[00:00:01] Speaker 03: Our next case is Weiland versus Apple Inc. [00:01:12] Speaker 03: As a preliminary matter, we received a request this morning to extend the time for oral argument. [00:01:20] Speaker 03: And the court has conferred on the request, and we've determined that to deny the request. [00:01:26] Speaker 03: So each side is going to have the 15 minutes to argue this for him. [00:01:32] Speaker 03: Mr. Cote, is that correct? [00:01:34] Speaker 04: Mr. Cote, yes. [00:01:34] Speaker 03: Cote, yes, sir. [00:01:35] Speaker 03: Are you ready? [00:01:36] Speaker 04: Yes, Your Honor. [00:01:37] Speaker 03: You're reserving five minutes for rebuttal time, correct? [00:01:40] Speaker 03: Correct, Your Honor. [00:01:42] Speaker 04: So it's Robert Cody for Appellant Weiland. [00:01:45] Speaker 04: May it please the court. [00:01:47] Speaker 04: The district court below. [00:01:48] Speaker 01: Mr. Cody, let me throw three questions at you, but I look at it as one, OK? [00:01:56] Speaker 01: So one, where does the patent mention complex multiplier? [00:02:03] Speaker 01: Where is it in the district court's original construction? [00:02:06] Speaker 01: And was that phrase raised in connection with the jury's instruction? [00:02:11] Speaker 01: So they're all complex multiplier questions. [00:02:14] Speaker 04: Correct. [00:02:14] Speaker 04: So if you look at the corresponding structure identified by the court, you'll see that both figures one and four are identified, along with citations to the specification. [00:02:24] Speaker 04: If you look at those citations, you'll find that there is a reference to the figure four description, would be column two lines 58 to 62 of the patent. [00:02:35] Speaker 04: At column two lines 58 to 62 of the patent, [00:02:38] Speaker 04: It actually identifies that the figure three transforms should be included as part of the corresponding structure. [00:02:45] Speaker 04: And you'll recall in figure four that we're talking about a multi-stage transform structure. [00:02:52] Speaker 04: And it just calls those endpoint transforms. [00:02:55] Speaker 04: To understand what those endpoint transforms are, the description of figure four, again, at column two, lines 58 to 62, tells you that those transforms include the transforms in figure three. [00:03:07] Speaker 04: And the patent will teach you at column 4, line 66, to column 5, line 12. [00:03:16] Speaker 04: Each of the transforms that you may use, each of the transforms, with the exception of one, the randomizer transform, is a multi-code transform that will do the multi-code spreading, which is the key to the invention. [00:03:28] Speaker 04: Turn a single-lane highway into a multi-lane freeway by using these multi-code spreading transforms. [00:03:34] Speaker 04: Those are all orthogonal transforms. [00:03:36] Speaker 04: to allow the lanes not to interfere with each other so that on the other end you get back the data that was in each of the highway lanes. [00:03:42] Speaker 04: So the essence of the invention is multi-code spreading. [00:03:45] Speaker 04: And then the other essence of the invention is if I want to combine these signals and actually replicate the transmission from a cell phone, I get these spikes when I combine all the highway lanes. [00:03:56] Speaker 04: It's kind of like you think of a signal. [00:03:57] Speaker 04: If the signals line up, you get a spike. [00:04:00] Speaker 04: It would take a stereo receiver to make it work in a cell phone. [00:04:04] Speaker 04: So the other key to the invention in the claim is the idea of complex randomization, which will actually eliminate the spike or greatly reduce it so that I can use a low-cost linear amplifier in a cell phone so I can get multi-code spreading in a cell phone. [00:04:19] Speaker 01: So you decided all that language. [00:04:21] Speaker 01: Where's complex multiplier? [00:04:23] Speaker 04: Well, so in the transforms that are identified as part of the figure four embodiment, [00:04:30] Speaker 04: You will see that, let me pull out the column. [00:04:36] Speaker 04: You go to four. [00:04:46] Speaker 04: You will see at the bottom of column four, up to the first portion of column five, you'll see a number of transforms. [00:04:54] Speaker 04: With the exception of one, they're all orthogonal spreading transforms that do the multicode. [00:05:00] Speaker 04: The only transform for randomization is the randomizer transform illustrated in Figure 8. [00:05:06] Speaker 04: So the corresponding structure identified the description of Figure 4 and that it must include the transforms in Figure 3. [00:05:15] Speaker 04: And those are all identified here in the portion that I've identified. [00:05:19] Speaker 04: And both experts agreed at trial that, in fact, the complex randomizer was the only structure, the Figure 8 structure, disclosed in the patent. [00:05:27] Speaker 04: and that it was part of the corresponding structure. [00:05:29] Speaker 04: And that's important. [00:05:30] Speaker 04: The expert didn't say it wasn't in the construction. [00:05:33] Speaker 04: He said it was part of the corresponding structure, their own expert. [00:05:37] Speaker 04: And so for the reasons I just identified, the cited passages. [00:05:40] Speaker 01: So that antedates the construction. [00:05:48] Speaker 01: So that again? [00:05:49] Speaker 01: That expert agreement that you argue antedates the original construction, right? [00:05:57] Speaker 01: So it's not in the construction that the court does. [00:06:00] Speaker 04: Well, it is. [00:06:01] Speaker 04: The expert agreed that the cited portions from the patent itself, the cited portions from the patent, taught that he transforms to use. [00:06:11] Speaker 05: I'm not following. [00:06:12] Speaker 05: You speak very fast. [00:06:14] Speaker 05: So there's nothing in figure one or figure four that actually requires complex numbers. [00:06:20] Speaker 05: You're getting me to the