[00:00:00] Speaker 07: Before we begin our argument this morning, we have four admissions. [00:00:05] Speaker 07: I have the pleasure of moving the admission of my four law clerks from this year who have done an excellent job and kept me on the straight and narrow, and I'm very appreciative. [00:00:19] Speaker 07: They have each one of them been terrific. [00:00:24] Speaker 07: So no particular order. [00:00:29] Speaker 07: I first move the admission of Anthony Shea, who's a member of the Bar and is good standing with the highest court of New York. [00:00:37] Speaker 07: I have knowledge of his credentials, and I'm satisfied that he possesses the necessary qualifications. [00:00:46] Speaker 07: Judge Taranto, I think you and Judge Hughes have to act on the motion. [00:00:50] Speaker 07: Granted. [00:00:54] Speaker 07: Second, I move the admission of Giovanni S. Sarman Gonzalez. [00:00:58] Speaker 07: who's a member of the bar and is good standing with the highest court of California. [00:01:02] Speaker 07: I have knowledge of his credentials and am satisfied that he possesses the necessary qualifications. [00:01:09] Speaker 07: Again, granted. [00:01:11] Speaker 04: We'll see how the rest go. [00:01:12] Speaker 07: Two for two, sir. [00:01:16] Speaker 07: I move the admission of Alexander Ryan Trasiak, who's a member of the bar and is good standing with the highest courts of Texas and the District of Columbia. [00:01:25] Speaker 07: I have knowledge of his credentials and I'm satisfied that he possesses the necessary qualifications. [00:01:31] Speaker 07: Granted once more. [00:01:32] Speaker 07: Okay. [00:01:33] Speaker 07: And then finally, I move the admission of Jimmy Jays Wang, who's a member of the bar and is in good standing with the highest court of Massachusetts. [00:01:41] Speaker 07: I have knowledge of his credentials and satisfied that he possesses the necessary qualifications. [00:01:47] Speaker 07: I think a clean sweep. [00:01:49] Speaker 04: Yeah, hit for the circuit, granted. [00:01:52] Speaker 07: So the four of you should take the oath of office. [00:01:55] Speaker 03: welcome welcome welcome all of you to the bar and and again thank you for all your excellent work [00:02:26] Speaker 07: Now, having done the easy part of this morning's work, we turn to the cases. [00:02:30] Speaker 07: The first of these is number 16-1742, Netlist Inc. [00:02:35] Speaker 07: versus Diablo Technologies Inc., Mr. Loy. [00:02:57] Speaker 01: Thank you, Your Honor, and may it please the Court. [00:03:00] Speaker 01: A single claim construction error infected all the Board's unpatentability decisions, and those decisions should not stand. [00:03:08] Speaker 07: What exactly is the claim construction that Netlist requested in the Board proceedings? [00:03:15] Speaker 01: We requested that the Board construed selectively electrically coupled as electrically coupling in response to a selection. [00:03:20] Speaker 01: We raised that construction at the institution stage, [00:03:24] Speaker 01: during the trial, and that's the construction that we've asked the court here to adopt as well. [00:03:29] Speaker 01: And on the merits of that, the board's construction is wrong for any of several reasons. [00:03:33] Speaker 07: I'm not really seeing that there's any clear difference between that and what the board did. [00:03:39] Speaker 07: Did the board explain that there was a difference between the two? [00:03:46] Speaker 01: The board rejected our claim construction consistently, Judge Dyke. [00:03:51] Speaker 01: And the board instead adopted a different construction [00:03:54] Speaker 01: And the core of the dispute here is about how the board ultimately understood its construction. [00:04:00] Speaker 01: The board understood that selectively electrically coupling can be accomplished by simply sending a signal to a memory device to activate or deactivate that memory device. [00:04:11] Speaker 01: And we consistently at the board challenged that interpretation and told the board that selectively electrically coupling could not be accomplished when you had a memory device that was permanently electrically connected to other parts of your computer system. [00:04:23] Speaker 01: The board, in adopting its construction, got it wrong because the board focused on different concepts. [00:04:29] Speaker 01: For example, the board focused on selecting between components. [00:04:32] Speaker 01: That was the first part of the board's construction. [00:04:34] Speaker 01: But the claim language doesn't say anything about selecting components. [00:04:37] Speaker 01: Indeed, the claim language already tells you the relevant components. [00:04:41] Speaker 01: A single instance of selectively electrically coupling occurs between a memory device data signal line and a common signal line. [00:04:48] Speaker 01: Those are the relevant components. [00:04:50] Speaker 01: And the circuit isn't selecting between those components. [00:04:53] Speaker 01: The circuit is electrically coupling those components. [00:04:57] Speaker 01: The board also got it wrong because the board focused on transferring power signal information. [00:05:03] Speaker 01: That's the other part of the board's construction. [00:05:05] Speaker 01: But the patents use a different term for that. [00:05:07] Speaker 01: They use the term transmit. [00:05:09] Speaker 01: That's in claims six and seven of the 150 patent and also in the specifications. [00:05:15] Speaker 04: But would I be right in understanding your position that, that the difference between what the board did and what you're doing is something like, you think that the act of coupling is creating a path along which certain things can happen, like transmitting signals or transmitting power. [00:05:32] Speaker 04: And the board took the view that it's not creating the path, but it's actually the transmission over a path. [00:05:39] Speaker 01: I think that's mostly right, Judge Schranu with the caveat that [00:05:43] Speaker 01: The word path can be a little bit loose. [00:05:46] Speaker 01: You can have electrical coupling even if there isn't necessarily current transfer across a wire. [00:05:51] Speaker 01: But that's right. [00:05:51] Speaker 01: An electrical coupling is about creating the electrical connection through which signals might be transmitted. [00:05:58] Speaker 01: But the board got it wrong because the board focused on that end result of transmitting. [00:06:02] Speaker 