[00:00:03] Speaker 06: Because I have a motion to make, I'm going to have to turn the gavel over to Judge Wallace for the morning, for the first part of the morning. [00:00:12] Speaker 05: Please proceed. [00:00:13] Speaker 06: I have the privilege of moving the admission of Glenn Chang. [00:00:20] Speaker 06: He's a member of the bar and is in good standing with the highest courts in both New York and New Jersey. [00:00:26] Speaker 06: I have knowledge of his credentials and am satisfied that he possesses the necessary qualifications. [00:00:32] Speaker 06: In fact, I have intimate knowledge of his credentials because he has been my law clerk for the last year. [00:00:38] Speaker 06: He has almost too many degrees, but that was a benefit to me. [00:00:46] Speaker 06: He is a graduate of the College of New Jersey Honors College, where he had a BS in biology. [00:00:52] Speaker 06: He has an MD from Rutgers New Jersey Medical School and a law degree from Rutgers School of Law. [00:01:01] Speaker 06: did all of those with great distinction. [00:01:03] Speaker 06: And he has served me with great distinction after having been both a practicing doctor and a practicing lawyer. [00:01:11] Speaker 06: So I've had the privilege of getting to know him and getting to know his beautiful family. [00:01:15] Speaker 06: His wife, Jew, and daughter, Gloria, are here with us today, too. [00:01:19] Speaker 06: I'm sure Gloria will never forget this moment. [00:01:21] Speaker 06: But at least I hope Glenn will never forget this moment. [00:01:26] Speaker 06: So with all of those credentials, I move his admission. [00:01:29] Speaker 05: I have a few questions. [00:01:31] Speaker 05: Doc, I've got this crick in my neck. [00:01:36] Speaker 05: Judge Trondell? [00:01:38] Speaker 05: I concur. [00:01:39] Speaker 05: Your motion is granted. [00:01:40] Speaker 06: But I should say that despite the fact that he's now being sworn in to the bar, that he actually had his first appearance before the court yesterday because we had a pro se applicant who could not speak English. [00:01:54] Speaker 06: And luckily, Glenn could translate the Mandarin Chinese for us. [00:01:59] Speaker 06: So both Judge Clevender and Judge Newman were very impressed with Glenn's appearance in the well yesterday. [00:02:06] Speaker 06: So welcome. [00:02:08] Speaker 06: Thank you. [00:02:09] Speaker 00: Please raise your right hand. [00:02:11] Speaker 00: Do you solemnly swear or affirm that you will appoint yourself as attorney and counsel of this court, up writing and according to law, and that you will support the Constitution of the United States of America? [00:02:22] Speaker 00: I do. [00:02:22] Speaker 00: Welcome to the bar of the United States Court of Appeal of the Federal Circuit. [00:02:25] Speaker 00: Thank you. [00:02:26] Speaker 00: You're welcome. [00:02:29] Speaker 06: Okay. [00:02:32] Speaker 06: With that happy event over, we now move to the argued cases for today. [00:02:39] Speaker 06: We have four argued cases on the calendar and two submitted cases. [00:02:48] Speaker 06: The first of the argued cases is docket number 151719, NetApp [00:02:55] Speaker 06: Inc. [00:02:55] Speaker 06: versus Electronics and Telecom. [00:02:58] Speaker 06: It is an appeal from the Patent Trial and Appeal Board. [00:03:05] Speaker 06: Mr. Kang, you want five minutes for rebuttal? [00:03:07] Speaker 06: Yes, ma'am. [00:03:08] Speaker 06: All right, you may begin. [00:03:12] Speaker 04: May it please the court. [00:03:14] Speaker 04: My name is Peter Kang, and I represent NetApp in this case. [00:03:17] Speaker 04: Your honors, there are three independent reasons why the board's decision in this case should be reversed. [00:03:24] Speaker 04: The board committed legal error in construing the claim term raid to require a system which forbids direct access between the host computers on one end and the disks of the raid on the other end, where that limitation does nowhere appear and has no support in the intrinsic evidence. [00:03:42] Speaker 04: And the only citation to the board in support of that limitation was attorney argument. [00:03:47] Speaker 03: Can I just ask? [00:03:49] Speaker 03: The way that I read this was that the claim construction [00:03:54] Speaker 03: didn't use language about direct access and we'll get back to that. [00:03:57] Speaker 03: It used language about single logical unit. [00:04:01] Speaker 04: Yes, Rob. [00:04:02] Speaker 03: Then on the application side, not the claim construction side, I'm pretty sure this is the way it was presented. [00:04:10] Speaker 03: by Dr. Conte in his declaration, the direct access comes up as an application issue, a familiar construction application divide, to which the substantial evidence standard would apply. [00:04:32] Speaker 03: So I'm not quite sure you're right in saying that the direct access was a piece of the claim [00:04:39] Speaker 04: As this court ruled in the Respironics opinion, which just issued last week, an application of a claim construction which adds a further limitation to the construction, in other words, a construction of the construction, is itself a claim construction. [00:04:53] Speaker 04: And that's exactly what happened here. [00:04:54] Speaker 03: Well, why is that