[00:00:00] Speaker 01: This is WSOU Investment versus Arista Networks, 2023, 2231. [00:00:26] Speaker 00: Good morning, Your Honours. [00:00:27] Speaker 00: Brett Mangrum for the Appellant, Brazos. [00:00:31] Speaker 00: This appeal concerns primarily two claim limitations. [00:00:34] Speaker 00: The first limitation is a database maintaining IP multicast snooping information, and the focus of the dispute was on maintaining. [00:00:44] Speaker 00: That first limitation applies to all but three of the challenge claims. [00:00:49] Speaker 00: Not all of the claims are challenged in the patent. [00:00:52] Speaker 00: The second limitation concerns [00:00:55] Speaker 00: The universal requirement recited in all of the challenge claims, that's 1 through 6 and 12 through 17, that the multicast index, a different claim element, must be determined locally at the primary aggregation switch. [00:01:09] Speaker 00: For both limitations, the board erred in conflating together distinct claim terms. [00:01:14] Speaker 00: In fact, terms that appear in the exact same claim. [00:01:17] Speaker 00: So as I mentioned, I'm going to start with the first limitation, the database maintaining IP multicast snooping information. [00:01:24] Speaker 00: Claim One qualifies a chassis management module as the element that's performing. [00:01:31] Speaker 00: I'm sorry. [00:01:32] Speaker 00: It qualifies a chassis management module as storing the snooping information within the database. [00:01:38] Speaker 00: In an antecedent limitation, the snooping information is introduced as being maintained by the database. [00:01:45] Speaker 00: And the use of maintaining and storing in the same claim gives rise to the rebuttable presumption that those terms connote different meanings under the patent. [00:01:55] Speaker 00: And nothing in the intrinsic evidence rebuts that presumption. [00:01:59] Speaker 00: On appeal, Arista actually admits that storing and maintaining connote different meanings. [00:02:04] Speaker 00: We agree. [00:02:05] Speaker 00: And that's been our argument from the beginning. [00:02:08] Speaker 00: The 447 patent consistently used the words maintain, maintaining, or maintains [00:02:13] Speaker 00: to describe a certain set of operations. [00:02:16] Speaker 00: And it separately uses storing or stores or store to describe different operations. [00:02:20] Speaker 00: And there's never an instance where you have those two terms used interchangeably. [00:02:25] Speaker 00: I also want to find out the emission from claim one, independent claim one. [00:02:29] Speaker 00: of the limitation, a database maintaining IP multicast snooping information, is rendered more important by its omission from the only other independent claim at issue. [00:02:41] Speaker 00: And that is independent claim 15. [00:02:43] Speaker 00: So there's two independent claims at issue. [00:02:45] Speaker 00: The first separately requires a database maintaining IP multicast snooping information. [00:02:51] Speaker 00: The second independent claim doesn't have that limitation. [00:02:53] Speaker 00: And to suggest when claim one says, [00:02:56] Speaker 00: you need to maintain IP multicast scooping information. [00:02:59] Speaker 00: The database is the element that does it. [00:03:01] Speaker 00: And a separate element in claim one says storing, that element stores the IP multicast scooping information in the database. [00:03:10] Speaker 00: Our position has always been those are two different operations attributed to two different claim elements, and both must be shown. [00:03:16] Speaker 00: The problem that we identified from the outset, even before trial was instituted, was the original petition says, [00:03:23] Speaker 00: You're maintaining because you store. [00:03:25] Speaker 00: They didn't separately address maintaining as a distinct operation, not only in the sense of the word maintaining, but the fact that it's attributed to a different claim element. [00:03:33] Speaker 03: So in your view, what is the difference in this patent between maintaining and storing? [00:03:37] Speaker 00: Sure. [00:03:38] Speaker 00: I think the most helpful description in the patent is figure eight. [00:03:42] Speaker 00: And that's where the word maintaining and maintain appears. [00:03:46] Speaker 00: The description at the highest level of maintaining in terms, it describes maintaining in terms of deleting or flushing outdated IP multicast snooping information and this is done to facilitate synchronization between separate aggregation switches. [00:04:03] Speaker 00: So you have a local one where the database resides and that's the one that's maintaining. [00:04:09] Speaker 00: You have a remote