01: But an electrical connection exists even if nothing is ever transmitted. [00:06:06] Speaker 01: And you know that because that's not only the way that claims use the language of coupling, but that's what the specification shows. [00:06:13] Speaker 01: at columns 6 and 7 of the 150 pattern. [00:06:16] Speaker 01: So this is at appendix 166. [00:06:21] Speaker 01: And the board's misunderstanding. [00:06:22] Speaker 07: You say there doesn't have to be a switch, you said. [00:06:25] Speaker 07: So I'm not sure, apart from a switch, what it is that, in your view, would satisfy your construction. [00:06:36] Speaker 01: That's right. [00:06:36] Speaker 01: Our position is not that a switch is required. [00:06:38] Speaker 01: But our position is that you can't have selective electric coupling, Judge Dyke. [00:06:42] Speaker 01: When the signal lines, remember, the claims focus on the signal lines. [00:06:46] Speaker 01: Electrical coupling is about coupling the signal lines. [00:06:49] Speaker 01: If those signal lines are permanently connected, you cannot selectively electrically couple them, because you already have a connection. [00:06:55] Speaker 01: There's always a connection. [00:06:57] Speaker 01: The board thought that that was enough if I'm simply at the end of those lines, you have these memory devices. [00:07:03] Speaker 01: And I simply say, this memory device, your turn to transmit. [00:07:07] Speaker 01: other memory device don't transmit. [00:07:09] Speaker 01: The board focused on that. [00:07:10] Speaker 01: What would be an example of something that would satisfy you that's not a switch? [00:07:14] Speaker 01: So one example that the specification gives is a multiplexer. [00:07:18] Speaker 01: A multiplexer is an electrical component where you might have four signal lines coming in. [00:07:22] Speaker 01: There's a control line. [00:07:23] Speaker 01: And then the multiplexer will electrically couple one of those four signal lines, say, to the output. [00:07:29] Speaker 01: That's one example. [00:07:30] Speaker 01: And one would not call that multiplexer a switch? [00:07:34] Speaker 01: I don't know that there's anything in the record either way, whether a person of skill in the art would think of that as a switch. [00:07:40] Speaker 01: But part of the problem here is that the word switch is, and there's no construction, it's a loosely defined term. [00:07:47] Speaker 01: But the core concept is about coupling, electrically connecting these lines. [00:07:52] Speaker 01: And this isn't just a semantic difference between the board and us. [00:07:56] Speaker 07: Did you suggest to them that a switch or multiplexer was what was required? [00:08:03] Speaker 01: No, what we suggested is what is required is electrically coupling in response to a selection. [00:08:07] Speaker 01: And when the board adopted at petition stage a construction that said you can have electrical coupling when you have permanent hardwired signal lines, we consistently attacked that and said that's not right. [00:08:19] Speaker 07: I don't understand how they're supposed to understand your position if you didn't tell them that you're talking about a switch or a multiplexer or something like that. [00:08:31] Speaker 07: You seem to have given them a somewhat vague claim construction and now on appeal you're suggesting that you're refining it to say it has to be a switch or a multiplexer, but you didn't tell that to the board, right? [00:08:47] Speaker 01: So Judge Dyke, I disagree in at least two ways with that. [00:08:50] Speaker 01: First, there was no confusion at the board about what our position was. [00:08:53] Speaker 01: You can read what the board said at appendix 29 where the board [00:08:57] Speaker 01: recognize that we're saying the exact same thing that we're saying now, that you can't have selective electric coupling when you have permanent connections. [00:09:04] Speaker 01: So appendix 29, that's in volume one of the appendix towards the back of the blue brief. [00:09:15] Speaker 01: Right at the top of the page, the board responds to this issue that we're discussing now. [00:09:18] Speaker 01: They say, additionally, we are unpersuaded by patent owners' argument [00:09:22] Speaker 01: that hardwired data signal lines, such as that taught by Amidi, cannot be electrically coupled in a selective fashion. [00:09:29] Speaker 01: And the other way I disagree with what you said is we did offer a switch as one example of selective electrically coupling. [00:09:37] Speaker 01: The point is simply that we're not trying to say that a switch is the only way to selective electrically couple. [00:09:42] Speaker 01: And there's no requirement. [00:09:45] Speaker 01: that we disclose all potential embodiments in our patent. [00:09:48] Speaker 01: We simply need to disclose enough to satisfy the written description enablement requirements. [00:09:53] Speaker 04: Can you say something about the reference in column 8 to the data buffer? [00:09:59] Speaker 04: Is that consistent with your view? [00:10:01] Speaker 04: Right at the bottom of column 8, certain embodiments, the memory module operates as having a [00:10:09] Speaker 04: a data path rank buffer, which advantageously isolates the ranks of memory devices or modules. [00:10:16] Speaker 04: That sounds like it's doing coupling. [00:10:20] Speaker 04: Is that? [00:10:23] Speaker 04: Not much is made. [00:10:24] Speaker 04: Maybe nothing is made in the red brief of that. [00:10:26] Speaker 04: And I think you have a footnote in your reply brief pointing that not much is made of it. [00:10:31] Speaker 04: But I guess I want to understand it better. [00:10:34] Speaker 01: So again, I think the fundamental dispute is that [00:10:38] Speaker 01: electrical coupling can't be accomplished when you have permanent electrical connections. [00:10:42] Speaker 01: So a buffer may, again, there hasn't been a lot of development on this, Judge Tronto, and so to the extent that we're an issue the other side thought were important, it was their burden to focus on that. [00:10:53] Speaker 01: A buffer could be one way to isolate things, but again, here the claim language is more particular than simply isolating, and it's more particular than simply coupling. [00:11:02] Speaker 01: It's selectively isolating a load in the load cases, so it would come down to, [00:11:07] Speaker 01: Is a buffer actually configured in a way within a circuit to selectively