what happened? [00:04:56] Speaker 03: I mean, the claim construction was done [00:04:59] Speaker 03: as I understand it, on the theory that the intrinsic evidence doesn't tell us a whole lot. [00:05:05] Speaker 03: So you have to look to what people in the art, in the extra patent world, would understand the term to mean. [00:05:14] Speaker 03: We think, the board said, it means a single logical unit. [00:05:18] Speaker 03: And then in applying it said, you don't have a single logical unit the way an expert would testify at a trial at the application stage after claim construction is done. [00:05:27] Speaker 03: to say as long as the host is able to get separate disk access, and I use here separate as a word like direct, then it's not a single logical file. [00:05:45] Speaker 04: There's nothing in the phrase single logical unit which implies or talks about that direct or separate access. [00:05:51] Speaker 03: That's what experts do in testifying about the application of a term that has been construed. [00:05:58] Speaker 03: I believe that these words read on the following and don't read on the following. [00:06:04] Speaker 04: And first, if I may point out, Dr. Conte's declaration of the testimony nowhere actually addresses this point. [00:06:11] Speaker 04: The cited testimony from Dr. Conte simply talks about the raid being a black box. [00:06:16] Speaker 04: Dr. Conte, there is no support in this record for any expert testimony that goes along the syllogism that Your Honor posited, that I am applying the single logical unit construction [00:06:27] Speaker 04: to understand it to mean that there cannot be this direct access. [00:06:30] Speaker 04: That's nowhere in the testimonial record. [00:06:33] Speaker 04: That's why the board simply cited to attorney argument in support of this construction. [00:06:39] Speaker 06: Again, I mean, I have to agree with Judge Taranto that there's a distinction between a construction and the application of that construction to your anticipation analysis, to an anticipation analysis. [00:06:53] Speaker 06: And isn't there other evidence in the record that the board could cite to conclude that [00:07:01] Speaker 06: that given the single logical unit construction and what that must mean that this doesn't anticipate because of the indirect versus direct analysis. [00:07:11] Speaker 04: So no your honor, to that point there is no evidence in the record that Hathorn actually forbids this kind of direct access. [00:07:18] Speaker 04: On page 53 of Patent Donors Brief they have a figure replicated from Hathorn which shows all the connections from the host computer to the disks always go through indirectly through the controllers in the middle. [00:07:30] Speaker 04: And so Hathorne does forbid exactly the kind of direct access that the board said has to be forbidden in the system. [00:07:37] Speaker 04: And so the board's decision on that issue doesn't stand up to substantial evidence. [00:07:42] Speaker 03: I'm sorry, what do you mean by Hathorne forbids the direct access? [00:07:50] Speaker 03: Is this because there's a disk specific controller sitting between the host and the disk drive? [00:07:58] Speaker 04: Yes, your honor. [00:07:59] Speaker 03: So why is it that IBM calls that direct access? [00:08:03] Speaker 04: The problem we have is we don't know what the board meant by lack on or forbidding direct access. [00:08:09] Speaker 03: It means what it meant is separate access, that the host gets separate access to each disk drive, and there may be a door in front of it, call it a disk-specific controller, but it's the one that gets to decide, that is entitled to decide, that the host wants to send something to a disk. [00:08:29] Speaker 04: And that's one of the problems with this application of this construction, in that the patent itself, the 346 patent, has that structure, has exactly that. [00:08:38] Speaker 04: The patent itself [00:08:39] Speaker 04: requires in the claim direct access between the hosts and the controllers. [00:08:44] Speaker 04: And so what the board has done is come up with an inconsistent, logically fallacious application of its construction, where the claim and the patent itself is that structure. [00:08:53] Speaker 04: So wait. [00:08:54] Speaker 06: So are you arguing that Hathorn requires direct access, or are you arguing that the board misinterpreted the patent that was challenged? [00:09:07] Speaker 04: As I understand, [00:09:08] Speaker 04: Your Honor, the board's logic, the claim forbids direct access between the hosts and the disks, and the board found that Hathorne does not forbid that direct access. [00:09:21] Speaker 04: If direct access means it can go via or through the controllers to get to the disks, then there is no individual access in Hathorne in any event, because Hathorne discloses that the disks shown in the figure is representative of a whole string [00:09:38] Speaker 04: of disks. [00:09:38] Speaker 04: And so there's no way to get at any one individual disk from the host. [00:09:42] Speaker 04: The host has no way to address which separate or individual disk it may want to, even if it tried. [00:09:48] Speaker 04: It's up to the controller in the middle to make that determination. [00:09:51] Speaker 03: I guess