaker 00: So it's not limited to combination of core builder. [00:24:12] Speaker 00: It's they've talked about the position on the claims. [00:24:15] Speaker 00: I think that's that's very, very clear. [00:24:17] Speaker 00: With respect to long felt need, Cisco argues that there's an inconsistency in the board's approach. [00:24:24] Speaker 00: And I think that's exactly wrong. [00:24:27] Speaker 00: The inconsistency was Cisco. [00:24:28] Speaker 00: The board correctly pointed out that the claims require protection at the control [00:24:34] Speaker 00: And the board could directly pointed out that Cisco's evidence supported evidence of a long quote unquote long felt need was not at the control plane level where it should be, but at a general denial of service level. [00:24:46] Speaker 00: And they argued that that was not a sufficient nexus to the claims, which is true. [00:24:50] Speaker 00: And what the board then did to say, it was also by the way, this 2013 article that Cisco submitted, which shows there were still denial of service problems at the general level. [00:25:02] Speaker 00: shows that even under Cisco's position, at the general level, it hadn't been solved. [00:25:06] Speaker 00: But what the board was correctly pointing out is that there was an inadequate amount of information that Cisco had provided to support its burden on the long-felt knee. [00:25:16] Speaker 00: That's all the board held. [00:25:17] Speaker 00: It's clearly true. [00:25:19] Speaker 00: I'd now like to turn to claim construction, if I may. [00:25:22] Speaker 00: So claim construction, the question is whether the control plane packets have to have normal port services in addition to regular port services. [00:25:32] Speaker 00: additional control plane support services. [00:25:35] Speaker 00: And that, I would submit, is clearly inconsistent with the claims and certainly inconsistent with the specification. [00:25:42] Speaker 00: But the starting point is that the standard here for the board is broadest reasonable interpretation. [00:25:48] Speaker 00: Cisco's interpretation is undeniably narrow. [00:25:52] Speaker 00: It's narrower because it adds a requirement that those control plane packets get an additional type of service rather than merely having control plane services. [00:26:00] Speaker 00: So that's narrower. [00:26:02] Speaker 04: And the question then is... Isn't that what the specification says then? [00:26:05] Speaker 00: I disagree completely. [00:26:06] Speaker 00: So the specification, and it starts with the point that your honor raised, at column 3, line 35, typically, typically certainly connotes, not always, that's not the end though, that's merely the starting point, and that is inconsistent with an argument that it has to always have normal port services applied. [00:26:25] Speaker 00: Column 2, line 63 to 66 squarely talks about a particular embodiment [00:26:31] Speaker 00: where a port has no configuration at all yet, and says in that context, if a port has no configuration, there are no normal port services. [00:26:39] Speaker 00: That's the whole point. [00:26:39] Speaker 00: It hasn't been configured. [00:26:41] Speaker 00: That's in the spec at column two. [00:26:43] Speaker 00: And so that's a situation where control plane packets coming through that un-configured port described in the specification would have no normal port services, would get control plane port services, and that's squarely inconsistent with Cisco's construction. [00:27:00] Speaker 00: Column 6, lines 24 to 32 talks about this administrator access during a denial of service attack. [00:27:07] Speaker 00: And that's referencing column figure 5 and also column 6. [00:27:12] Speaker 00: That's a context in which there's a specific port which the administrator is going through, which doesn't receive any of the normal port services. [00:27:19] Speaker 00: And if you look at column 5, rather figure 5, I'm sorry, figure 5 is what that's describing. [00:27:29] Speaker 00: Figure five says allow trusted host traffic, for example. [00:27:37] Speaker 00: And that's what column six is describing. [00:27:40] Speaker 00: As you're going to allow that traffic, regardless of whatever those port services are that may limit the rate or anything else, because it's, for example, an administrator who wants direct access during a dialed service attack, which makes complete sense. [00:27:53] Speaker 00: You wouldn't want the administrator subject to those port services when he's trying to access the control plane to fix it. [00:28:00] Speaker 00: And that's exactly what that's talking about in column six. [00:28:03] Speaker 00: Column seven, lines 46 to 56, also inconsistent with Cisco's construction, because it's talking about in the context of figure five, that the packets would go to the control plane, quote, without rate limits. [00:28:17] Speaker 00: So the rate limits are the normal port services, not applicable to those control plane packets, because you want the administrator to be able to reach the control plane. [00:28:26] Speaker 00: But if there's any question about all of that, [00:28:29] Speaker 00: It's removed by column six, column six lines 36 to 42. [00:28:34] Speaker 00: And column six lines 36 to 42 could not be cleared. [00:28:39] Speaker 00: And I'll read it. [00:28:42] Speaker 00: Thus, the user is afforded significant control over the flow of traffic destined to the control plane, just as if the control plane were a hardware interface. [00:28:53] Speaker 00: Since control plane destined packets will invoke [00:28:57] Speaker 00: only control plane services, transit traffic and system performance is minimally impacted. [00:29:04] Speaker 00: Since control plane destined traffic will invoke only control plane services, there is no way to reconcile that statement with their contention that the claim must require normal port services in addition to control plane services being applied to those packets destined for the control plane. [00:29:23] Speaker 00: Impossible to reconcile. [00:29:25] Speaker 00: And that's the specification. [00:29:27] Speaker 00: It's also inconsistent with their own experts' testimony in the ITC, who read that claim language and said, I don't see it meaning that you have to apply normal port services to control plane packets. [00:29:38] Speaker 04: Do you know what it means to have services applied to ports? [00:29:44] Speaker 00: I think the normal, as I understand it, is that a port will be configured so that, for example, a rate limit