aggregation switch and to synchronize [00:04:12] Speaker 00: when you delete or flush outdated information you maintain. [00:04:15] Speaker 00: That's what the description says. [00:04:17] Speaker 00: Let me read to you. [00:04:19] Speaker 00: When Figure 8 is introduced, the description of Figure 8 in the written description, it says, quote, Figure 8 illustrates a schematic block diagram of an embodiment for maintaining the map. [00:04:30] Speaker 02: Where exactly are you reading, if you could give us? [00:04:32] Speaker 00: Sure. [00:04:32] Speaker 00: That is on column 16, lines 38 through 40. [00:04:35] Speaker 00: Right. [00:04:38] Speaker 00: Go ahead. [00:04:40] Speaker 00: And I'm just reading those two lines. [00:04:42] Speaker 00: The word maintaining is used in describing what figure eight is showing. [00:04:47] Speaker 00: And the very next line states, and this is a column 16, lines 41 through 44, when a MAC address in the MAC HDI forwarding table has not been updated during the aging time, the entry will be deleted or flushed from the table, end quote. [00:05:02] Speaker 00: If you go down two lines 48 through 50 of the same column, [00:05:05] Speaker 00: It describes a mechanism used to decide when to delete or flush obsolete information. [00:05:12] Speaker 00: And then in the following column, column 17, lines 26 to 27, you again have this theme that similarly describes the maintaining of figure 8 as involving deleting an entry from the MAC HDI forwarding tables. [00:05:26] Speaker 00: That's 26 to 27. [00:05:29] Speaker 00: Now, on appeal, Arista attempted to discount figure 8 and its description as irrelevant to maintaining. [00:05:34] Speaker 00: And their argument was, well, it's a process for keeping MAC forwarding tables synchronized. [00:05:40] Speaker 00: But there could be no question that the word maintaining and maintains appears in the description of figure 8. [00:05:44] Speaker 00: And therefore, it certainly is relevant to what maintaining means in the context of the patent. [00:05:49] Speaker 00: But I also want to draw your honest attention to there's a parenthetical that appears later on. [00:05:53] Speaker 00: It's at column 19, lines 38 to 41. [00:05:58] Speaker 00: And there, in a parenthetical, the patent says, it describes, let me just read the statement. [00:06:07] Speaker 00: The IP multicast snooping information can be transferred between the aggregation switches using proprietary messages via the interchassis IPC, shown in Figure 8. [00:06:18] Speaker 00: So the argument that figure eight has nothing to do with IP multicast snooping information that's something else is defeated by what the patent says about figure eight. [00:06:27] Speaker 00: The patent says figure eight is all about IP multicast snooping information. [00:06:31] Speaker 00: It just describes it at a very detailed level. [00:06:33] Speaker 00: And it describes it in the context of maintaining. [00:06:36] Speaker 00: And so our position has always been that maintaining does not mean storing. [00:06:41] Speaker 00: Instead, in fact, in some ways, it's just the opposite, because maintaining refers to [00:06:47] Speaker 00: The way in which switches that are not physically connected through communication but physically remote from one another is the way in which they coordinate when they delete or flush obsolete information. [00:06:59] Speaker 00: Now, to be fair, we didn't offer an affirmative construction of maintaining below. [00:07:05] Speaker 00: No party did, and the board didn't ever affirmatively construe maintaining. [00:07:09] Speaker 00: Instead, our focus was, when we looked at the petition and saw that maintaining was completely omitted from the arguments, [00:07:15] Speaker 00: It was sufficient for us to say, well, maintaining is not storing. [00:07:18] Speaker 00: And you say maintaining is storing. [00:07:20] Speaker 00: And so we've always offered and relied upon a negative construction that if maintaining isn't storing, then the petition is deficient on its face for failing to address a distinct claim limitation. [00:07:31] Speaker 00: However, if the case should be remanded to at least have the board consider an affirmative definition to clarify the record, [00:07:40] Speaker 00: then that would be a basis to remand. [00:07:42] Speaker 00: But our position would be, even on remand, the petitioner, Arista, will necessarily, by law, be restricted to the argument set forth in its petition. [00:07:51] Speaker 00: And in its petition, it ignored that limitation. [00:07:54] Speaker 00: So it's kind of stuck. [00:07:55] Speaker 00: I mean, even if it wants to describe what maintaining means, the petition doesn't give an