what I'm remembering is, just correct me if I'm wrong, you rely heavily on column seven of Hafthorne describing figure two, which figure three is just a variant of. [00:10:03] Speaker 03: So in this respect, it's the same. [00:10:06] Speaker 03: rely on lines that say the host can get to a particular disk in two ways. [00:10:13] Speaker 03: Either it can direct its information to that disk, that particular disk, again, by way of the disk-specific controller, or it can direct its information to its primary controller, and the primary controller can do the selection. [00:10:29] Speaker 03: And I took it your position on Hafthorn was that as long as that [00:10:34] Speaker 03: first option was available, then it is teaching a raid whose point, even under the claim construction adopted by the board, is that the host isn't doing the selecting of the disk. [00:10:52] Speaker 03: And it was the option of the primary host controller [00:10:56] Speaker 03: choosing disks that you were relying on. [00:10:59] Speaker 03: Did I misunderstand? [00:11:00] Speaker 04: The control there is not the disk controller. [00:11:03] Speaker 04: There's a network controller in Hathorne in the middle there that is actually performing the option. [00:11:08] Speaker 04: That's why the host computers cannot direct themselves which disk to get to in Hathorne. [00:11:14] Speaker 04: They have to go through the network controllers in the middle. [00:11:17] Speaker 04: And that's why there's no individual access to disks on the disk side of Hathorne. [00:11:22] Speaker 04: And that's why the board's decision here doesn't stand up to substantial evidence. [00:11:26] Speaker 04: Furthermore, going to the point, the intrinsic record here for the 346 patent is completely ignored by the board. [00:11:35] Speaker 04: The board should not have resorted to extrinsic evidence. [00:11:39] Speaker 06: The board didn't ignore the intrinsic evidence. [00:11:42] Speaker 06: You say that repeatedly the acronym is explained or defined, but it happens twice, one of which was in the abstract, which we know doesn't mean very much. [00:11:56] Speaker 06: So twice they tell you what the words of the acronym are in a colloquial form, but that doesn't mean that that explains what that word means for purposes of this patented invention. [00:12:09] Speaker 04: And beyond expanding the acronym, the 346 specification at column one, lines 18 says, a RAID is a storage system based on a large capacity and a high performance. [00:12:20] Speaker 03: I guess my problem with that is that that expression is [00:12:25] Speaker 03: is sometimes definitional and is sometimes not. [00:12:28] Speaker 03: If I said a rectangle is a four-sided polygon, I would not be defining a rectangle. [00:12:33] Speaker 03: I would be describing a feature of it. [00:12:37] Speaker 03: Right? [00:12:37] Speaker 03: Yes. [00:12:38] Speaker 03: So why is this something other than that? [00:12:40] Speaker 04: There is nothing in the 3-4-6 specification which talks about single logical unit. [00:12:45] Speaker 04: There is nothing in this definitional language or this descriptive language. [00:12:48] Speaker 03: No, no, no, but the point is this, that the only [00:12:52] Speaker 03: thing of informational value in the intrinsic evidence, and here I mean in the intrinsic evidence, is basically that sentence, and there's a version of it in the abstract. [00:13:05] Speaker 03: But the board, it seemed to me, was perfectly entitled to say, we don't know whether that's definitional or not. [00:13:11] Speaker 03: In order to tell whether it's definitional, I'm asking myself the same question as if it said a rectangle is a four-sided polygon. [00:13:18] Speaker 03: And I need to go and find out whether, in the extra patent world, people use the thing in front of the is, rectangle, to mean something more. [00:13:28] Speaker 03: It's one of the kinds of four-sided polygons, but it's not all of them. [00:13:33] Speaker 04: And, Your Honor, in this case, as in vitronics, if the intrinsic evidence is not ambiguous, resort to ambiguous extrinsic evidence is improper. [00:13:43] Speaker 04: We pointed out that the- No, it's not. [00:13:45] Speaker 06: Phillips made it clear that it's not. [00:13:47] Speaker 06: I mean, I don't know why we're still going back to Petronix. [00:13:49] Speaker 06: It's not improper. [00:13:53] Speaker 06: The intrinsic evidence might be more persuasive at times, but it's never improper to look at it. [00:13:58] Speaker 04: I over-spoke, Your Honor. [00:13:59] Speaker 04: The extrinsic evidence in this case on this record is certainly ambiguous. [00:14:03] Speaker 04: There is extrinsic evidence that predates the priority date of this patent, which has no such single logical unit limitation. [00:14:12] Speaker 04: The YGANT reference expressly relied on by the board says, [00:14:16] Speaker 04: raids can be configured in many ways and only uses the single unit language as an alternative and does not require it. [00:14:23] Speaker 06: There is no definitive... But as Philip said, we are supposed to interpret words in light of the entirety of the specification and in light of what it is that that specification discloses. [00:14:35] Speaker 06: The board said it's not very clear what