isolate a load? [00:11:13] Speaker 01: And is it configured in a way that it's selectively electrically coupling the relevant data signal line? [00:11:18] Speaker 01: So buffers may be a part of that, but I think it would depend on the circumstances of how the circuit is configured. [00:11:27] Speaker 01: And Judge Dyke, the dispute that we raised, it is the same dispute at the board. [00:11:31] Speaker 01: And it goes to the core of what the invention is about. [00:11:33] Speaker 01: What the board said was selectively electrically coupling. [00:11:36] Speaker 01: is what the prior art did. [00:11:39] Speaker 01: If the board had looked at the specification, it would have understood that. [00:11:42] Speaker 01: But the board's analysis on the claims, it bypassed the specification. [00:11:46] Speaker 01: It said, well, the specification doesn't expressly define selectively electrically coupling, and it went straight to technical dictionary definitions. [00:11:52] Speaker 01: But the claims are the single, or the specification is the single best guide to understanding the claims. [00:11:57] Speaker 01: And here, and we walk through this in the blue brief, if you look at the prior art memory device in figure two, it operates in the same way that the board thought was selectively electrically coupling. [00:12:06] Speaker 01: You have permanent electrical connections. [00:12:08] Speaker 01: We reproduced this on page 19 of our gray brief side by side. [00:12:14] Speaker 01: In the prior art, you had a memory device, and it was permanently electrically connected to the rest of the system. [00:12:19] Speaker 01: And the problem with that type of connection, the specification details this at column six, is that the system is always exposed to the loads. [00:12:26] Speaker 01: When there is no selective electric coupling, the system will always be exposed to the loads of both memory devices. [00:12:33] Speaker 01: That was OK with the prior art. [00:12:34] Speaker 01: Because the computer system knew how many memory devices to expect on a module. [00:12:38] Speaker 01: The computer system was designed to handle, say, eight memory devices on a module. [00:12:42] Speaker 01: So if all of the devices are being coupled, the system can handle that load. [00:12:47] Speaker 01: But our invention was about doubling or quadrupling the number of memory devices on a module beyond what the system was designed to handle. [00:12:55] Speaker 01: And if you configured it in the way that the board thought and the way the prior art shows. [00:13:01] Speaker 01: Didn't Klein do that in the prior art? [00:13:03] Speaker 01: So Klein did disclose putting more memory devices on a module. [00:13:08] Speaker 01: But Klein did not disclose this aspect of the invention, which was about selectively coupling or isolating loads. [00:13:17] Speaker 01: In fact, what Klein did was the same as the prior art. [00:13:19] Speaker 07: I thought it did. [00:13:20] Speaker 07: I thought it showed switches where you could turn it on and off. [00:13:24] Speaker 07: That's right. [00:13:25] Speaker 01: So I was actually thinking of a meaty. [00:13:28] Speaker 01: On Klein, Klein does show switches. [00:13:30] Speaker 01: But switches are neither here nor there. [00:13:32] Speaker 01: Our claims don't require switches. [00:13:33] Speaker 06: Wait, wait, wait. [00:13:35] Speaker 06: You said they don't require switches, but they do cover switches. [00:13:40] Speaker 01: They cover switches configured in the specific way that the claims detail. [00:13:43] Speaker 01: The claims here are very specific. [00:13:45] Speaker 01: There has to be selective electric coupling between a device data signal line and a common signal line. [00:13:51] Speaker 01: There are switches inclined. [00:13:52] Speaker 01: We don't dispute that. [00:13:53] Speaker 01: But look at what the switches do incline. [00:13:56] Speaker 01: That's at appendix 1630. [00:14:09] Speaker 01: This is for one of them, embodiment inclined. [00:14:12] Speaker 01: But if you look at the description, it's volume two, Judge Stein. [00:14:19] Speaker 01: What page? [00:14:20] Speaker 01: 1630. [00:14:22] Speaker 01: It's about maybe a fifth of the way through volume two. [00:14:26] Speaker 01: They're big. [00:14:31] Speaker 01: OK. [00:14:31] Speaker 01: So this is showing one embodiment inclined. [00:14:33] Speaker 01: But if you read the description, it applies to all. [00:14:36] Speaker 01: Yes, there are switches here. [00:14:37] Speaker 01: But if you look at how the switches are configured, the switches are configured to couple or decouple an entire memory module. [00:14:43] Speaker 01: That block that says memory on the right, that's an entire memory module. [00:14:47] Speaker 01: That's not an individual memory device. [00:14:50] Speaker 01: And that's not what our claims require. [00:14:53] Speaker 01: And it also wouldn't work for our invention. [00:14:55] Speaker 07: I thought there was testimony that the switches in Klein were disclosed [00:15:04] Speaker 07: apply to individual memory devices and not just the module? [00:15:11] Speaker 01: I think there was testimony that the switches could be put on a memory module. [00:15:15] Speaker 01: And in some instances, the switches could actually be built into the memory devices. [00:15:19] Speaker 01: But the way the switches are all configured, which that's what the claims are about, is how the circuit is configured. [00:15:25] Speaker 01: goes to it couples or decouples the entire module. [00:15:29] Speaker 07: I mean, we have to deal not with your argument as to what Klein discloses, but since this is really complicated stuff, but with what the testimony and expert declaration said about this. [00:15:43] Speaker 07: And I thought the petitioner had expert testimony that Klein disclosed using the switches [00:15:51] Speaker 07: to apply to the individual devices as well as to the module. [00:15:55] Speaker 07: Am I wrong about that? [00:15:58] Speaker 01: I'd have to look. [00:15:58] Speaker 01: I didn't see that in the red brief, Judge Dyke. [00:16:01] Speaker 01: But we also have expert testimony saying the same, appendix 22 or 2490. [00:16:06] Speaker 01: And I agree that this is complicated stuff, which is why this issue of whether Klein would disclose under our construction isn't properly before the court. [00:16:15] Speaker 01: The board made all of its findings based on its construction. [00:16:19] Speaker 01: And it consistently rejected our expert testimony as based on our construction and not the board's. [00:16:23] Speaker 01: So to the extent that there's factual disputes about what Klein discloses and whether it discloses under our construction, that's an issue the board needs to resolve in the first instance. [00:16:32] Speaker 01: If I may, I'd like to reserve the remainder of my time, but I'm happy to take it. [00:16:36] Speaker 07: My question is whether there's conflicting testimony as to what Klein shows in this respect. [00:16:43] Speaker 01: Yes. [00:16:43] Speaker 01: So at appendix 2490, that's our expert. [00:16:57] Speaker 07: OK. [00:16:59] Speaker 07: So where am I supposed to look here? [00:17:03] Speaker 01: So in paragraph 63, what is this? [00:17:12] Speaker 01: I'm sorry. [00:17:12] Speaker 01: I gave you the wrong site, Judge Dyck. [00:17:16] Speaker 01: It is at appendix 21, 22. [00:17:19] Speaker 01: I'm sorry. [00:17:21] Speaker 01: OK. [00:17:29] Speaker 01: So this is starting at the bottom of the page, Judge Ike. [00:17:32] Speaker 01: This is our expert in his deposition. [00:17:35] Speaker 01: He says, well, actually, in all the shown embodiments, that can selectively decouple or sort of decouple a whole memory module. [00:17:42] Speaker 01: That is the granularity. [00:17:44] Speaker 01: And then at the bottom of PENIX 2123, about line 19, he reconfirms that testimony. [00:17:51] Speaker 01: So that is why I'm going to say that it decouples or couples an entire memory module, because that embodiment meets the objective of the summary of the invention [00:18:00] Speaker 07: talking about client. [00:18:05] Speaker 05: Am I misreading the next sentence that says one could argue, I suppose, if you went to figure 10, there's a possibility of decoupling a memory device? [00:18:14] Speaker 01: So that is what he says. [00:18:16] Speaker 01: And if you look at the figure 10, it shows switches being built into the memory devices themselves. [00:18:25] Speaker 01: That's not consistent with our claims for two reasons. [00:18:30] Speaker 01: It has to be the circuit that selectively electrically couples, and the claims are written in a way that's clear the memory devices are different from the circuit. [00:18:36] Speaker 01: But also, even when it's built into the chip, it's still coupling all of the devices at the same time. [00:18:43] Speaker 07: Thank you. [00:18:44] Speaker 07: OK, we'll give you two minutes for a bottle. [00:18:46] Speaker 07: Thank you, Judge. [00:18:52] Speaker 02: Mr. Maria. [00:18:55] Speaker 02: Good morning, Your Honors, and may it please the court. [00:18:58] Speaker 02: Appellee asked this court to affirm the board's rejections of all three IPRs on three separate grounds. [00:19:06] Speaker 02: The first and most important one, as we just heard from Natlis, the board actually made determinations on what was the prior art that was being used to invalidate. [00:19:21] Speaker 02: And those determinations went beyond just applying the board's own construction. [00:19:25] Speaker 02: And in fact, as [00:19:26] Speaker 02: Your Honor, as you were in Natalie's part of the argument, it did specifically address whether the combination of prior art, in fact, included a switch. [00:19:36] Speaker 02: And the board had to do that because there were two dependent claims, 17 and 33 in the 150 patent, that expressly required a switch. [00:19:44] Speaker 02: And the board determined factually that, in fact, the switch was a present element of that combination. [00:19:51] Speaker 02: And I'm going to focus on the combination of a midi and client, just to keep things simple. [00:19:56] Speaker 02: So in a median climb, regardless of whether Netlis arguments are correct on claim construction, there is the complete circle that's required by the claims, including the critical switch element that Netlis admitted is not necessarily required, but it would meet the claim limitation of selectively electrically coupling and selectively isolating the other and the rest of the claims. [00:20:21] Speaker 07: The argument appears to be that while Klein discloses switches, it's a switch which acts on the module as a whole and not the individual. [00:20:30] Speaker 02: But first of all, I don't know that that argument was preserved on appeal, but let's assume it was preserved because I want to answer it factually. [00:20:37] Speaker 02: They point into the wrong figure in Klein. [00:20:39] Speaker 02: That's not the figure that the board relied upon. [00:20:42] Speaker 02: The figure relied on figure six of Klein, which is at appendix page 1632. [00:20:53] Speaker 02: which I actually think was part of what Dr. Sechen was deposed about. [00:20:59] Speaker 02: And on Figure 6, you can clearly see that there are transfer gates 64, each one of which is connected to a memory device, which are labeled 62 in that figure. [00:21:11] Speaker 02: And then there is a dashed box around all of those solid line boxes, and that's the memory module in that figure. [00:21:18] Speaker 02: So in Figure 6, clearly, [00:21:21] Speaker 02: there are individual switches for each memory device and multiple memory devices on a single memory module. [00:21:29] Speaker 05: I'm sorry, where does this show your individual switches? [00:21:33] Speaker 05: The transfer gate 64. [00:21:35] Speaker 05: Are those the switches? [00:21:37] Speaker 02: Those are the switches. [00:21:38] Speaker 02: And then you see the memory module 62. [00:21:41] Speaker 02: And if you read the specification, it actually makes it clear those are separate switches for each of the memory module or memory devices. [00:21:49] Speaker 07: Did your expert explain that? [00:21:51] Speaker 02: Yes, he explained that in his declaration. [00:21:53] Speaker 02: What word did he use? [00:21:53] Speaker 02: He gives a page. [00:21:57] Speaker 02: Is it that? [00:21:59] Speaker 02: He needs supplemental declaration. [00:22:06] Speaker 02: So in the appendix at 8525, paragraph 41. [00:22:09] Speaker 02: And I will also refer to paragraph 42. [00:22:34] Speaker 02: 85 what? [00:22:35] Speaker 02: So it's 8525. [00:22:40] Speaker 02: Paragraphs 41 and 42 address this particular aspect of client. [00:22:45] Speaker 02: And you will see that in paragraph 42, there's an express reference to figure 6. [00:22:52] Speaker 02: And it refers to the memory device. [00:22:53] Speaker 02: And it says that client teaches that control signals