argument. [00:08:00] Speaker 02: I can understand how you could argue that you can have storing without maintaining. [00:08:08] Speaker 02: Explain to me how you can have maintaining without storing. [00:08:12] Speaker 00: Well, what's being maintained is a type of information, right? [00:08:17] Speaker 02: Well, it's information. [00:08:18] Speaker 00: Right, right. [00:08:18] Speaker 00: It's a type of information. [00:08:19] Speaker 02: If it's being maintained, then it's being stored while it's being maintained, right? [00:08:24] Speaker 00: Well, I think that's the problem with conflating those two terms. [00:08:27] Speaker 00: You have to understand what maintaining means in the context of the patent. [00:08:29] Speaker 00: And maintaining is a process by which you decide whether or not information in a database needs to be deleted or perched, because it's not [00:08:37] Speaker 00: It's obsolete. [00:08:40] Speaker 00: And you want all these switches to have that same type of timing in terms of when information is deleted or purged. [00:08:46] Speaker 00: And so maintaining refers to a coordination of deletion. [00:08:52] Speaker 02: But to go back to my question, maintaining also implies, it seems to me, storing. [00:08:59] Speaker 00: Well, maybe as a prerequisite, but not in terms of its scope. [00:09:03] Speaker 00: If you're deleting something from a database, it has to be there to be deleted. [00:09:08] Speaker 00: But our position is, when you store something, you're not maintaining it. [00:09:11] Speaker 02: That was my question. [00:09:14] Speaker 02: I can understand storing without maintaining, but not maintaining without storing. [00:09:18] Speaker 00: That's fair as long as the court recognizes that they are different operations. [00:09:24] Speaker 00: They're being performed by different elements. [00:09:26] Speaker 00: And the database job is to say, should this information be deleted? [00:09:31] Speaker 00: And if it is, it deletes it. [00:09:32] Speaker 00: And communications are sent to other switches saying, we're deleting, and you should too. [00:09:37] Speaker 00: And whereas storing is simply what the chassis management module does. [00:09:41] Speaker 00: It's a different element. [00:09:42] Speaker 00: to put information into the database. [00:09:45] Speaker 00: So we submit that those are distinct limitations attributed to different elements and both need to be shown and Arista felt to do that. [00:09:55] Speaker 00: The board also erred in its final decision in saying it need not address patent owner's argument that maintaining is distinct from storing. [00:10:03] Speaker 00: The board failed to recognize that we raised that argument. [00:10:05] Speaker 00: So ARISTA offered two alternative theories for the database limitations. [00:10:10] Speaker 00: But our argument that storing doesn't mean maintaining applied equally. [00:10:13] Speaker 00: It was a universal argument to both theories. [00:10:15] Speaker 00: The board said, well, we're only going to address, we're going to find this argument reasonable and persuasive, and therefore we're not going to address patent owners' argument that destroying doesn't mean maintaining. [00:10:24] Speaker 00: That was error, because our argument applied equally to both, and we addressed that same argument in addressing both of the alternative theories that Arista raised. [00:10:33] Speaker 00: I also want to point out that the argument's not new, and the board found that it wasn't new. [00:10:37] Speaker 00: Arista argues on appeal that we never offered a negative construction, and the board disagreed, and the board recognized that was our position even before trial. [00:10:44] Speaker 00: Now, Arista's belated attempt on appeal to rewrite maintaining as keeping or holding should not only be rejected as untimely, unexplained, and unsupported, but it would also exclude preferred embodiments. [00:10:56] Speaker 00: As I mentioned, maintaining is the process of deletion, right? [00:11:00] Speaker 00: It's the process of coordinated deletion. [00:11:04] Speaker 00: To the extent the idea is well you keep in hold indefinitely That would that would defeat the deferred embodiment says at specific times and under specific conditions we coordinate flushing I'm going to reserve the remainder of my time for the chassis management module determining unless there's other questions Okay, so on this issue as well. [00:11:25] Speaker 00: It's a problem with the board at least implicitly [00:11:28] Speaker 00: interpreting, determining to mean using. [00:11:31] Speaker 00: Our position has always been that because the 447 patent equates determining to creating, and because the 447 patent requires the