the specification discloses, but it does disclose one thing. [00:14:41] Speaker 06: It says the raid. [00:14:43] Speaker 06: And so [00:14:44] Speaker 06: They had to find out what the raid meant for purposes of this patent. [00:14:48] Speaker 04: And as I said, at column one, it goes on to say, a raid is a fault-tolerance system in which the disks or controllers have a redundant nature. [00:14:57] Speaker 04: There is specific disclosure in the specification as to what a raid is for purpose of this patent, which is absolutely consistent with the extrinsic evidence. [00:15:06] Speaker 04: All the extrinsic evidence does, in fact, comport with this definition, redundant nature of fault tolerance. [00:15:11] Speaker 04: The extrinsic evidence is not [00:15:13] Speaker 04: wholly consistent with the additional limitation of single logical unit. [00:15:17] Speaker 04: That in fact the phrase single logical unit does not appear in any of the extrinsic evidence. [00:15:22] Speaker 03: But I mean just assume with me that I think that variants are everywhere with the possible exception of Weygand and there was certainly evidence from Dr. Conte and the board I think was signaling [00:15:38] Speaker 03: some uncertainty about what they have to make of Wagant. [00:15:41] Speaker 03: Dr. Conte said it was nonsensical. [00:15:44] Speaker 03: It's not entirely clear what the language of Wagant meant, whether it was limited to some arrangement that uses both striping and mirroring. [00:15:55] Speaker 03: And the board's language, I think at one point it said Wagant indicates this, and then the second time on page nine of the opinion it said that [00:16:07] Speaker 03: that term might extend to certain non-singological units. [00:16:13] Speaker 03: And ultimately, why couldn't the board say, we're just not really sure what to make of Wagant. [00:16:19] Speaker 03: It's the odd man out here. [00:16:20] Speaker 03: And a skilled artisan would really, quite generally, understand that when you refer to a raid, you're referring to something that is acting as a single logical unit in the particular sense that allocations inside it [00:16:35] Speaker 03: are something that the host is blind to. [00:16:39] Speaker 04: So first, Your Honor, the Weigandt is not... Under Chanter, we have to look at what the board actually did here. [00:16:46] Speaker 04: And their construction of Singological Unit relies expressly on Weigandt and does not squarely address why the alternatives... It doesn't say, in fact, that they're not sure why it's talking about alternatives. [00:16:58] Speaker 04: They mention it [00:16:59] Speaker 04: But it's counsel's argument that there's some uncertainty there. [00:17:03] Speaker 03: The actual construction that is adopted says single logical unit without exception, doesn't it? [00:17:10] Speaker 03: Yes, Your Honor. [00:17:10] Speaker 03: And I think that's what the other side defends. [00:17:13] Speaker 03: It's only in, I think, two sentences that don't get incorporated into the formal claim construction that says, well, we're not so sure about whether you need to have a single logical unit if you have a combination of striping and mirror. [00:17:26] Speaker 04: But Wygant says that, first of all, Wygant says that mirroring is raid level one. [00:17:31] Speaker 04: Chen says mirroring is raid level one. [00:17:34] Speaker 04: And Hathorne, their own expert, sorry, patent owner's expert, admits Hathorne discloses raid level one. [00:17:39] Speaker 03: I'm not sure it says that all mirroring is raid level one. [00:17:46] Speaker 04: And in fact, we would submit, Your Honor, that that's the- Where does he admit that? [00:17:56] Speaker 05: Sorry, Your Honor. [00:17:58] Speaker 05: I'm looking at your discussion around 57, 58 of the Red Beret. [00:18:03] Speaker 04: He admits it at JA 2100, starting at line 25, and ending at 2101, line 2. [00:18:11] Speaker 04: Did you raise that argument before? [00:18:12] Speaker 05: Yes, we did, Your Honor. [00:18:15] Speaker 05: Did you submit the figure in page 58 to the board? [00:18:22] Speaker 04: Yes, Your Honor. [00:18:25] Speaker 04: Where is that in the record? [00:18:33] Speaker 04: You can find it, Your Honor. [00:18:34] Speaker 04: It was in the reply brief, I believe. [00:18:38] Speaker 04: Is that correct? [00:19:00] Speaker 04: Our reply before the board pointed to Dr. Conti's admission with reference to the mirroring operation at JA337. [00:19:07] Speaker 04: And we presented the figure below, which is identical to that in our brief. [00:19:10] Speaker 04: And if you compare our brief at page 55 with JA1770 and then also look at the citation at JA389 starting at line 5 and JA426 starting at line 8. [00:19:27] Speaker 06: Okay, I think you're out of time. [00:19:28] Speaker 06: I'll give you three minutes for a rebuttal. [00:19:30] Speaker 04: Thank you. [00:19:39] Speaker 05: Start by discussing your statement in the red break at 56 that the illustration at 58 was not presented to the board. [00:20:01] Speaker 05: You're referring, excuse me, excuse me your honor. [00:20:09] Speaker 01: You are referring to the statement in page 56 of the red brief, your honor. [00:20:13] Speaker 05: Yes, that begins the drawing on page 58, Anne. [00:20:22] Speaker 01: I do not believe that the drawing on page 58, the annotated figure of, figure three of Hathorne, was presented below to the board. [00:20:31] Speaker 05: Well, it's not a question I do not believe. [00:20:33] Speaker 05: You made a statement in your red brief. [00:20:37] Speaker 01: Yes. [00:20:38] Speaker 01: That's a certification to the court. [00:20:42] Speaker 01: Yes. [00:20:43] Speaker 01: I do not recall that this was presented to the board below. [00:20:46] Speaker 01: If it was... I couldn't find it. [00:20:50] Speaker 01: I couldn't find it. [00:20:51] Speaker 01: If I overlooked it, I apologize. [00:20:53] Speaker 01: If it was below, it was such a minor point that it escaped me. [00:21:01] Speaker 06: Hi. [00:21:04] Speaker 06: Let's go to the construction issue. [00:21:07] Speaker 06: The RAID, it is true that twice, in parens, it defines it. [00:21:15] Speaker 06: Does it not in the spec? [00:21:18] Speaker 01: No. [00:21:18] Speaker 01: Twice in the specification, the patent expands the acronym. [00:21:22] Speaker 01: It says that a RAID is an acronym for redundant array of independent disks. [00:21:27] Speaker 01: That is not presented as a definition of the term RAID, and it [00:21:32] Speaker 01: That expansion of the acronym by itself does not convey the full meaning of the term to one of ordinary skill in the art. [00:21:41] Speaker 01: That's what the board recognized, and that's why the board turned to the extrinsic evidence. [00:21:46] Speaker 06: And there are lots of times, though, when acronyms just calling out the acronyms, like RAM or DRAM. [00:21:53] Speaker 06: I mean, they tell you what they are, right? [00:21:55] Speaker 01: Sometimes they do. [00:21:56] Speaker 01: It's case by case, Your Honor. [00:21:58] Speaker 01: Some acronyms, when expanded, may fully and completely [00:22:02] Speaker 01: indicate the meaning of that acronym to those skilled in the art. [00:22:06] Speaker 01: And sometimes they do not. [00:22:07] Speaker 01: In this case, the acronym RAID does not. [00:22:10] Speaker 01: In fact, it's even misleading. [00:22:13] Speaker 01: The word redundant from the letter R is actually optional because RAID level 0 is a non-redundant form of RAID. [00:22:22] Speaker 01: The word inexpensive is actually not correct because those of skill in the art understand that the [00:22:31] Speaker 01: array or the disk may be independent or inexpensive. [00:22:35] Speaker 01: And in fact the D for disk is imprecise. [00:22:39] Speaker 01: It actually means disk drives to those skilled in the art, not disks. [00:22:45] Speaker 01: And that's true for many acronyms. [00:22:47] Speaker 01: They're catchy. [00:22:48] Speaker 01: They're shorthand. [00:22:49] Speaker 01: They can be pronounced. [00:22:50] Speaker 01: That's why they were chosen, not because they present a precise definition of the term as understood by one skilled in the art. [00:22:58] Speaker 03: Can you address what you think the board said about Wagant and whether to be a raid there needs to be a single logical unit full stop without exception or basically on pages 8 and 9 it seemed to me that the board [00:23:28] Speaker 03: was saying, WGAN, I think one term was, indicates that maybe that's not always true, but that the board didn't come to a firm conclusion about that and the actual claim construction it adopts at the end of that little section of its opinion doesn't have any limitation in it. [00:23:49] Speaker 03: And I'm trying to figure out what to make of what the board did [00:23:56] Speaker 03: bowing toward Wagant and then applying a, there must be a single logical unit. [00:24:03] Speaker 01: Yes, so the board did correctly recognize that Wagant does say that perhaps some aspect, in some sense, there may be a raid that is not a single logical unit. [00:24:14] Speaker 01: But the board credited the testimony of Dr. Conte, the expert in this case, who said, I can't make sense of what that means. [00:24:22] Speaker 03: The board didn't cite that. [00:24:24] Speaker 03: Did it? [00:24:24] Speaker 03: I don't think it did, but you told me. [00:24:26] Speaker 01: Perhaps they didn't cite that in their opinion, but they were well aware of it. [00:24:30] Speaker 01: That was testimony that was presented by Dr. Conte. [00:24:35] Speaker 06: But didn't Dr. Conte say that if you're going to have mirroring or striping, you have to have two? [00:24:42] Speaker 01: You have to have two disk drives to do mirroring or striping. [00:24:44] Speaker 01: You must have multiple disk drives, yes. [00:24:48] Speaker 06: All right, so how does that encompass a raid if a raid has to be a single logical unit? [00:24:54] Speaker 01: So the multiple disk drives that are the constituents of a RAID are presented to the host as a single logical unit. [00:25:03] Speaker 01: Dr. Conte explained that that is what a RAID does. [00:25:07] Speaker 01: It presents a single interface to the host. [00:25:10] Speaker 01: It's like a black box. [00:25:12] Speaker 01: So the host computers look at this black box as if it were just another storage device. [00:25:17] Speaker 01: And in fact, the host computers don't know that the storage device is a RAID or a simple disk drive. [00:25:24] Speaker 01: That is the nature of a black box, and