specification. [00:06:23] Speaker 04: Right. [00:06:23] Speaker 04: And so if I go to column two. [00:06:25] Speaker 05: Bottom of page four, it says preferred embodiments. [00:06:29] Speaker 05: Right? [00:06:29] Speaker 05: Is that where you were reading from? [00:06:31] Speaker 05: Am I reading from the right place? [00:06:33] Speaker 04: Column two of the bank. [00:06:34] Speaker 05: Well, you directed us to column four earlier. [00:06:38] Speaker 04: Well, just to go through it one more time. [00:06:42] Speaker 04: If you look at figure four, you'll see there are a series of Ns. [00:06:45] Speaker 05: No, I've got figure four. [00:06:47] Speaker 05: But you have to admit, there is nothing in figure four that says complex numbers. [00:06:51] Speaker 04: That is correct. [00:06:52] Speaker 04: However, there are cited portions from the patent to go along with figure four, so you know what the endpoint transforms are. [00:06:59] Speaker 04: You wouldn't know what any of the transforms are, whether it's the randomizer transform that does randomization, or which spreading transforms to use, unless you went to the cited pattern. [00:07:09] Speaker 05: So in column two, we say figure four is the schematic [00:07:12] Speaker 05: showing, and then there's a bunch of stuff. [00:07:14] Speaker 05: Is there something in there that says it has to have complex numbers? [00:07:17] Speaker 04: So figure four is a drawing that uses the codes generated in figure three. [00:07:24] Speaker 04: When you go to the discussion in the patent at the bottom of column four, through the top of column five, you'll see it says the examples of the transformers. [00:07:34] Speaker 05: This is where you're speaking too fast and I can't follow it. [00:07:38] Speaker 05: That's where it says preferred embodiments. [00:07:40] Speaker 05: Is that where you're reading from? [00:07:42] Speaker 05: the language I prefer to write. [00:07:44] Speaker 05: And then it says examples of the endpoint transforms in figure three are a bunch of words that are all kinds of different transforms. [00:07:54] Speaker 04: Except for one, which is the randomizer transform of figure eight. [00:07:57] Speaker 04: That is the only transform that does [00:07:59] Speaker 05: complex randomization is the only... But a bunch of these don't require complex. [00:08:03] Speaker 05: Say again? [00:08:04] Speaker 05: A bunch of these don't require complex numbers. [00:08:05] Speaker 04: Well, there are two steps, right? [00:08:07] Speaker 04: There's the spreading step and the randomization step. [00:08:09] Speaker 04: The spreading uses any one of the other transforms. [00:08:12] Speaker 04: The experts agree those are spreading transforms, multi-code spreading transforms. [00:08:18] Speaker 04: And the only transform in the patent identified, both experts agree, is the randomizer transformer figure eight. [00:08:24] Speaker 04: And that's what's identified here. [00:08:25] Speaker 04: So one of ordinary skill in the art would understand reading this, [00:08:28] Speaker 04: that this is teaching you here are your options for your spreading transforms. [00:08:33] Speaker 04: That's your modulation of your data symbols. [00:08:35] Speaker 04: And here is the option for your complex randomization, the randomizer transform in Figure 8. [00:08:40] Speaker 04: And those are the two pieces of the invention, spreading to modulate. [00:08:44] Speaker 05: I'm sorry. [00:08:45] Speaker 05: Sure. [00:08:46] Speaker 05: This is hard. [00:08:47] Speaker 05: But when it says examples of the endpoint transforms in Figure 3, maybe I should look at Figure 3, because you're talking about Figure 4. [00:08:57] Speaker 05: Is that the same thing as figure? [00:09:00] Speaker 04: Well, so just so you guys, so that you understand. [00:09:04] Speaker 05: I have figure four in front of me, and that's what I've been looking at, because that's what you all cite. [00:09:09] Speaker 05: That has a number of transforms in it. [00:09:11] Speaker 05: It has up to as n number of many as you want, but then it doesn't say that any of those have to be complex. [00:09:21] Speaker 04: Well, we're dealing with means plus function terms. [00:09:25] Speaker 04: And so you go to the patent to see what structures perform the function, what's necessary to performing the function. [00:09:32] Speaker 04: The only structure identified that does randomizing is descriptively called the randomizer transform of figure 8. [00:09:40] Speaker 04: And so it's clearly linked to the function. [00:09:42] Speaker 04: In the patent, it's clearly linked to the function of randomizing the figure 8 structure. [00:09:47] Speaker 05: And I think where the confusion lies, I could- So you're saying that none of these other transforms do randomizing ether? [00:09:52] Speaker 05: Correct. [00:09:52] Speaker 04: And both experts agree. [00:09:54] Speaker 05: I thought that one of the spreading itself is randomizing. [00:09:58] Speaker 05: I mean, the use of pseudo-random code is a randomization. [00:10:03] Speaker 04: The pseudo-random code is not complex randomization. [00:10:08] Speaker 04: That is distinct. [00:10:09] Speaker 05: Well, you're not answering the question. [00:10:12] Speaker 05: You're presuming the answer. [00:10:14] Speaker 05: that it requires complex randomization. [00:10:16] Speaker 05: But it seems to me that I don't know what any of these formulas do, but I do know what pseudorandom code is to a certain extent. [00:10:23] Speaker 05: And that is a form of randomization, is it not? [00:10:25] Speaker 04: It is, but not complex randomization. [00:10:27] Speaker 04: So figure 8 does complex randomization. [00:10:31] Speaker 04: And you need the complex randomization to get the spikes out of the signal so you can put it into a cellular device. [00:10:36] Speaker 03: This is the panel with respect to figure 3. [00:10:41] Speaker 03: That's what we're looking at. [00:10:42] Speaker 03: say that any other reversible