for controlling a switch may be generated on modules in figure 6. [00:23:02] Speaker 02: And there's other portions of this declaration. [00:23:05] Speaker 04: But where did the board make findings about this specific issue? [00:23:09] Speaker 02: Thank you, Aaron. [00:23:10] Speaker 02: That's exactly my next point. [00:23:11] Speaker 02: So if we look back at actually the very same page in the board's decision, the council referred to page 29 of the appendix. [00:23:22] Speaker 02: And I'm also going to be referring in a moment to page 32 of the appendix. [00:23:25] Speaker 02: But let's start with 29. [00:23:27] Speaker 02: Council referred to the first half of the page [00:23:31] Speaker 02: where the board addressed this issue of what Amidi teaches and rejected Natalie's argument that hardwired lines are uncovered. [00:23:41] Speaker 02: But then, in the very last sentence, they said, furthermore, patent owners declare on Dr. Sachin, testify decline discloses the use of MOSFET switches, i.e. [00:23:52] Speaker 02: data bus switches, for decoupling select memory circuits from the data bus. [00:23:57] Speaker 02: So if Netflix Arguments were correct, the board was just applying its own construction. [00:24:03] Speaker 04: But that sentence doesn't specifically address the distinction between decoupling individual devices within a module and decoupling the entire module. [00:24:12] Speaker 04: And this, at least as I look at columns, there aren't columns, I guess, in Klein. [00:24:17] Speaker 04: But page three of Klein, where the transfer gate 64 are discussed in figure six, I can't quite tell. [00:24:25] Speaker 04: whether all the transfer gates are decoupled, thereby decoupling the entire module. [00:24:29] Speaker 04: There's at least one sentence that says that. [00:24:31] Speaker 02: That is correct, Your Honor. [00:24:32] Speaker 02: This paragraph doesn't address that issue. [00:24:34] Speaker 02: My argument is slightly different. [00:24:36] Speaker 02: Let me just complete my argument. [00:24:37] Speaker 02: I'm going to go back to that point, which I honestly don't think was raised on appeal. [00:24:41] Speaker 02: That's why I'm focusing on the earlier point. [00:24:43] Speaker 02: So first, the board clearly was cognizant of Netlis' argument that a switch might be required. [00:24:49] Speaker 02: They made an express finding of this, which was disclosed in the combination. [00:24:52] Speaker 02: But if you then go back, go on to page 32, which is actually the finding of the board, where they say they find these claims obvious in light of the combination of a median client. [00:25:03] Speaker 02: This is on page 32. [00:25:04] Speaker 02: It's the first full paragraph. [00:25:07] Speaker 02: They set forth that conclusion. [00:25:08] Speaker 02: And then they tell you exactly what the combination is. [00:25:12] Speaker 02: They say that they're persuaded that a MIDI teaches a circuit with a logic element, a register, and a PLL. [00:25:18] Speaker 02: Those are the structural elements of the circuit. [00:25:20] Speaker 02: We're also persuaded that the teachings of a MIDI could have been implemented using the common data signal line and switch system disclosed incline, so that selective electrically coupling a first data signal line to a common data signal line. [00:25:35] Speaker 02: That's the disputed element. [00:25:37] Speaker 02: So they are relying not on Amidi alone, but on the circuit of Amidi, which is multiple memory devices on a single board with all the required circuit elements modified to include the transfer gates of Klein, for example, in figure six. [00:25:54] Speaker 02: So the actual combination that the board applied wasn't Amidi and Klein separately. [00:26:01] Speaker 02: It was a combination. [00:26:03] Speaker 04: But how do we know to read it that way, given that [00:26:06] Speaker 04: Two pages earlier, the board said pretty clearly that hardwired data signal lines meet the selective electrical coupling. [00:26:16] Speaker 02: Because I think they're making a further finding. [00:26:18] Speaker 02: So in the very sentence that you just referred to, they then say, furthermore, Dr. Sachin actually second admits the client, the secondary reference, expressed in these closed switches. [00:26:31] Speaker 02: And then when they do the combination, they point to Klein. [00:26:34] Speaker 04: But again, it didn't admit that it did it at the device level. [00:26:39] Speaker 05: Which reference do we get the fact that they're coupling devices rather than whole modules? [00:26:48] Speaker 05: Is it a meaty or is it Klein? [00:26:50] Speaker 05: Both. [00:26:52] Speaker 02: So figure six of Klein-Kuhn. [00:26:53] Speaker 05: Klein is what you just walked us through. [00:26:55] Speaker 05: Let's assume that I find that a little unclear. [00:26:58] Speaker 05: Is it in a MIDI, too? [00:27:00] Speaker 02: Yes. [00:27:00] Speaker 02: So there are separate signals. [00:27:02] Speaker 02: So the Netflix argument is in a MIDI, there are data lines that go to each individual memory module, memory circuit within the memory module. [00:27:12] Speaker 02: And there are no switches. [00:27:13] Speaker 02: So Klein is supplying the teaching with switches. [00:27:16] Speaker 05: The switch in a MIDI gets you to the devices. [00:27:18] Speaker 05: Where's your testimony on the fact that a MIDI has that structure? [00:27:23] Speaker 02: I want to understand what structure your honor is referring to. [00:27:27] Speaker 05: The device that they coupled devices, not the entire memory. [00:27:33] Speaker 05: I'm trying to get the terms right here. [00:27:35] Speaker 05: I forget what the patent used, but there's the bigger memory module on the individual devices. [00:27:40] Speaker 05: Yeah, I understand the question. [00:27:41] Speaker 02: So there is testimony throughout Dr. Jagannathan's declaration and in the board's description where they address what I mean he teaches. [00:27:50] Speaker 02: So for example, [00:27:52] Speaker 02: If you look at appendix page 21 and 20, those are the board's findings on the immediate reference. [00:28:15] Speaker 02: Okay, so they first walk through the elements of the claim. [00:28:18] Speaker 02: And as the court might recall, the circuit requires a register, a CPLD, and a PLL. [00:28:24] Speaker 02: Those are shown in Figure 4A on page 20 of the determination. [00:28:29] Speaker 02: Then they go on to the next page. [00:28:31] Speaker 