that is the nature of a raid, that it presents itself as a single logical unit. [00:25:36] Speaker 01: And that is why the board was correct to conclude that a raid is a firm requirement for a raid to be a single logical unit, and that is their construction. [00:25:46] Speaker 01: And even if there were some evidence to the contrary in the record below, [00:25:51] Speaker 01: Your honors can and should affirm under the substantial evidence standard, because these are factual findings we're talking about. [00:25:57] Speaker 03: So what do you make of column seven of Hathorne, which says that if the host, the primary host up on the top portion of the diagram, wants to send, put stuff in a number of the different disk drives, [00:26:19] Speaker 03: or the DASDs, I guess, as they're called. [00:26:23] Speaker 03: There are kind of two ways that can happen. [00:26:25] Speaker 03: The host can direct that to happen through the storage controller or can give it to the storage controller in the north portion of the diagram, and the storage controller can do the allocating. [00:26:38] Speaker 03: Why isn't the first way of doing it something that amounts to a rape? [00:26:46] Speaker 01: If that were the only way to do it, then perhaps there would be a raid. [00:26:51] Speaker 01: But the mere existence of the capability to directly access that second disk drive, a constituent disk drive of the alleged raid, means that the pair of disk drives is not a single logical unit. [00:27:05] Speaker 01: They are separate logical units. [00:27:07] Speaker 01: And that is why each one is independently or individually accessible. [00:27:12] Speaker 03: Can you connect that thought [00:27:13] Speaker 03: in a way I might be able to understand to the statement that Dr. Conte made, I think this is 1745 and 1746, that says a raid is something that you can just, and you referred to this a moment ago, swap out for a single large expensive disc, is that the sled, without changing the interface. [00:27:36] Speaker 03: Why would, if you started in a system with a single large disk and some interface between the host and the sled, and you then created the Hawthorne system, why would the interface have to change? [00:27:52] Speaker 03: The Hawthorne system, column seven, which allows two options to get the data to or from a particular disk under the control of the host or under the control of the storage controller directly connected to the host. [00:28:07] Speaker 03: When... Why would there have to be a different interface? [00:28:11] Speaker 01: There is a different interface in Hathorne. [00:28:14] Speaker 01: There's the direct access to the remote DASD or there's the access that happens by the remote mere copy operation. [00:28:24] Speaker 01: So there's two different ways to access the remote DASD in Hathorne. [00:28:29] Speaker 01: And what Dr. Conte was saying on pages JA 1745 and 46 is that with a raid [00:28:38] Speaker 01: the host computers accesses the raid as a single logical unit through just one interface and cannot look behind the wrapper of the black box. [00:28:49] Speaker 03: What's the interface? [00:28:52] Speaker 03: I'm not adept at this to know what the connection is between the two different ways described in column seven and the term interface and yet Dr. Conte uses the term interface. [00:29:05] Speaker 03: He says [00:29:06] Speaker 03: The idea of a RAID is that if you have a single disk, the sled, and you get rid of it and stick in a RAID, the interface doesn't have to change, the interface between the host and the memory system calling. [00:29:18] Speaker 01: Right. [00:29:18] Speaker 01: So the interface to a RAID is its RAID controller. [00:29:24] Speaker 01: And there may be multiple RAID controllers for the same RAID, each of which presents an interface [00:29:29] Speaker 01: that a host computer can go through, a door through which the host computer can go through to access the RAID. [00:29:38] Speaker 06: Dr. Conte's point is that the interface is the RAID controller and there are multiple RAID controllers. [00:29:45] Speaker 06: Then how is that a logical single unit? [00:29:50] Speaker 01: There are different RAID controllers for the same RAID. [00:29:57] Speaker 01: The key distinction between [00:29:59] Speaker 01: a RAID and separate logical disk is that when you have separate logical disks that are not a RAID, the host controllers can individually access those individual disk drives. [00:30:12] Speaker 01: DASD, D-A-S-D in the IBM terminology in Hathorne. [00:30:16] Speaker 01: It can look behind the wrapper of the black box and access the constituent disk drives. [00:30:22] Speaker 01: Whereas in a RAID, the host computers simply access [00:30:30] Speaker 01: the entirety, the single logical unit, whether it's from RAID controller A or RAID controller B for the same RAID, it doesn't matter. [00:30:39] Speaker 01: And Dr. Conte's point that you can substitute a RAID for a sled means that electronically the host computer just writes to or reads from one of those RAID controllers exactly like it would write to or read from a controller for a single simple disk drive. [00:30:58] Speaker 01: electronically. [00:31:00] Speaker 03: And your position is that as in column seven and a half horn, as