multicast index to not only be created, but each claim also says once it's created or determined, it's used. [00:11:53] Speaker 00: So our problem with the board's findings is the board said, look, [00:11:56] Speaker 00: We find that the multicast index was shown to be determined. [00:12:00] Speaker 00: It would have been obvious because it's used to build something else. [00:12:05] Speaker 00: That finding is problematic for several reasons. [00:12:07] Speaker 00: Again, all claims require that the multicast index be locally determined at the primary switch. [00:12:14] Speaker 00: But as a separate and distinct limitation, that same multicast index must then be used universally by all switches. [00:12:20] Speaker 00: So to suggest that determining means using would conflate those two distinct limitations. [00:12:25] Speaker 02: What exactly do you mean by determining the term? [00:12:28] Speaker 00: Yeah, the full-form patent is, in this instance, it is helpful in the sense that there are two different words that are used as synonyms. [00:12:36] Speaker 00: For maintaining, it was always the word maintain. [00:12:38] Speaker 00: But for determining, the full-form patent uses the word creating, and that's at 23, lines 14 through 18. [00:12:45] Speaker 00: And it also uses on the same column, lines 57 through 59, the word computing. [00:12:49] Speaker 00: Now, it's described in a lexicographic statement, and the parties don't dispute that this is lexicography. [00:12:58] Speaker 00: It is described as a unique identifier of an ingressing multicast flow, and that it is determined based on three input elements. [00:13:07] Speaker 00: One, the IP source, two, the destination addresses, and three, [00:13:11] Speaker 00: the Ingress VLAN. [00:13:12] Speaker 02: But again, that uses the term determined. [00:13:14] Speaker 02: And what I'm really trying to understand is, what is your position as to what that term itself means? [00:13:20] Speaker 00: Well, so again, we offer the negative construction below that it's not using. [00:13:25] Speaker 00: And no party offered an affirmative. [00:13:27] Speaker 02: What do you think it is? [00:13:28] Speaker 00: Yeah, our position is creating. [00:13:31] Speaker 00: So it doesn't exist. [00:13:34] Speaker 00: It's not something that's received from an external source. [00:13:37] Speaker 00: It is actually locally created. [00:13:39] Speaker 00: It's computed, it's calculated based off the input elements. [00:13:42] Speaker 00: And before that creation process, it actually doesn't even exist. [00:13:46] Speaker 00: So it is the primary switch that comes up with this new multicast index, and then it shares it elsewhere. [00:13:54] Speaker 00: It says, here's what we're going to use. [00:13:55] Speaker 00: And here's how it's important. [00:13:57] Speaker 00: We want to share this elsewhere so there's a coordination between switches. [00:14:00] Speaker 00: Now, the problem with the theory set forth in the petition is that... You're well into your bottle time. [00:14:06] Speaker 00: Sure, yeah. [00:14:07] Speaker 01: Do you want to save it or continue? [00:14:09] Speaker 00: I'm going to continue it on this point. [00:14:12] Speaker 00: All right. [00:14:13] Speaker 00: The problem with the theory in the petition was they say, they acknowledge, even their own declarant acknowledges, that what they rely on for the multicast index is received from an external source. [00:14:25] Speaker 00: and then it's only used locally. [00:14:28] Speaker 00: And our argument from the beginning is, because the claim language distinguishes receiving things, because that's also other elements in the claims, from an external source of determining, the determining requires local creation, and that kills their argument. [00:14:40] Speaker 00: They've also newly argued on appeal, a new claim instruction, improperly belated introduction on appeal, that determining means assigning, and opposition has always been, that can't be the case either, because when you look at dependent claim 13, [00:14:54] Speaker 00: There's an additional limitation for the same element, the chassis management module. [00:14:58] Speaker 00: And that element is described, the chassis management module is further described in 13 as allocating a multicast index. [00:15:07] Speaker 00: Once it's created, you allocate it to a multicast flow. [00:15:10] Speaker 00: And so to suggest, well, the primary switch receives from an external source the index, and then it allocates it or assigns it to a multicast flow that's determining, that's defeated by the claim language itself. [00:15:22] Speaker 00: And I've used my time, so thank you. [00:15:25] Speaker 01: Well, you've just