transform can apply. [00:10:50] Speaker 03: It doesn't say that it must be a complex multiplier. [00:10:55] Speaker 04: Well, so that's where a patent is viewed through the eyes of one of ordinary skill in the art. [00:11:00] Speaker 04: And both experts agreed that one of ordinary skill in the art, reading this patent, would understand. [00:11:05] Speaker 05: Can you put me to the specific lines for this specification? [00:11:08] Speaker 05: that you think requires complex randomization? [00:11:13] Speaker 04: If you go to column five, if you go to line three, I'm sorry, line two, it says a randomizer transform is the one illustrated in figure eight. [00:11:28] Speaker 04: And if you go to figure eight, the structure of figure eight, [00:11:32] Speaker 04: Well, it doesn't say the words complex randomization. [00:11:35] Speaker 05: Wait, but OK, I get your argument. [00:11:39] Speaker 05: I agree. [00:11:41] Speaker 05: Figure 8 is a complex transformation. [00:11:44] Speaker 05: But this says, for examples of the endpoint transforms in figure 3, a whole list. [00:11:50] Speaker 05: It doesn't require a randomizer transform. [00:11:54] Speaker 05: It requires one of these or one or more of these whole lists. [00:11:58] Speaker 05: So why isn't the use of one or more of these and not [00:12:02] Speaker 05: the randomizer transform into your eighth sufficient structure? [00:12:06] Speaker 04: Because this is where patents are looked through the eyes of one of ordinary school in the art. [00:12:11] Speaker 04: Each expert agreed that the randomizer transform was the only structure disclosing the patent for meeting the function. [00:12:17] Speaker 04: Remember, I have to link to the function. [00:12:19] Speaker 04: I've got to spread, and I have to have to randomize. [00:12:22] Speaker 04: The randomizing function is achieved, both experts agreed, by the complex randomizer and figure eight. [00:12:27] Speaker 04: And while there's a list, one of ordinary skill in the art, and they both agreed, would understand that is the transform for randomizing to meet the function in the claim. [00:12:36] Speaker 04: But the other transforms are also understood to be spreading transforms. [00:12:39] Speaker 04: That's, first I take my data symbols and I put them on different highway lanes. [00:12:44] Speaker 04: I spread them on highway lanes using one of these transforms. [00:12:48] Speaker 05: That would be a reasonable reading [00:12:50] Speaker 05: What it says was, examples of the spreading things are this whole list. [00:12:54] Speaker 05: And then examples of the randomizing are this list. [00:12:57] Speaker 05: But what it says, examples of the endpoint transforms in figure 3, are any of this list. [00:13:03] Speaker 04: Well, the literal language is as you described, your honor. [00:13:09] Speaker 03: I will agree, but one more thing. [00:13:11] Speaker 03: Sure. [00:13:11] Speaker 03: To add to that goes back to, we're looking at the language in lines 1, 2, 3, and 4 of column 5. [00:13:20] Speaker 03: And I pointed you to line six. [00:13:23] Speaker 03: This is in any other reversible transform. [00:13:27] Speaker 03: So it's speaking about reversible transforms. [00:13:29] Speaker 03: It gives you a list, but nowhere does it talk about a complex transform. [00:13:34] Speaker 04: Well, the randomized transform in figure eight that's identified is one of the example, one of the transforms you can use. [00:13:41] Speaker 04: actually is the complex randomizer. [00:13:43] Speaker 04: Both experts agreed. [00:13:45] Speaker 05: And while there's a list. [00:13:46] Speaker 05: Can I pull back a little bit? [00:13:48] Speaker 05: Sure. [00:13:48] Speaker 05: Because let's just assume that we disagree with you on the claim of construction. [00:13:52] Speaker 05: It seems like the district court actually did. [00:13:55] Speaker 05: I don't know. [00:13:58] Speaker 05: I didn't follow it very well. [00:13:59] Speaker 05: And I didn't find the reasoning particularly persuasive, but relied at least in part on expert testimony to get to that claim of construction. [00:14:08] Speaker 05: Even if we disagree, is that something that presents a Teva issue for us about deference? [00:14:16] Speaker 04: I don't believe so. [00:14:20] Speaker 05: You mean the... So if we think the claim construction is wrong, we can just reverse it and send it back. [00:14:28] Speaker 04: If you think the claim construction is wrong, you could. [00:14:30] Speaker 04: Correct. [00:14:31] Speaker 04: You could do that. [00:14:33] Speaker 04: The claim construction based on the deference is right. [00:14:36] Speaker 05: I'm sorry. [00:14:36] Speaker 05: Maybe I'm not being clear. [00:14:38] Speaker 05: I think you're answering the question adverse to your interest, but I see this as a disputed claim construction issue that was at least in part relied, the district court relied upon expert testimony to reach his claim construction. [00:14:52] Speaker 05: Under TEVA, aren't we required to defer to that claim construction since it's based upon extrinsic evidence? [00:15:01] Speaker 05: Yes, Your Honor. [00:15:06] Speaker 05: Okay. [00:15:08] Speaker 05: It does seem to me a little bit problematic, the claim construction. [00:15:12] Speaker 05: Even if that's the case, though, well, you know what? [00:15:15] Speaker 05: We've taken up too much of your time on this issue. [00:15:17] Speaker 05: Do you want to move on to something else? [00:15:19] Speaker 05: Yeah. [00:15:20] Speaker 04: I think the theme here today is that intellectual property trial should be about an intellectual property right, as the court has defined it. [00:15:27] Speaker 04: And then the jury should ask the question, is that intellectual property right found in the accused product? [00:15:33] Speaker 04: In this particular case, the district court heard in not enforcing its claim constructions [00:15:37] Speaker 04: At trial, Apple urged two limitations that are not