long as the host has the option of individually addressing a particular DASD, then it's not a raid, even if it can also deal with the manager at the front desk of the storage unit. [00:31:22] Speaker 01: Yes, exactly, Your Honor. [00:31:24] Speaker 03: Can I get back for a minute to this thorny [00:31:28] Speaker 03: issue of the board and Wagant. [00:31:31] Speaker 03: The board said, maybe Wagant suggests this, but then it had some sentence that said, in effect, striping and mirroring doesn't really seem to be part of this patent. [00:31:47] Speaker 03: Do you remember the phrase? [00:31:52] Speaker 01: So what I believe the board meant there is that the details of striping or mirroring [00:31:59] Speaker 03: It's that we have not been shown evidence that the rate of this patent, the 346, is configured in a striped and mirrored configuration. [00:32:10] Speaker 03: And that I took to kind of be, that was ultimately the board's reason for saying, well, whatever WAGAN says, it's not applicable here. [00:32:19] Speaker 03: Can you? [00:32:19] Speaker 03: And that seems like a point, sort of an intrinsic evidence point, not an extrinsic evidence point. [00:32:25] Speaker 03: And so I want to understand [00:32:27] Speaker 03: to the extent, and I realize there's an argument that that's not really what's dispositive here, but to the extent that's what the board was saying. [00:32:35] Speaker 03: What is it about the 346 patent that would confirm, if you think it does, that striping and mirroring are pretty clearly from the intrinsic evidence not something that this patent is about? [00:32:49] Speaker 01: Well, I believe what the board was saying is that what Wagon, when it refers to [00:32:55] Speaker 01: a situation where there's not a single logical unit, has in mind some combination of striping and mirroring. [00:33:02] Speaker 01: The experts aren't sure what that combination is, but some combination of striping and mirroring that Wagen thinks would still qualify as a raid. [00:33:10] Speaker 01: Like raid 1 slash 0. [00:33:11] Speaker 01: Exactly. [00:33:12] Speaker 01: I believe. [00:33:14] Speaker 01: And the board looked at the 346 patent and realized that the 346 patent talks about raid [00:33:22] Speaker 01: at a high level does not distinguish between particular levels of RAID, does not attempt to call out specifically that particular combination of RAID, RAID 1 slash 0. [00:33:34] Speaker 06: Well, it certainly doesn't exclude it. [00:33:36] Speaker 06: I mean, it's one of the shortest patterns that I've ever read. [00:33:39] Speaker 06: So it doesn't exclude mirroring and striping, does it? [00:33:43] Speaker 01: It does not exclude mirroring. [00:33:45] Speaker 01: It does not exclude striping. [00:33:47] Speaker 01: But to the extent that there's a combination of striping and mirroring that the experts don't understand, [00:33:52] Speaker 01: that one reference says is a raid, it would exclude that, because those skilled in the art would not consider that to actually be a raid. [00:34:01] Speaker 06: So what you think the board really meant here was not that this patent doesn't disclose this, but that we don't understand what this is that White House talked about? [00:34:12] Speaker 01: I believe so, yes. [00:34:13] Speaker 01: And to the extent [00:34:14] Speaker 01: This is relevant. [00:34:15] Speaker 01: It's not what the patent is specifically directed at. [00:34:17] Speaker 06: One last question. [00:34:18] Speaker 06: I thought that their strategy of trying to give us a different opinion, a different board decision to analyze, was an interesting one. [00:34:25] Speaker 06: Because I know you'd rather be here on appeal from that decision, because it was much more fulsome. [00:34:31] Speaker 06: But isn't the mere fact that that decision is more fulsome? [00:34:35] Speaker 06: Doesn't that present a problem for you in this case? [00:34:38] Speaker 06: In other words, if we can look at that, we could say, oh, well, this is the way the board really should do its analysis. [00:34:44] Speaker 06: But it didn't necessarily do that here. [00:34:48] Speaker 01: No, I think the second decision by the board simply underscores the fact that it reached the right conclusion in this case below that's on appeal here. [00:34:57] Speaker 01: The exact same issue with the exact same evidence, plus some additional evidence that clearly supports us. [00:35:03] Speaker 06: More evidence, more analysis. [00:35:06] Speaker 06: I mean, I would think that that's kind of the opposite of what you had hoped we would take from that second decision. [00:35:14] Speaker 01: I hope not. [00:35:15] Speaker 01: I bet that if your honors were presented with a similar issue in a second appeal, you would have additional insights that would allow you to draft an opinion that's perhaps tighter and more complete, but that in no way would... You don't think we get it right the first time every time? [00:35:31] Speaker 05: You don't think we rule 36? [00:35:32] Speaker 01: You always get it right the first time. [00:35:35] Speaker 01: And that's just natural. [00:35:36] Speaker 01: I mean, that's the way things happen if you look at the same issues the second time. [00:35:40] Speaker 01: The important point is they reached the