about consumed all of your time, entitling you to no rebuttal time, but we'll give you a minute. [00:15:31] Speaker 00: Yes, sir. [00:15:31] Speaker 00: Thank you, Your Honor. [00:15:35] Speaker 01: Mr. Macher. [00:15:45] Speaker 04: Thank you. [00:15:46] Speaker 04: May it please the court, Hamid Macher for Appellee. [00:15:49] Speaker 04: The board's decision here is supported by more than substantial evidence, which is the standard of review. [00:15:54] Speaker 04: and it should be affirmed. [00:15:56] Speaker 04: The petition demonstrated why a person of ordinary skill would start with Sane and combine that with the teachings of Sue Bermanian to arrive at the challenge claims. [00:16:05] Speaker 04: The combination, the motivations, and the understanding of a person of ordinary skill was supported by the unrebutted declaration of Dr. Black. [00:16:13] Speaker 04: Brazos makes two primary arguments on appeal, as you heard, one toward the database limitation, one towards the multicast index limitation, putting aside that neither argument was timely raised below. [00:16:24] Speaker 04: Neither argument has merit either. [00:16:27] Speaker 04: The fundamental problem with Brazos' arguments is twofold. [00:16:31] Speaker 02: The board addressed the arguments on their merits, I take it. [00:16:37] Speaker 02: Given that, is your position that nonetheless, we could dismiss the arguments on the ground, that even though the board addressed them, that it could have thrown them out on waiver and should have or could we uphold the decision on waiver grounds? [00:16:57] Speaker 04: I think the decision could be uphold on numerous grounds. [00:17:00] Speaker 04: One. [00:17:01] Speaker 04: There's also waiver as well. [00:17:05] Speaker 04: And this court could find on a case-by-case basis that there was waiver. [00:17:08] Speaker 02: But when the board actually addresses the merits, I'm trying to get at a category of cases in which, OK, an argument isn't made. [00:17:14] Speaker 02: But the board says, OK, we're going to go ahead and address it on the merits. [00:17:18] Speaker 02: Can we then, within the construct of shennery, can we come in and say, ah, [00:17:25] Speaker 02: We don't have to reach the merits because this was not timely raised before the board, even though the board, in effect, waived the waiver argument. [00:17:35] Speaker 04: I don't think that the board waived the waiver argument. [00:17:37] Speaker 02: But even though the board went to the merits, you understand the point I'm trying to get at. [00:17:42] Speaker 02: I want to see if we can effectively rule on waiver notwithstanding the board's having [00:17:50] Speaker 02: what case says we can do this. [00:17:52] Speaker 04: I disagree that the board did that. [00:17:55] Speaker 04: I think, as my opponent here is pointing out, well, there's arguments that they raised, and they're complaining that they weren't addressed. [00:18:01] Speaker 04: And so to the extent the board was segregating out, as it should do under its discretion when there's these new arguments, it did that. [00:18:08] Speaker 04: This court can then look back and say, either that was correct, or yes, we hold that waived as well. [00:18:14] Speaker 04: I point the court to the Fingen case that we cited, where there's a very similar situation. [00:18:19] Speaker 04: Which case? [00:18:20] Speaker 00: I'm sorry. [00:18:21] Speaker 04: Fingen that we cited? [00:18:22] Speaker 04: Yeah. [00:18:22] Speaker 04: The patent owner in that case said, hey, we raised this argument in surreply. [00:18:27] Speaker 04: And this court said, sorry, it's been waived and disregarded that argument. [00:18:32] Speaker 02: That was a situation where the board didn't address the argument, I take it. [00:18:36] Speaker 04: Yeah. [00:18:38] Speaker 04: didn't say, hey, you've waived it, and I'm not going to address it. [00:18:40] Speaker 04: But the board didn't address the argument. [00:18:42] Speaker 02: With respect to arguments that were addressed by the board, again, my question is, can we override, in effect, the board's decision to decide the issues? [00:18:52] Speaker 04: I think that would be under an abuse of discretion review, if that was brought to you in that way. [00:18:58] Speaker 04: Now here, on the merits, the board got it right anyway. [00:19:01] Speaker 04: And so there's no reason to reach that issue. [00:19:03] Speaker 04: But what Patent Owner does in its arguments and fundamentally misses in attacking the board's decision is two problems. [00:19:11] Speaker 04: One, it ignores the arguments that it actually did make in its Patent Owner response, which had to be addressed by the board. [00:19:18] Speaker 04: And it