02: On page 21, they show you Figure 6A of a MIDI that has the connections between the CPLD and the module connector, which then goes to the memory controller. [00:28:42] Speaker 02: And by the way, in the previous figure, in Figure 4A. [00:28:47] Speaker 05: I mean this is very difficult and it seems like at this point we're trying to figure out whether even under their construction the board would have affirmed. [00:28:54] Speaker 05: But it's very unclear to me, this language you're pointing to, whether it's specifying the difference between devices or modules. [00:29:02] Speaker 05: The last sentence of that paragraph right before to overview of clients says, a media explains that system chips select signal control the ranks of individual memory modules. [00:29:14] Speaker 05: Do you understand that word module to be different than device when they're talking about it? [00:29:19] Speaker 02: Word module is different than device. [00:29:20] Speaker 02: There is a single module shown in a MIDI, and that's the one in figure 4a. [00:29:24] Speaker 02: There are not multiple modules. [00:29:26] Speaker 02: So everything in a MIDI that the board relied on is depicted on figure 4a, which is a single module, which is this box labeled 400. [00:29:36] Speaker 02: So all of these memory devices 404 that you see on the top line of that circuit, those are all individual memory devices on the same memory module. [00:29:45] Speaker 02: And the connections that are discussed throughout a MIDI are always referring to the same module. [00:29:51] Speaker 02: Now, if we go back to the board's decision on pages 22 and 23, that's when they address figure 6 of Klein. [00:30:03] Speaker 02: And that's the same figure I referred you to. [00:30:05] Speaker 02: And in that figure, again, you see that all those memory devices labeled 62 are connected to the transfer gates 14. [00:30:18] Speaker 02: And the board relies expressly on that figure. [00:30:20] Speaker 02: Figure six, not the figure that council cited in its arguments that that figure applies to all the environments. [00:30:28] Speaker 02: No. [00:30:28] Speaker 02: It is inclined and expressed [00:30:31] Speaker 02: environment figure 6 that has multiple memory devices connected through transfer gates to the common data bus. [00:30:39] Speaker 02: So that's what the board was looking at when it made its determination of obviousness of the MEDI plus Klein combination. [00:30:46] Speaker 04: I would also submit that... Can I just double check? [00:30:49] Speaker 04: Is there something in this passage describing Klein that says there are... there clearly are multiple memory devices in the [00:31:01] Speaker 04: What stack? [00:31:02] Speaker 04: Is that 62? [00:31:02] Speaker 04: Correct. [00:31:05] Speaker 04: But that the transfer gates basically turn them on and off individually as opposed to simply making a disconnect between the whole group of them on the module and the rest and the data line. [00:31:26] Speaker 04: I'm not seeing anything that I think at least not [00:31:30] Speaker 02: in a way I can understand, gets at the device specific issue. [00:31:34] Speaker 02: If you look on page 23, the very last part before the section entitled three analysis says, according to Klein, integrated circuit and transfer gate output are connected to data buffer registers, which I think in that case, you're referring. [00:31:49] Speaker 02: Well, that's the other side. [00:31:51] Speaker 02: So you're right. [00:31:51] Speaker 02: This section doesn't expressly address that issue. [00:31:54] Speaker 04: Can I just ask, that's sort of at the global level, why doesn't it make sense if [00:32:00] Speaker 04: We set aside the board's construction to send it back for the board to sort out what seems like a fairly complicated set of issues, not so much about a media, but about some of the other references and not just Klein. [00:32:14] Speaker 02: Well, in the first instance, because quite honestly, that wasn't raised on appeal. [00:32:19] Speaker 02: The only issue that Natalie's raised on appeal is the claim construction. [00:32:23] Speaker 02: They didn't dispute any of the board's findings. [00:32:26] Speaker 05: Sure, but I mean, I think what we're talking about now is presuming that the board's claim construction is wrong and still trying to see whether it comes out the same way. [00:32:38] Speaker 05: And if we can't figure out that it comes out the same way, then shouldn't we send it back? [00:32:46] Speaker 05: You're arguing lots of complicated stuff to us about what a meaty, inclined teacher is based upon the board's opinions, which it seems to me vague. [00:32:56] Speaker 05: in your explanation of the drawings, but you don't have specific or you haven't pointed to specific expert testimony that particularly discussing this figure from Klein we've been talking about. [00:33:09] Speaker 05: Is there anything more on that? [00:33:11] Speaker 02: No, there is actual in the supplemental declaration, Dr. Jagannath, and he goes into through this in detail. [00:33:16] Speaker 05: And it says that that figure in Klein teaches that the transfer gates can [00:33:21] Speaker 02: And go to individual memory devices. [00:33:23] Speaker 02: Where? [00:33:24] Speaker 07: Would you show us then? [00:33:25] Speaker 02: Yeah. [00:33:25] Speaker 02: Looking for that. [00:33:27] Speaker 02: Because that was one of the issues that quite frankly came up during the trial before the PTAB. [00:33:31] Speaker 02: So I'm confident that that was presented. [00:33:35] Speaker 02: But while we're looking for that citation, let me get to my second argument, which is, and I think disappointed, might have come up in the argument already. [00:33:46] Speaker 02: I think the dispute that Natalys raised, while it's being characterized as a dispute of claim construction appeal, was really a dispute on whether the claim meaning construction they proposed and the board's construction actually applied to the references. [00:34:02] Speaker 02: That is not a claim construction dispute. [00:34:04] Speaker 02: That is a factual finding for the board to make. [00:34:08] Speaker 02: And that is not reviewed de novo. [00:34:09] Speaker 02: That is reviewed for clear error. [00:34:13] Speaker 02: natural size for the interstainer case? [00:34:16] Speaker 05: Well, it depends on how you view the constructions, don't you? [00:34:18] Speaker 05: If you view the board's construction as