same conclusion on the same issue presented by two different challengers. [00:35:48] Speaker 01: And so the board is really confident in its conclusion. [00:35:51] Speaker 03: OK. [00:35:52] Speaker 03: And it's the same board, right? [00:35:54] Speaker 01: Same panel of the same three judges, yes. [00:36:00] Speaker 06: OK. [00:36:01] Speaker 06: Three minutes. [00:36:03] Speaker 04: Thank you, Your Honor. [00:36:04] Speaker 04: Council argued that the acronym for RAID is not a definition. [00:36:09] Speaker 04: However, Dr. Conti, [00:36:10] Speaker 04: admitted at JA 2065 starting at line 15 that that expansion of the acronym is a definition and he agrees with it. [00:36:17] Speaker 03: A definition of the acronym or definition of the words that define the acronym? [00:36:22] Speaker 04: Well, at that point in the transcript he agreed that the expansion of the acronym is a correct definition. [00:36:27] Speaker 04: He agreed with it. [00:36:28] Speaker 04: Later in the transcript he agreed that the portion I read from the 346 patent, column 1, line 18, is he agreed with that that's a correct definition of raid. [00:36:36] Speaker 04: And so that's a JA 2063 starting at line two. [00:36:40] Speaker 06: You're not contending that every time there's an acronym that the correct construction of this to pull out all the words? [00:36:49] Speaker 04: I don't need to argue that every time there's an acronym that that's the correct construction. [00:36:52] Speaker 04: In this case, it is the correct construction, especially where it's supplemented by that further definitional or descriptive language at column one. [00:36:59] Speaker 04: The only intrinsic evidence. [00:37:01] Speaker 06: Are we supposed to assume that it's inexpensive or independent? [00:37:05] Speaker 04: Which word? [00:37:06] Speaker 04: The fuller description says it has to be fault-tolerant and redundant in nature, so I would argue that's a description of independent disks. [00:37:17] Speaker 06: Okay, but the word independent isn't really in the acronym, is it? [00:37:23] Speaker 04: As expanded in the specification itself, that's correct. [00:37:27] Speaker 03: But the only intrinsic- To use an example Judge O'Malley used earlier, if I wanted to know what the definition of RAM is, you think it would cover every memory system in which access was random? [00:37:41] Speaker 03: There is- You wouldn't. [00:37:42] Speaker 04: No, no. [00:37:44] Speaker 04: And I'm not arguing that every time there's an acronym that it always is, that is always the corrective construction. [00:37:48] Speaker 04: What I'm saying is, even if you look at the extrinsic evidence here, the extrinsic evidence supports the construction [00:37:54] Speaker 05: The testimony upon which you are relying is at best ambiguous. [00:37:58] Speaker 05: I'm sorry, Your Honor? [00:37:59] Speaker 05: I said the testimony of Dr. Conte upon which you are relying is at best ambiguous because of the ambiguity of the questions. [00:38:09] Speaker 04: Well, he expressly says he agrees with those definitions. [00:38:12] Speaker 04: He doesn't caveat that in any way. [00:38:15] Speaker 05: When he says he agrees with the definitions, he doesn't say anything more. [00:38:19] Speaker 05: then the letters R-A-I-D have an accepted meaning. [00:38:25] Speaker 04: And that's correct, Your Honor. [00:38:26] Speaker 04: But he does go on, as I said, later in the transcript at 2063, starting at line two, to agree that the fuller passage at column one, lines 18, which I read, he agrees with that definition. [00:38:37] Speaker 04: Really? [00:38:38] Speaker 05: I think it's at best ambiguous. [00:38:40] Speaker 04: I don't believe he caveat. [00:38:41] Speaker 05: We just reread those pages. [00:38:44] Speaker 04: The issue here is, counsel argued that somebody of ordinary skill would not understand [00:38:49] Speaker 04: raid to have the meaning as set forth in the specification. [00:38:53] Speaker 04: Dr. Conti himself certainly didn't deny that those were appropriate, and so that argument by counsel was not well supported in the record. [00:39:01] Speaker 04: Further, the only intrinsic evidence that the board really looked at here is the phrase A raid. [00:39:05] Speaker 04: They relied on A to be construed as single. [00:39:08] Speaker 06: No, they also relied on the raid. [00:39:09] Speaker 04: But the claim says A. [00:39:10] Speaker 04: The claim says A-RAID doesn't say it isn't limited to a single. [00:39:13] Speaker 04: And they elevated A-RAID to mean one and only one RAID, meaning a single logical unit. [00:39:18] Speaker 04: If I can go very briefly to column seven of Wigandt, Your Honor, and counsel's arguments there. [00:39:24] Speaker 04: Of Haythorne? [00:39:25] Speaker 04: Of Haythorne, I'm sorry. [00:39:26] Speaker 04: All right. [00:39:27] Speaker 04: It's clear in Haythorne that passage in column seven is forbidding direct access between the host and the backup disks. [00:39:35] Speaker 04: It says it has to go through the first controller to get to the second controller. [00:39:41] Speaker 04: And so there's nothing in the board's decision that supports a finding that that disclosure somehow avoids the claim language even as construed. [00:39:50] Speaker 04: In other words...