found in the court's claim construction. [00:15:44] Speaker 05: Well, it seems to me you're arguing one way on one claim construction and one way on the other one. [00:15:49] Speaker 05: On one claim construction you say not enforcing it more explicitly and then on the other one that [00:15:55] Speaker 05: goes in your favor, you're saying he can't add an additional language after the fact. [00:15:59] Speaker 04: But the difference is on the complex multiplier, both experts agreed that it was part of the corresponding structure explicitly. [00:16:05] Speaker 04: And in this case, we don't agree. [00:16:06] Speaker 05: Was that same construction ever presented to the jury? [00:16:10] Speaker 04: The complex multiplier? [00:16:11] Speaker 04: Yes, absolutely. [00:16:13] Speaker 01: Was it part of the jury instructions? [00:16:16] Speaker 04: Well, the figure one and four embodiments plus the description. [00:16:20] Speaker 05: No, not the embodiments and the specification. [00:16:21] Speaker 05: The district court made a specific claim construction [00:16:24] Speaker 05: on this disputed claim language, didn't he? [00:16:26] Speaker 05: And it didn't use the words complex multiplier. [00:16:30] Speaker 05: He added that in on his JMAL. [00:16:33] Speaker 04: He said it wasn't explicitly stated to be a complex multiplier. [00:16:37] Speaker 04: But I don't believe that he said he pointed to the fact that both experts agreed that the corresponding structure included the complex multiplier. [00:16:47] Speaker 04: And both our experts agreed that what was stated in the court's construction, at least on cross from [00:16:53] Speaker 04: of their expert, he agreed as well, that it was included within the transforms identified that the corresponding structure linked to the transform. [00:17:01] Speaker 05: So if the jury was making its factual findings on invalidity, was it explicitly aware that it had defined prior art, that it had a complex multiplier? [00:17:11] Speaker 04: Yes. [00:17:12] Speaker 05: Where in the record? [00:17:14] Speaker 04: It would have been explained to the jury as part of the trial as well as the closing arguments by both sides. [00:17:21] Speaker 04: that the claim required, the corresponding structure required, a complex multiply. [00:17:26] Speaker 04: So they were well aware. [00:17:27] Speaker 03: Before you sit down, I'd like to hear some argument on the sequencing of the modulated data symbols. [00:17:33] Speaker 03: Can you address that, please? [00:17:35] Speaker 04: Yes. [00:17:37] Speaker 04: So we're dealing with means plus function language, means to combine modulated data symbols. [00:17:43] Speaker 04: And the court below found that the modulated data symbols need only be spread [00:17:49] Speaker 04: And so taking the court's construction as it was given, as it was given, there's no question and there's no dispute that Apple's products combine modulated data symbols. [00:18:01] Speaker 04: The question is, should the error that occurred was the court did not enforce its construction before the jury or on JMAW and had it, there'd be no question that a sequence is not required. [00:18:15] Speaker 04: And the sequence is, [00:18:17] Speaker 04: Does the modulated data symbols, the means to combine the modulated data symbols, have to happen last? [00:18:22] Speaker 04: It does not if modulated data symbols mean spread data symbols as the court construed the term. [00:18:28] Speaker 03: Where the pen reads and combines the modulated data symbols, what does D refer to? [00:18:34] Speaker 04: It refers back to the modulated data symbols in the first computing means. [00:18:39] Speaker 03: So doesn't that require a certain sequence? [00:18:42] Speaker 04: It does not because we're dealing with means plus function claims. [00:18:46] Speaker 04: And the means, the first computing means, if you look at figure four, has a stage where you modulate and a stage where you randomize, two endpoint transforms. [00:18:57] Speaker 04: The experts agreed the Qualcomm Prior Art, the Guildhausen Prior Art, and the Qualcomm chip specifications, which talk about how the product works. [00:19:06] Speaker 04: Everybody agreed, both experts. [00:19:08] Speaker 04: that how you order the combining of the modulated data symbols, whether or not you do it at the end after the computing means is done as modulating or spreading and done as randomizing, or whether you include the means to combine within the modulation and between the modulation and the randomization within the first computing means. [00:19:28] Speaker 04: Either order gives you the exact same output. [00:19:30] Speaker 04: And so when we look at means plus function claims, we ask, what structure is necessary for performing the function? [00:19:36] Speaker 04: It's a necessary test. [00:19:39] Speaker 04: And if both experts agree that it's not necessary to have the means to combine last, and that it can occur between the modulator and the randomizer within the first computing means, then you would not read a requirement of a sequence into these means plus functions. [00:19:56] Speaker 03: OK, I think I got it. [00:19:58] Speaker 03: Any other questions? [00:19:59] Speaker 03: All right, thank you very much. [00:20:01] Speaker ?: OK. [00:20:08] Speaker 03: Mr. Davies, so you divided your time into, you're reserving two minutes for your cross appeal, is that correct? [00:20:16] Speaker 00: I'm at your service, Your Honor. [00:20:18] Speaker 00: If we can talk about validity in this argument, then we can do it all now. [00:20:21] Speaker 03: I'm going to add two minutes here, 13 minutes. [00:20:25] Speaker 05: I'm here for you, Your Honor. [00:20:26] Speaker 03: You may proceed. [00:20:26] Speaker 05: Can I start with my Teva question? [00:20:28] Speaker 05: Yes. [00:20:29] Speaker 05: Because I think, honestly, that I have some big problems with the district court's construction, but it also seems to me that [00:20:37] Speaker 