your friend does, then the board's construction concentrates on selecting the memory device, not electrically coupling or decoupling from that memory device. [00:34:32] Speaker 05: And it seems that there is a distinction there in the patent. [00:34:36] Speaker 05: I mean, one, maybe the whole point of [00:34:39] Speaker 05: selectively coupling or decoupling electrically is to activate that device, but it has the additional purpose, at least from their arguments and their patent and specification, of reducing load and things like that. [00:34:54] Speaker 05: Whereas if the line is always electrically connected, but you just use the selection from this prior art to go in and get data from it, it's different than being electrically isolated. [00:35:09] Speaker 02: I would like to make a couple of points on that. [00:35:12] Speaker 05: If I've got the science wrong, I'm sorry. [00:35:14] Speaker 05: This is why we need... You think you've characterized the argument correctly? [00:35:19] Speaker 02: There are two problems with that. [00:35:21] Speaker 02: And I think there was a question during council's argument whether coupling simply creates the path, and then the transmission of the signal happens thereafter. [00:35:32] Speaker 02: The answer is no, because the claim requires electrical coupling, not just coupling. [00:35:36] Speaker 02: And as Courtois knows, coupling is a very broad term. [00:35:39] Speaker 02: Electrically, as the word construction says, actually requires transfer of electrons of electricity. [00:35:46] Speaker 02: That's the electrical. [00:35:47] Speaker 02: It's like a switch. [00:35:48] Speaker 02: In fact, the switch, Netflix admits, is the primary embodiment described in the pot. [00:35:52] Speaker 02: If you turn the switch on, there is electrical flow. [00:35:54] Speaker 02: If you turn the switch off, there is no electrical flow. [00:35:57] Speaker 02: So it's operation of the circuit. [00:36:00] Speaker 02: It's not just creating the path or not. [00:36:03] Speaker 02: And how do we know that? [00:36:04] Speaker 02: We look at the structure of the claim. [00:36:06] Speaker 02: In claim 15, we have a circuit [00:36:09] Speaker 02: with three required elements, the logic element, the register, and the PLL. [00:36:14] Speaker 02: And it says wherein the circuit is configurable to, and one of the things it's configurable to do is selective electrically coupling or selective isolating. [00:36:23] Speaker 02: So the phrase that we're construing here is part of the operation of the circuit. [00:36:28] Speaker 02: It's not part of the structure of the circuit. [00:36:31] Speaker 02: And what Netflix is arguing is really they want to add a structural requirement, either a switch or some other type of device being required to perform that function. [00:36:43] Speaker 02: But the inventors knew how to set forth structural limitations in the claim, and they knew how to set forth functional limitations. [00:36:50] Speaker 02: And they chose to include this sentence in the functional limitations. [00:36:56] Speaker 07: Do you have the reference to the testimony? [00:37:07] Speaker 02: I think the best one is on appendix 1285. [00:37:10] Speaker 02: 1285, which volume is it? [00:37:13] Speaker 02: It's in volume one, I believe. [00:37:32] Speaker 02: And that's the very beginning of paragraph 86. [00:37:35] Speaker 02: That's Dr. Jagannathan's testimony. [00:37:38] Speaker 02: The switch was taught by Klein electrically coupled data signals from a selected memory element, not a memory module, to the common data signal bus and do so in response to the input signals received. [00:37:51] Speaker 02: So he's actually testifying. [00:37:53] Speaker 05: How do we know memory element means memory device and not memory module? [00:38:00] Speaker 05: Because in the paragraph before, when it's talking about memory device, it says memory device. [00:38:07] Speaker 05: And the sentence before that, it uses memory element, presumably to mean something different. [00:38:23] Speaker 02: Well, so he also says in that same, he doesn't address it squarely. [00:38:26] Speaker 02: That is the answer. [00:38:27] Speaker 02: But in the same paragraph, he says, one ordinary skill would understand that decoder is a logic element. [00:38:33] Speaker 02: Element, as its ordinary meaning, means one thing. [00:38:37] Speaker 02: So I think in this, in circuitry, you're talking about one of the things in the figure. [00:38:42] Speaker 05: But it could mean one module or one device within the module. [00:38:47] Speaker 02: I wouldn't read it that way, to be honest. [00:38:49] Speaker 02: I don't know that there is a contention. [00:38:50] Speaker 07: It says it's construed in section 3C, paragraph 1516. [00:38:54] Speaker 07: Where is that? [00:38:58] Speaker 07: I'm sorry, I didn't catch that. [00:38:59] Speaker 07: It says on page 1285, so one ordinary skill in the art would understand that the coder is a logic element. [00:39:07] Speaker 07: is construed in Section 3C. [00:39:10] Speaker 07: This is page 1230 and 1231. [00:39:12] Speaker 02: Yeah, so there is on page 1231, there's a definition of logic element. [00:39:32] Speaker 02: And I'm going to perform some kind of logic function or an element that comprises a logic circuit. [00:39:39] Speaker 02: So it's suggesting the word element usually is used to refer to an individual component, not to a whole memory module. [00:39:47] Speaker 02: So I agree that 62 has to find these figures. [00:39:57] Speaker 02: I know, but how does it relate to the testimony? [00:40:01] Speaker 02: Because this is from Klein. [00:40:03] Speaker 02: What does he say in the testimony? [00:40:11] Speaker 02: In the actual body of the patent for Klein. [00:40:15] Speaker 02: Is that the patent for Klein? [00:40:16] Speaker 02: Yes. [00:40:17] Speaker 02: They actually use the word element to describe the memory device as 62. [00:40:23] Speaker 02: And that is? [00:40:26] Speaker 02: In paragraph 33, page 1637 of the appendix. [00:40:38] Speaker 02: And I think that makes that point clear. [00:40:41] Speaker 02: So the memory elements are 62, which are the blocks in figure 6 that I was referring to. [00:40:48] Speaker 02: So now we know that those transfer gates, plural, [00:40:51] Speaker 02: are each one for controlling a specific memory element or memory device 62 and figure 6 of 5. [00:40:57] Speaker 02: That's part of the testimony that the board had available