ignores how one of ordinary skill in the art understands these claims. [00:19:22] Speaker 04: and understands how these switches work. [00:19:24] Speaker 04: The only evidence of what one of ordinary skill in the art understands came from Dr. Black. [00:19:29] Speaker 04: And the board rightly credited that testimony as it was unrebutted. [00:19:33] Speaker 04: And so by relying on that testimony and the other evidence with the petition, the board's conclusion is supported by more than substantial evidence and should be firm. [00:19:41] Speaker 02: So do you see stored and maintained as being either A, distinct, B, overlapping, or C, effectively the same word? [00:19:56] Speaker 04: Yeah, I look at that as two sides of the storage coin. [00:19:59] Speaker 04: You have one entity. [00:20:01] Speaker 04: The claim is laid out as an apparatus claim, structural elements, chassis management module, and a database. [00:20:07] Speaker 04: So one of those structures is storing into [00:20:11] Speaker 04: the database, and the other one is the receptacle that then receives and keeps that information and holds onto it. [00:20:18] Speaker 04: I want to address the intrinsic evidence that my opponent is pointing to. [00:20:22] Speaker 04: And Figure 8, if you look through that discussion, it's well before the discussion of IP multicast snooping information. [00:20:30] Speaker 04: I heard him make a reference to IP multicast snooping information in the context of Figure 8. [00:20:35] Speaker 04: That's not there. [00:20:36] Speaker 04: Figure 8 is discussing unicast communications only. [00:20:39] Speaker 04: That's clear from the context. [00:20:41] Speaker 04: And this MAC and HDI forwarding tables, those are not multicast snooping information in that context. [00:20:48] Speaker 03: What about the column 19 parenthetical? [00:20:50] Speaker 03: Doesn't that tie IP multicast back to figure 8? [00:20:54] Speaker 04: It ties it back just for the sense of you're sharing information through an IPC message. [00:20:59] Speaker 04: and it's just saying, hey, we pointed it and showed you that in figure eight. [00:21:01] Speaker 04: It's not talking about the operations that are happening in figure eight. [00:21:04] Speaker 04: It's talking about sharing IP multicast snooping information. [00:21:08] Speaker 04: The discussion of the database starts in column 19 of the patent and is used interchangeably throughout until column 24. [00:21:17] Speaker 04: And in that discussion, the database is constantly referred to as a receptacle for storage and later retrieval. [00:21:24] Speaker 04: And so in what we're talking about in the context of the claims and looking at in the context of the intrinsic evidence, that's what it is. [00:21:31] Speaker 04: And maintaining here is the other side of the storage coin. [00:21:36] Speaker 04: And so if you look at the other evidence that they pointed to in their brief, for example, column 23, line 10 to 13, I think if you start a little earlier, line 4, you'll see what's happening there. [00:21:47] Speaker 04: And it's almost using storing and maintaining as synonyms there. [00:21:50] Speaker 02: What column again? [00:21:51] Speaker 04: I'm in column 23, starting at line 4. [00:21:53] Speaker 04: And what's being described there [00:21:57] Speaker 04: is with respect to a replication vector. [00:22:00] Speaker 04: But similar approach, which is chassis management module in the primary switch, stores information in its database, shares it across to the secondary switch, it's maintained in that database. [00:22:12] Speaker 04: And so there you see they're being used almost synonymously. [00:22:15] Speaker 04: And to the extent there is a difference here, it's really that the [00:22:20] Speaker 04: The maintaining is sort of giving you a little bit of emphasis on this continued storage. [00:22:28] Speaker 04: I point the court to the LA biomedical case. [00:22:32] Speaker 04: In that case, you had a very similar situation where we had two phrases in one claim, like we do here. [00:22:40] Speaker 04: Factually, also similar, in that the patent owner in that case was trying to add limitations, arguing that otherwise claim language obese or perfluous. [00:22:50] Speaker 04: And so same thing here, Brazos is saying there needs to be maintaining operations, even though operations isn't in the claim. [00:22:55] Speaker 04: And there, what this court said was that even though you have overlapping claim terms, [00:23:04] Speaker 04: That's not fatal, and it doesn't compel us to adopt an otherwise unsupported construction of the claims. [00:23:09] Speaker 04: And that's what we have here is overlapping terms from the two sides of the storage