05: he did rely on expert testimony. [00:20:40] Speaker 05: So I'm not sure that I can do anything about it. [00:20:43] Speaker 05: Setting aside the question of whether this was presented to the jury, which I don't see that it was, so it needs to go back. [00:20:51] Speaker 05: But it seems to me that we can't reverse the district court's construction outright, because we have to defer to its factual findings of what a person of the ordinary skill would find, unless we find them clearly erroneous. [00:21:04] Speaker 00: But happily, there was actually a clear answer on that, Your Honor, is that it was not a claim construction. [00:21:10] Speaker 00: What the district court did was cited expert testimony, but never keyed that testimony to a particular term. [00:21:16] Speaker 00: There is no construction that you have to defer to under TEVA, would be the straightforward answer. [00:21:22] Speaker 00: Because he used the word complex multiplier, he cited our expert, but he never keyed it back to the terms in the patent. [00:21:28] Speaker 00: So there was not a TEVA issue, because it's not a claim construction issue. [00:21:31] Speaker 00: But beyond that, you can't have an expert come in and just start adding words to patents. [00:21:37] Speaker 00: That's not what the public relies on. [00:21:38] Speaker 00: It relies on what's in the patent. [00:21:40] Speaker 00: So I guess you could also go off the clearly erroneous rationale as well. [00:21:44] Speaker 00: I mean, there is no complex multiplier as you started. [00:21:47] Speaker 00: There's no complex multiplier required in the patent. [00:21:50] Speaker 00: And that's what the district court started out by saying. [00:21:53] Speaker 00: That's at Appendix 73. [00:21:54] Speaker 00: Everybody agreed there was no complex multiplier in the patent. [00:21:57] Speaker 00: The jury was never told. [00:21:59] Speaker 00: And the jury instructions are [00:22:01] Speaker 00: or 8th, appendix 3377, Your Honor. [00:22:05] Speaker 00: Our export at 1033 makes it clear that there is no complex multiplier required. [00:22:10] Speaker 00: Claim 8, I think, would also help Your Honor's understanding there is no complex multiplier required. [00:22:15] Speaker 00: The only reason the district court got there was our export testimony, which we think was misunderstood, and I can elaborate on that. [00:22:21] Speaker 00: But TEVA itself, if TEVA stands for anything, it certainly stands for a wide swath of reading the patent itself. [00:22:29] Speaker 00: I think you could really stop and end there. [00:22:31] Speaker 00: I have many more things to say, but the sequencing argument, your honor, I think is one of, I was rereading your Motorola decision, and it's really very directly on point here. [00:22:45] Speaker 00: The appendix 1026, our expert told the jury, the language of the claim itself tells us it's an order. [00:22:52] Speaker 00: First you randomize, and then you combine. [00:22:54] Speaker 00: And that just follows naturally from the reading of the claims, very similar to the Apple Motorola case. [00:23:00] Speaker 00: It's confirmed by the figures, [00:23:01] Speaker 00: and the Oak Tech case is also directly on point. [00:23:05] Speaker 00: So that's the sequencing rationale. [00:23:07] Speaker 00: We have the validity that we've been over, and there is this complex multiplier issue that I'm happy to answer questions on as well. [00:23:18] Speaker 05: Can I ask you about the sequencing issue? [00:23:21] Speaker 05: So to the extent I understand your friend's argument, it is that even though the structure may have, and the claims may have a certain sequence, [00:23:31] Speaker 05: it's not necessary to the invention, so they can get there through the doctrine of equivalence. [00:23:38] Speaker 05: So that seems like a very odd argument to me, though, when we're in the means plus function world, because the means plus function world means you have to provide the specific structure. [00:23:48] Speaker 05: So how could you have a doctrine of equivalence for a structure that's different than the structure and the pattern? [00:23:56] Speaker 00: Yeah, you could not. [00:23:57] Speaker 00: You could not. [00:23:58] Speaker 00: And obviously, it would be a jury call. [00:24:00] Speaker 00: To the extent you would make the argument, it would be a jury call. [00:24:02] Speaker 00: And here we have the briefing. [00:24:05] Speaker 00: We didn't get a final brief, so the briefing is a little unclear on this. [00:24:07] Speaker 00: But the actual exchange in the record of 1054 is, their suggestion is, well, there's 300 million transistors. [00:24:15] Speaker 00: There's a little bit of a change of 20, so that doesn't really matter. [00:24:20] Speaker 00: But our expert comes right back and tells the jury, no, that's a big deal in our universe, that that's actually a 10% optimization of the particular sections that a team's in charge of. [00:24:29] Speaker 00: And that, if anything, is a jury call, Your Honor. [00:24:31] Speaker 00: So if there's anything to defer to, I don't think it's a TAVA point. [00:24:34] Speaker 00: It's this doctrine of equivalence finding by the jury, expressly rejecting their doctrine of equivalence argument. [00:24:40] Speaker 00: So recall, we have a jury verdict. [00:24:43] Speaker 00: We're not up [00:24:44] Speaker 00: on anything unusual. [00:24:46] Speaker 00: We have a jury that sat for a week, eight people. [00:24:48] Speaker 00: They heard the arguments, and they made a decision. [00:24:50] Speaker 00: And this court is rightfully respectful of that process. [00:24:54] Speaker 00: And it's not one, not two, but three reasons to uphold that verdict. [00:24:59] Speaker 00: There is the validity issue, which I think is been over. [00:25:06] Speaker 00: And I'm here for you, Your Honor, so I'm happy to answer further questions. [00:25:10] Speaker 05: Do you want to go over there? [00:25:13] Speaker 05: I'm stumped as to