to it. [00:41:02] Speaker 02: When it made the determination in the 8A2 inter-partis review that, in fact, the combination of a MIDI, one memory module with multiple memory devices that has all the elements of the circuit [00:41:16] Speaker 02: But for what Netlis argues is a requirement potentially for a switch or some other kind of device to decouple and couple the lines selectively, the board took the teaching of Klein of MOSFET switches to modify the immediate figure in the middle of the data lines to provide a switch for each memory device. [00:41:36] Speaker 02: And that is the integration board made. [00:41:38] Speaker 02: I think we're about out of time. [00:41:39] Speaker 02: Thank you. [00:41:40] Speaker 02: Thank you. [00:41:51] Speaker 07: So, Mr. Lloyd, if the petitioner's expert said that Klein showed control of the individual devices and not just the module, was there testimony to the contrary by your expert? [00:42:11] Speaker 07: That testimony that we looked at earlier didn't seem to contradict that. [00:42:16] Speaker 01: I think it does contradict it, Judge Dyke, because [00:42:19] Speaker 01: Again, the testimony that was just presented here went to, is the switch connected to the memory devices? [00:42:25] Speaker 01: But again, our claims are more specific. [00:42:28] Speaker 01: Our claims require selectively electrically coupling a device data signaling to a common signal line. [00:42:34] Speaker 01: And what Klein shows, even if switches are connected to individual memory devices, there's no disclosure of operating those switches in a way to selectively electrically couple the individual devices. [00:42:48] Speaker 01: On the figure six, for example, this is the paragraph. [00:42:52] Speaker 07: Where does your expert contradict this? [00:42:54] Speaker 07: Say what you just said. [00:42:56] Speaker 01: So one is the testimony we read before, where what he said is that Klein operates the switches at the module level. [00:43:06] Speaker 07: But your experts seem to concede that maybe it also shows operating with respect to individual devices. [00:43:15] Speaker 01: No, I don't think that's right. [00:43:17] Speaker 01: What page were we at? [00:43:18] Speaker 01: Let me pull it up. [00:43:25] Speaker 01: Sorry, notes are over here. [00:43:30] Speaker 01: And it's 21, 22. [00:43:31] Speaker 01: 21, 22. [00:43:41] Speaker 00: Okay. [00:43:45] Speaker 01: Well, it actually, in all the shown embodiments that can selectively decouple, it's decoupling a whole memory module. [00:43:52] Speaker 01: That is the granularity. [00:43:54] Speaker 01: So he's not talking about necessarily whether the switches are connected to individual memory devices. [00:43:59] Speaker 01: He's talking about how the coupling or decoupling operates. [00:44:02] Speaker 01: And it's operating at the whole memory module level. [00:44:06] Speaker 07: But then on the next page, in the sense that Judge Hughes referred to, he says, well, yeah, one could argue, I suppose, if you want to figure 10, [00:44:14] Speaker 07: there's a possibility of decoupling a memory device. [00:44:17] Speaker 01: That's right. [00:44:18] Speaker 01: So there is an embodiment where the switches are built into the memory devices. [00:44:23] Speaker 01: And that certainly creates the possibility, as he admits, of coupling or decoupling the devices. [00:44:28] Speaker 01: But then he continues on. [00:44:29] Speaker 01: At the end of the day, it decouples or couples an entire memory module, because that's the embodiment that meets the objectives of the summary of the invention. [00:44:37] Speaker 05: Is what you're saying that the difference from what he's describing here in your invention is, [00:44:44] Speaker 05: The switch incline is in the memory module itself to the extent it can control devices, and therefore it's not shutting off the power from the controller level, which is where you're shutting off the power from. [00:45:00] Speaker 01: I think maybe slightly differently from that, Judge Hughes. [00:45:03] Speaker 01: It's not necessarily about where the switches are located or even what they're connected to, but it's about whether the switches selectively couple [00:45:10] Speaker 01: a signal line from the memory device to a common signal line. [00:45:14] Speaker 01: And the reason that matters, I mean, here's the takeaway. [00:45:17] Speaker 05: Wait, let me just make sure I've got this. [00:45:19] Speaker 05: So under that explanation, even if Klein selects a memory device, your view is it's not uncoupling or coupling to the common signal line. [00:45:33] Speaker 05: The common signal line is still either, is all on. [00:45:38] Speaker 01: Even if Klein [00:45:39] Speaker 01: has switches connected to individual memory devices. [00:45:43] Speaker 01: I think that's the only part of what you said that I'm not sure about. [00:45:47] Speaker 01: Even if it has switches connected to individual memory devices, it's coupling or decoupling all of the devices at the same time. [00:45:53] Speaker 01: That's what our expert said. [00:45:55] Speaker 01: In the paragraph, Judge Shrano, that you pointed to, paragraph 35, incline, it's not particularly clear that that's at 16. [00:46:03] Speaker 07: But that's not really what he says on the line. [00:46:05] Speaker 07: He says you could argue, I suppose, [00:46:07] Speaker 07: there's a possibility of decoupling a memory device. [00:46:12] Speaker 01: I agree, Judge. [00:46:14] Speaker 01: Because of where the switches are placed, there is a possibility. [00:46:18] Speaker 01: But he doesn't say that that's how Klein was configured to operate. [00:46:22] Speaker 01: And again, everything we're talking about, these are factual issues that the board simply never reached. [00:46:28] Speaker 01: All of its findings were under its construction. [00:46:31] Speaker 01: And on Appendix 32, the page that my counsel on the other side focused on, the very next page, Appendix 33, [00:46:38] Speaker 01: The board discredits our expert for relying on our construction and not the board's. [00:46:43] Speaker 01: These are issues that the board should reach in the first instance. [00:46:45] Speaker 01: There are factual disputes on this. [00:46:48] Speaker 01: And so we ask that the court vacate and remand. [00:46:52] Speaker 07: All right. [00:46:52] Speaker 07: I think we're out of time. [00:46:53] Speaker 07: Thank both counsel. [00:46:54] Speaker 07: The case is submitted.