will be applied. [00:29:53] Speaker 00: that rate limit is exceeded, that's a typical example of a normal port service. [00:29:58] Speaker 00: If that rate limit is exceeded, it will stop additional flow of those packets. [00:30:03] Speaker 00: It might be also, we're gonna block packets from this source because we deem it to be unreliable. [00:30:09] Speaker 00: That's a port service. [00:30:11] Speaker 00: That's a port service being applied to a port, but you could have ports that are not configured, and therefore none of that is happening, because it hasn't been configured to happen. [00:30:19] Speaker 00: That's what column two teaches us. [00:30:21] Speaker 00: And column six is square, [00:30:23] Speaker 00: telling us that in many situations, control plane packets will only, only invoke control plane services. [00:30:33] Speaker 00: Cisco's evidence is primarily figures four and six. [00:30:36] Speaker 00: The specification is unambiguous at columns four and eight, describing those figures, saying these are steps that may be performed. [00:30:46] Speaker 01: Not required. [00:30:48] Speaker 00: Thank you. [00:30:52] Speaker 01: So we'll add a couple more minutes to your time to even out here. [00:30:57] Speaker 02: Thank you, Chief Judge Prost. [00:30:59] Speaker 02: Starting on the issue of secondary considerations and copying, I think it's very telling. [00:31:03] Speaker 02: Mr. Powers does not attempt to defend the board's reasoning here in terms of its analysis of copying. [00:31:11] Speaker 02: And of course, we're here under APA review. [00:31:13] Speaker 02: And it's not enough for the board to just simply say, I agree or I disagree, as this court has made clear in a number of recent cases, like ICON Health, [00:31:22] Speaker 02: in Renewvasive and recently Google versus Intellectual Ventures, the board has to explain why it agrees and why it disagrees. [00:31:29] Speaker 02: And the point, by the way, that he points to on page 22, where the board purports to adopt their initial contentions, well, none of that goes to any of the issues of copying, because none of the stuff about copying would have been part of Arista's initial contentions. [00:31:43] Speaker 02: The suggestion from Arista that secondary considerations are just relevant to the issue of motivation to combine [00:31:49] Speaker 02: is unsupported by case law. [00:31:51] Speaker 02: It obviously goes to the ultimate question of obviousness. [00:31:54] Speaker 02: And what you have here is the board engaging in a legally defective analysis. [00:32:00] Speaker 02: It did not even consider the inference, as this court has articulated. [00:32:05] Speaker 02: When a party has access to the technology, there is a strong inference of copying that can be drawn. [00:32:12] Speaker 02: The board does not address this. [00:32:15] Speaker 02: And if you accepted the board's analysis here without sending it back for them [00:32:18] Speaker 02: to have to look at this more carefully, it would mean that you could never prevail on showing copying without direct evidence of copying the feature itself, that the notion of circumstantial evidence would be gone. [00:32:31] Speaker 02: Here, we're not dealing with amorphous evidence of copying. [00:32:34] Speaker 02: We are talking about evidence of copying the very feature names, the very command line expressions used by Cisco to implement the features of this very patent. [00:32:46] Speaker 02: And if the board applies a proper analysis to copying, it will lead to a different analysis. [00:32:52] Speaker 02: Now, coming back to the question of claim construction, Mr. Powers makes a lot of arguments today, none of which, except for the last one, are in his brief. [00:33:03] Speaker 02: He does refer to the argument about column 6, lines 37 to excuse me, yeah, lines 39 to 45, [00:33:16] Speaker 02: And we address this in our brief. [00:33:17] Speaker 02: He latches on to the first of the two sentences there. [00:33:21] Speaker 02: And we said that it is clear there's a misplaced modifier. [00:33:23] Speaker 02: When you look at the second sentence, it defines what's going on in the first sentence. [00:33:28] Speaker 02: It says, that is, transit packets will not invoke control plane port services, but will continue to invoke normal input and output port services. [00:33:40] Speaker 02: That's all this sentence is saying. [00:33:42] Speaker 02: It's not that the control, what enhances [00:33:45] Speaker 02: The features here is not that control plane port services don't, excuse me, that control plane destined packets don't get normal port services. [00:33:55] Speaker 02: What enhances the operation is that normal transit packets don't also get the control plane port services. [00:34:03] Speaker 02: That's one of the things that the ART was improving. [00:34:07] Speaker 02: Now, in terms of some of the new arguments that we heard today, the point about column six and figure five, [00:34:13] Speaker 02: Ultimately there, all this is saying is that the administrator can apply different kinds of control plane port services if you're coming from a trusted host. [00:34:22] Speaker 04: Just going back to column six, are you saying that the word only is a typo and we shouldn't read it to be there? [00:34:30] Speaker 02: I'm saying that what this is trying to say is that the control plane port services, excuse me, the control plane destined packets [00:34:40] Speaker 02: are the only packets getting control plane port services. [00:34:46] Speaker 04: The normal... The only packets that get control plane services are the control plane packets. [00:34:53] Speaker 02: That's exactly right. [00:34:54] Speaker 02: And I realize that it is not a artful sentence, but when you look at the very next sentence, it very, very clearly explains exactly what it means. [00:35:05] Speaker 02: And that is consistent, of course, with everything else in the specification, [00:35:09] Speaker 02: And there is no reason to then ignore, by the way, the plain meaning of element B, which isn't modified to say some packets. [00:35:18] Speaker 02: And there was no response to the fact that if you read B that way, you'd have to read D2 that way. [00:35:23] Speaker 02: And if you read D2 that way, that means some of the packets going to the brain, to the control plane, are not going to get control plane port services. [00:35:30] Speaker 02: And that makes no sense. [00:35:32] Speaker 02: These need to be read consistently. [00:35:34] Speaker 01: Thank you, Chief Judge Kors.