what to do on this. [00:25:16] Speaker 05: Because the validity of the thing seems to me like, even though you say it's not a new claim construction, it seems to me that at least it's implicitly a new claim construction by saying the patent requires a complex number randomizer. [00:25:30] Speaker 05: So in that, you know, I assume you disagree with your friend that that was presented to the jury. [00:25:38] Speaker 05: Yeah, no, no. [00:25:41] Speaker 05: That's fine. [00:25:41] Speaker 05: Let's assume the jury did not know about this requirement. [00:25:48] Speaker 05: At the very least, it should come back. [00:25:50] Speaker 05: I assume you want us to reinstate the invalidity thing, but I'm having a little trouble getting around the fact that the district court found that the patent actually did include this. [00:26:00] Speaker 00: I'm happy to speak to that again, Your Honor. [00:26:03] Speaker 00: This court obviously sees this issue time and time again. [00:26:06] Speaker 00: A party goes to trial on one set of claim instructions, doesn't get the answer. [00:26:10] Speaker 00: And all of a sudden, oh, I've got a new claim for instruction. [00:26:12] Speaker 00: Oh, it's not new, but it's come up before. [00:26:15] Speaker 00: There are loads of cases on the topic. [00:26:17] Speaker 00: So you're not inventing them. [00:26:19] Speaker 00: It's not the first time it's caught. [00:26:20] Speaker 00: It's not a novel problem. [00:26:21] Speaker 00: And we cited them as HB, the Solvay, there's Moba, there's Quartus. [00:26:25] Speaker 00: And they really all come down to whether the claim instruction was presented and whether it makes sense. [00:26:31] Speaker 05: Are there any that meet these facts? [00:26:32] Speaker 05: I can understand where a claim construction was presented to the jury. [00:26:36] Speaker 05: They argue and the jury finds them validity. [00:26:39] Speaker 05: On a JMAW motion, they argue a new claim construction. [00:26:41] Speaker 05: The district court says, no, that's a new claim construction, and we affirm. [00:26:45] Speaker 05: But that makes sense. [00:26:47] Speaker 05: Are there any situations where the district court says that this new claim construction actually should apply, and therefore, I'm going to flip the invalidity decision? [00:26:57] Speaker 05: And we go back and say, no, you can't do that, rather than just vacate and remand. [00:27:03] Speaker 00: Yeah, I think the HP and I think Solvay are in that procedural posture, but the logic would be the same regardless. [00:27:10] Speaker 00: It would be an unfortunate outcome if that tactic would work here. [00:27:16] Speaker 05: Well, I understand that, but isn't the district court free if even after a jury trial of invalidity, somebody come in and says, you know, you really got your claim construction wrong. [00:27:26] Speaker 05: You should change it. [00:27:27] Speaker 05: He's not bound. [00:27:28] Speaker 05: He could say, I agree with you. [00:27:30] Speaker 05: If he does, what he should do is send it back to the jury. [00:27:32] Speaker 03: But the jury's already ruled. [00:27:34] Speaker 05: Under what he now thinks is an incorrect legal claim construction. [00:27:37] Speaker 00: Well, but litigation has a sequence. [00:27:41] Speaker 00: It has a sequence. [00:27:42] Speaker 00: And the reason why Markman comes in the process when it does is it doesn't come after the jury ruled. [00:27:48] Speaker 00: post trial, post jury markman. [00:27:51] Speaker 00: That is just not the system we have and it would be unfortunate. [00:27:54] Speaker 00: This is a heavily litigated patent. [00:27:57] Speaker 00: It really is. [00:27:58] Speaker 03: We've upheld a district court where the district court goes back and clarifies the construction. [00:28:04] Speaker 00: Sure, sure, that happens. [00:28:05] Speaker 03: That's a clarification. [00:28:07] Speaker 03: Exactly. [00:28:07] Speaker 03: And in our cases we say that's okay because the clarification itself was minor to the point that the jury understood both sides and understood [00:28:18] Speaker 03: what's being sought now. [00:28:20] Speaker 03: Is this a clarification in this case? [00:28:22] Speaker 03: Isn't this a real game changer to suddenly come back and add the complex randomizer into the construction? [00:28:31] Speaker 00: And that's where this argument began. [00:28:33] Speaker 00: A game changer, it's a U-turn or whatever metaphors we want to use, Your Honor. [00:28:38] Speaker 00: At A73, there's an agreement by the party that the complex [00:28:43] Speaker 00: that the definition of figure one does not include complex multiplier. [00:28:49] Speaker 01: Was there a basis for the jury to not accept what is characterized as the agreement of the two experts? [00:28:59] Speaker 01: Was there a factual basis? [00:29:00] Speaker 01: Could they have disbelieved them? [00:29:03] Speaker 00: Thank you, Your Honor. [00:29:04] Speaker 00: Yes. [00:29:05] Speaker 00: Our expert certainly said things that the jury could have credited. [00:29:08] Speaker 00: So even if our expert had said what it [00:29:11] Speaker 00: been mischaracterized as adding the complex multipliers. [00:29:13] Speaker 00: For example, at 1059, he says, the court's claim construction told us what the first computing means is, and it didn't say complex multipliers. [00:29:22] Speaker 00: The jury could have just said, yes, I believe that part of the testimony. [00:29:25] Speaker 00: Depending 1074, the prior art, which all agrees has no complex multiplier, 1074, our expert says the prior art still invalidates. [00:29:33] Speaker 00: So there's plenty of testimony for the jury to resolve, even if our expert was, in a sense, contradicting himself, plenty of expert [00:29:40] Speaker 00: testimony for the jury to, and that's what, I mean, that's really what you should be deferring to. [00:29:46] Speaker 03: What role does the construction in the Acer case have in this