point. [00:23:16] Speaker 03: So the bottom line is your view is maintaining and storage overlap in meaning in this patent, correct? [00:23:23] Speaker 04: They overlap in meaning. [00:23:24] Speaker 04: I'm not saying that they're, sorry. [00:23:26] Speaker 03: And you're not saying they're complete synonyms. [00:23:29] Speaker 03: They have some overlap. [00:23:31] Speaker 03: Where did you take that position in the petition? [00:23:37] Speaker 04: And so, again, remember, procedurally, this argument was not raised at any time in the patent owner response. [00:23:43] Speaker 04: In the petition, we've applied the Plaint and Ordinary Meeting. [00:23:45] Speaker 04: And we think that the Plaint and Ordinary Meeting is sufficient here because maintaining, just like maintaining prices or maintaining files in a file cabinet, [00:23:54] Speaker 03: Includes holding on to this information and so that's shown in the petition is how the database is used Detail about how maintenance or overlap Is is new on appeal You would say that's okay because the argument is never presented by the patent owner [00:24:16] Speaker 04: It's not that it wasn't presented to the board. [00:24:20] Speaker 04: Procedurally, this claim construction argument was raised only in sir reply. [00:24:23] Speaker 04: Our first opportunity was an oral argument. [00:24:26] Speaker 04: We did raise that then. [00:24:28] Speaker 04: It's not in the joint appendix, but in the transcript. [00:24:30] Speaker 04: I'd point to pages 18 to 20. [00:24:32] Speaker 02: And what was the position you took? [00:24:34] Speaker 04: Exactly. [00:24:35] Speaker 04: What we've taken in our briefs here is that maintaining just means keeping it, holding onto it. [00:24:40] Speaker 04: It's the same position we have on appeal. [00:24:42] Speaker 04: We just were only able to raise it during oral argument, because the argument was raised for the first time in syrup line. [00:24:46] Speaker 02: So maintaining equals storing. [00:24:48] Speaker 04: Not that it equals storing, but maintaining is keeping that. [00:24:51] Speaker 04: I made the same argument, two sides of the same coin. [00:24:54] Speaker 04: You have the chassis management module storing into the database, and then the database holds on to that information or keeps it. [00:25:04] Speaker 04: OK. [00:25:07] Speaker 04: So I pointed to the LA Biomedical case and how that case explains that you can have overlap. [00:25:14] Speaker 04: I would also point to the Bancorp case. [00:25:15] Speaker 04: Both of these cases were not addressed by Brazos in reply. [00:25:18] Speaker 04: The Bancorp case, similarly, you had two phrases, one claim. [00:25:24] Speaker 04: And the court said, well, that gives rise to an inference, not a presumption, but an inference that there would be different meanings. [00:25:30] Speaker 04: In that case, the court said, [00:25:33] Speaker 04: That inference, however, is not conclusive, because it's not unknown for different words to be used to express similar concepts, or as it goes on later, in fact, to be synonyms. [00:25:43] Speaker 04: And that's, again, like the situation we have here. [00:25:45] Speaker 04: We have two sides of the storage coin, and it's OK from a claim construction perspective for that to be the case, especially when it's so supported by the intrinsic evidence. [00:25:55] Speaker 02: I want to remind the court what [00:25:58] Speaker 02: Do you want to talk about determining? [00:26:00] Speaker 04: If I may finish my point, then I will move to determining. [00:26:03] Speaker 04: The last point I wanted to make on the maintaining is that the intrinsic evidence is so supportive of our view. [00:26:15] Speaker 04: What the other side's position is, remember, is not that some other entity is maintaining. [00:26:21] Speaker 04: The database itself is maintaining. [00:26:23] Speaker 04: That's their argument. [00:26:24] Speaker 04: No one has disputed that. [00:26:26] Speaker 04: a database is an organized collection of data. [00:26:28] Speaker 04: It's not software instructions. [00:26:30] Speaker 04: It's not a program. [00:26:32] Speaker 04: There's nothing in the spec that would tell you a database itself is performing any maintenance operation, as they suggest that it should be. [00:26:39] Speaker 04: And so it's a completely unsupported construction. [00:26:42] Speaker 04: And frankly, it does not make sense. [00:26:43] Speaker 04: It's not consistent with that. [00:26:45] Speaker 04: And so the pointing to what happens in figure eight, I just don't figure eight in the discussion, it's not applicable here. [00:26:52] Speaker 04: And I don't think it should be considered applicable. [00:26:56] Speaker 04: Now, to your point, Judge Price, and I'll move to the multicast index limitation. [00:27:01] Speaker 04: Now, Bob's argument is that the evidence just says that the multicast index is coming from some external source. [00:27:08] Speaker 04: That's a made up fiction. [00:27:10] Speaker 04: I don't know where that comes from. [00:27:11] Speaker 04: There's nothing in the prior art. [00:27:13] Speaker 04: There's nothing in Dr. Black's declaration. [00:27:14] Speaker 04: There's nothing in the decision that says it's coming from an external source. [00:27:19] Speaker 04: What did the board say? [00:27:20] Speaker 04: The board said, that is on page 36 of the appendix of the opinion, [00:27:24] Speaker 04: In the combination Sane and Supermanian, the control plane determines the global multicast index in conjunction with determining the claim forwarding vector. [00:27:34] Speaker 04: The control plane of a Sane switch is local and internal to a Sane switch. [00:27:40] Speaker 04: The other side has never argued otherwise. [00:27:43] Speaker 04: in their patent owner response, this is 1182 to 83 of the appendix, they agree that Sane's control plane is in the Sane switch. [00:27:51] Speaker 04: And so that finding by the board was supported by the prior art that they cited, numerous paragraphs from Dr. Black's declaration that they cited. [00:28:01] Speaker 04: And so what's happening here is they're disagreeing with that factual finding and asking this court to reweigh the evidence. [00:28:08] Speaker 04: But on substantial evidence review, [00:28:10] Speaker 04: That's not what we do here. [00:28:12] Speaker 04: And so that factual finding was also well supported and should be affirmed. [00:28:17] Speaker 04: If there's no questions on that, I'll cede the rest of my time. [00:28:21] Speaker 01: Thank you, counsel. [00:28:24] Speaker 01: Mr. Mangrum has mentioned you elected to use all your time, but we'll give you one minute for a brief reply. [00:28:32] Speaker 00: Yes, I want to start with the end of what he talked about. [00:28:35] Speaker 00: He said Dr. Black said determining is not receiving from an external source. [00:28:40] Speaker 00: Look, he said the opposite. [00:28:41] Speaker 00: It's in paragraph 125 of his original declaration. [00:28:44] Speaker 00: He said the MGID was what they rely for for the index. [00:28:48] Speaker 00: It's distributed to all line cards. [00:28:50] Speaker 00: It's provided to all line cards. [00:28:51] Speaker 00: And then it's used. [00:28:52] Speaker 00: And he said the use is what's determining. [00:28:55] Speaker 00: The board found the same thing. [00:28:56] Speaker 00: I mean, the board summarized at page 44 of their opinion [00:29:02] Speaker 00: that the [00:29:04] Speaker 00: I'm quoting, petitioner contains that each of Subramanian chassis receives the MGID and EPI multicast indexes and uses these indexes to program the hardware, MGID, something else, with all ports. [00:29:18] Speaker 00: And the board said that that something else is building fording vectors. [00:29:21] Speaker 00: So the use, locally, the board attributed to building fording vectors, not to determining in the first place. [00:29:26] Speaker 00: And the board recognized the argument they raised. [00:29:29] Speaker 00: And their own declarant said is the, what they're alleging is the multicast index that was provided [00:29:34] Speaker 00: globally, and every switch is something it's receiving. [00:29:38] Speaker 00: Now I want to address the waiver. [00:29:41] Speaker 00: We addressed the waiver argument in our briefing, so I'm out of time. [00:29:43] Speaker 00: But I would just submit that, oh, it looks like it's coming up. [00:29:47] Speaker 00: For the waiver argument, the board found that for both the negative construction of storing is not maintaining and that determining is not using, the board recognized that that's been our argument from the beginning, even prior to trial. [00:30:03] Speaker 00: And when you look at how we address the maintaining limitation, just look at arguments in the panel in response. [00:30:08] Speaker 00: We circle the whole thing and say, what they're alleging is maintaining it a database is not maintaining a database. [00:30:15] Speaker 00: They say today that maintaining and storing overlap, in their appeal brief, they say it's not the same thing. [00:30:22] Speaker 00: They say today that a database maintaining is not a database maintaining. [00:30:26] Speaker 00: It's something else maintaining, or the database doesn't have to do it. [00:30:29] Speaker 01: Counsel, a red light is a red light. [00:30:31] Speaker 01: We have your argument and the case is submitted. [00:30:34] Speaker 00: Thank you.