case? [00:29:51] Speaker 00: I mean, doctrinally, I don't think it has anything to do with it. [00:29:55] Speaker 00: It's a separate case. [00:29:57] Speaker 00: But I think in terms of this court understanding the real world dynamics of a party that advances arguments in one setting and elects not to argue them in another setting, I think it's useful background. [00:30:08] Speaker 00: And obviously the judge there made some decisions that were well thought through and they're instructive, but they're not mine. [00:30:19] Speaker 00: One more time. [00:30:26] Speaker 00: So I'll close with, I have a few more minutes and I do think the tech is worth trying to understand since we all spend a lot of time on it and it took me a little while to really understand the way it works. [00:30:37] Speaker 00: which is this, that each data symbol that is transmitted has to get its own unique code. [00:30:45] Speaker 00: If you get that, then a lot of things follow from that. [00:30:47] Speaker 00: And the reason why each data symbol has to get a unique randomizing code is because they're being transmitted at the same time. [00:30:53] Speaker 00: So to decode them when they're all sent at the same time, they each have to have their own unique code or you can't unpack it. [00:31:00] Speaker 00: Once you know that, once you know that each data symbol that's transmitted at the same time has to get its unique code, [00:31:06] Speaker 00: then they have to be on separate paths. [00:31:09] Speaker 00: And once they're on separate paths, and N is the same number of data symbols as paths, then you know there's only one symbol at a time. [00:31:15] Speaker 00: And that's what figure one shows in more detail. [00:31:18] Speaker 00: And that's the N argument. [00:31:21] Speaker 00: But that's a lot. [00:31:23] Speaker 00: But the first issue, there's a lot of technical terms, but it actually comes down to a very simple question, which is this. [00:31:30] Speaker 00: The jury had the claim construction of N. N equals the number of parallel data symbols. [00:31:36] Speaker 00: Then you go to the example 2, which is at page 43 of our red brief. [00:31:41] Speaker 00: And it's just a picture. [00:31:42] Speaker 00: And it's a picture of three yellow dots. [00:31:45] Speaker 00: And all we really have to say is, are those three yellow dots in parallel or not? [00:31:51] Speaker 00: And they're not. [00:31:52] Speaker 00: If you look at the three dots, they're just there. [00:31:56] Speaker 00: The two are serial. [00:31:57] Speaker 00: And page 43. [00:31:58] Speaker 00: I got it right. [00:32:01] Speaker 00: So it's page 50 of the blue brief or page 43 of the red brief. [00:32:06] Speaker 00: The only question the jury was asked is whether those two yellow dots and the one dot below are in parallel. [00:32:12] Speaker 00: The top two are serial and the bottom one is kind of in the middle. [00:32:16] Speaker 00: The streams may be in parallel, but the data symbols, which the jury was told was the constructions, the data symbols have to be in parallel. [00:32:22] Speaker 00: The jury just looked at that and said, well, you know, I don't think so. [00:32:25] Speaker 00: And this court doesn't even have to agree with the jury. [00:32:27] Speaker 00: They just have to say that was a reasonable determination. [00:32:33] Speaker 00: And if you actually start to unpack the patent, if you go to figure one, you'll see that there are separate symbols. [00:32:41] Speaker 00: If you go to the background section, more than one code at the same time, and the whole point, and this is a column two, line 16, is you're gonna get these low complexity parallel operations. [00:32:49] Speaker 00: So that's the heart of the technical explanation, but the jury result that this court's reviewing is simply whether N equals parallel data symbols can match up with the example two, [00:33:00] Speaker 00: 1027 is what our experts told the jury. [00:33:02] Speaker 00: Example two does not comply with the court's claims instruction. [00:33:13] Speaker 00: I'm here for you, Your Honor. [00:33:14] Speaker 00: Any other questions? [00:33:16] Speaker 03: No. [00:33:16] Speaker 03: Any questions? [00:33:17] Speaker 03: Thank you. [00:33:37] Speaker 04: Turning to the converter construction, the court, and you can see this on pages 16 and 17 of our opening brief, we show the constructions proposed by Apple for the term converter. [00:33:50] Speaker 04: On page 16, we show the constructions proposed by Apple for the function of the converter, for converting the first stream of data symbols into plural sets of N data symbols each. [00:34:05] Speaker 04: So pages 16 and 17. [00:34:08] Speaker 04: In both the converter construction on page 16 and both the converting function on page 17, you'll see that Apple proposed what is equivalent to an even distribution of the symbols on the output of the converter. [00:34:27] Speaker 04: If you look at the claim language, the claim language only requires that a converter convert the incoming stream of symbols into plural sets [00:34:36] Speaker 04: of n data symbols each. [00:34:38] Speaker 04: The court construed that as even separation of the incoming stream into groups. [00:34:43] Speaker 04: Three at a time, four at a time, that's n data symbols each. [00:34:46] Speaker 04: Each group has to hit the same number. [00:34:48] Speaker 04: So even separation of the incoming stream and the language of the claim does not speak to how the data symbols need to be output because it's not important to the invention. [00:34:59] Speaker 04: The invention is about having two or more parallel streams of data symbols [00:35:05] Speaker 04: And how many symbols from each group you distribute on those parallel paths is not important to the invention. [00:35:11] Speaker 04: The invention's turning a single-lane highway, because you're going to mine all these lanes together and transmit them as one composite signal, into a multi-lane freeway because you'll be able to get out those parallel paths on the other end. [00:35:24] Speaker 04: How many cars or how much traffic, how many symbols you put on any highway is not important to the invention. [00:35:30] Speaker 04: And so they imposed, [00:35:34] Speaker 04: during trial, in addition to the court's constructions, that was rejected during markment. [00:35:40] Speaker 04: The court agreed that even separation from the incoming stream was required. [00:35:46] Speaker 04: You can see that on page 17 of the opening brief. [00:35:49] Speaker 04: But he rejected and re-bolded the language that would then say, I need to then separate each group. [00:35:55] Speaker 04: Once I've evenly separated it from the incoming stream, I've got my group of three symbols. [00:36:00] Speaker 04: They then tried to add the limitation that you need to then evenly distribute those symbols on the output paths. [00:36:06] Speaker 04: You will find no language in the claim that calls for anything about how you output those symbols. [00:36:11] Speaker 04: It's not important to the invention. [00:36:13] Speaker 04: The invention is about having two or more parallel orthogonal paths, a multi-lane highway, and you can place [00:36:19] Speaker 04: Any number of symbols from the group on the top path, any number on the bottom path, you have three lanes. [00:36:25] Speaker 04: You can distribute them in any way you would like across the three lanes. [00:36:28] Speaker 04: It's, again, creating a multi-lane highway. [00:36:31] Speaker 04: That's the essence of the invention. [00:36:33] Speaker 04: And then the complex randomization to get rid of the spikes in the signal that you get when you go to combine these paths for transmission. [00:36:41] Speaker 04: And that allows for it to work in a cellular device. [00:36:45] Speaker 04: By getting rid of the spikes, I can use a low-cost linear amplifier. [00:36:48] Speaker 04: Because now I don't have to replicate this big signal, which would take a big amplifier. [00:36:52] Speaker 04: And so that's the essence of the converter. [00:36:55] Speaker 04: Back to just a couple of points that I would like to raise. [00:37:01] Speaker 04: And that is you'll find in Appendix Site A73 that the court included the corresponding structure for the first computing means. [00:37:16] Speaker 04: And in there, you will find the cited passages I took you through that both experts agreed disclosed that the corresponding structure included the complex multiplier. [00:37:26] Speaker 04: So you will find that actually set forth in the court's construction on Appendix A73, which is page 56 of the Jamal order. [00:37:35] Speaker 04: Back to the issue of the sequence. [00:37:39] Speaker 04: The question of whether or not an order is necessary to performing the functions of [00:37:45] Speaker 04: combining the modulated data symbols and doing the functions of the first computing means, spreading to modulate, create modulated data symbols, followed by the function of randomizing those modulated data symbols. [00:37:59] Speaker 04: That is a question that goes to the literal scope of the claim. [00:38:03] Speaker 04: How much of the corresponding structure do I read into the claim's literal scope? [00:38:08] Speaker 04: So order is not necessary. [00:38:11] Speaker 04: The experts agreed. [00:38:12] Speaker 04: It's in the record. [00:38:14] Speaker 04: The prior art, the Qualcomm prior art, the guild housing reference, says that how you order the combining and randomizing structures is not, it gives you the exact same output and it's interchangeable. [00:38:27] Speaker 04: And then the Qualcomm data sheets for how their chips work say you can reverse the order. [00:38:33] Speaker 04: You can do the order either combined in between modulating and randomizing, or you can combine at the end and you get the exact same result. [00:38:42] Speaker 03: But it affects the hardware. [00:38:44] Speaker 04: Right, but that's an equivalence question. [00:38:47] Speaker 04: Order under means plus function, we look at is it, I have to decide what's the corresponding structure. [00:38:53] Speaker 04: And do I have to read in an order to these structures or not is a question of is it necessary? [00:38:59] Speaker 04: Your law says that you only read in so much of the structure disclosed in the patent as is actually necessary for performing the functions and all experts agree you get the same output. [00:39:11] Speaker 04: The first computing means will modulate [00:39:15] Speaker 04: If you combine, put the combiner within the first computing means, it will combine, it will then randomize, and the output will be modulated symbols that now have been randomized. [00:39:25] Speaker 04: You will get that regardless of where you put the combiner. [00:39:29] Speaker 04: The first computing means, if I put the combiner last, I'll modulate in the first computing means, I'll randomize, and then I'll combine. [00:39:37] Speaker 04: Either order gets me the exact same functional result required by the claim. [00:39:42] Speaker 04: And your law says you only read in [00:39:44] Speaker 04: that of the corresponding structure necessary to perform the function. [00:39:49] Speaker 04: When you move over to equivalence, we never had an opportunity to argue equivalence where we knew what the literal scope of the claim was because the court didn't enforce its markman ruling, which said modulated data symbols didn't have to be randomized. [00:40:03] Speaker 04: And so when you get into DOE, that's when you're now into the insubstantial differences test. [00:40:09] Speaker 04: And when you get to that, that's where differences in hardware matter. [00:40:12] Speaker 04: But here, the overwhelming evidence was. [00:40:14] Speaker 04: I'm not sure you're out of time. [00:40:15] Speaker 04: Yeah, sure. [00:40:16] Speaker 04: I think we have your argument. [00:40:17] Speaker 04: Thank you very much. [00:40:18] Speaker 04: You're welcome.