# Oral argument — Hangartner v. Intel Corporation > Briefs, arguments, decisions, and more. URL: https://www.frixlaw.com/law-library/documents/argument%3A16193 ## Record - **Collection:** Oral argument - **Document type:** Oral argument - **Published:** January 6, 2016 ## Text 015-1293 Angartner v. Intel Please proceed, Mr. Mayor. Thank you, Your Honor. Good morning, Your Honor, and may it please the Court. Again, my name is Philip Mann. I'm here on behalf of the appellant, Rick Angartner. In the proceedings below, Your Honor, the fundamental error that the district court made was misinterpreting the word element, or elements, as used in Claim 1 of the 422 patent. Can I move past what you want to talk about? Certainly, Your Honor. I'm here to help the Court. Frankly, the only thing that is centrally on my mind. For purposes of this question, assume that I think you're right about the language of the claim taken as language of claim. Nevertheless, you added this element to overcome what amounts to an anticipation rejection, that is statutory double patenting. And under your view, the claim would include systems in which the synchronizing element doesn't synchronize two different nondeterministic logic elements, which would seem to be captured by the 518. How does that not amount to, let's call it a disclaimer, of single element systems that would be covered under your view? I'm glad you asked that question, Your Honor, because I agree that's a central question. If I may rephrase the question, it seems sort of common sense. How can you synchronize only one? If you have one element, how do you talk about synchronizing when there's only one? Doesn't synchronizing imply two? Is that, in essence, what's on your mind, Your Honor? Well, that's, I think, related to it. I mean, you can synchronize components, but let's assume for purposes of this question that I think that what is required to be synchronized in the last element of the claim is actually two different logic elements rather than the components that spit out a single result of a logic element. So assume that in order for the synchronizing to be doing anything, there has to be at least two logic elements. Well, I would respectfully disagree with that characterization, Your Honor, for this reason. I'm interested in that dispute about the premise, but I guess I'm primarily interested in let's assume, just for purposes of this question, that the synchronizing element here, element of the claim, has to apply. In order for there to be synchronizing, there has to be two logic elements. At that point, how is it that you didn't disclaim a single logic element system? Well, that's begging the question, Your Honor. It's basically saying if the claim requires two elements, if we're going to assume that the claim requires two elements, how can it be satisfied with only one element? And again, if the hypothetical is that the claim requires two, I can't answer that. That's sort of like saying if we assume that two plus two equals five, how can it equal four? If it doesn't require two in order for, how does the claim not include systems covered by, I guess, what's it, the 518 patent, which was distinguished? Oh, the 518. Well, the difference is, I think what was happening here, that's not, it was a double patenting rejection. There was that. And they simply added the synchronizing means. Now, what the inventor has said, this is not in the record, but they did not intend to include synchronizing two separate elements. All they were doing was adding the synchronization means, which in our view, what that means is you've got these two cross-coupled transistor inverters, and in order to generate a random number, you have to make sure that those inverters are both released very, very close in time. If you favor one over the other, you're no longer… But that's true every time you want to generate a random number, and the 518 thereby captured the same thing because it claimed generating random numbers. So every single logic element in the 518 had to have two transistors firing at the same time, or you wouldn't generate the random number. I don't believe that was in the claim. It certainly wasn't in the broadest claim of the 518 patent, and that may have been disclosed because these were related disclosures. But the claims of the 518 patent, with all due respect, Your Honor, did not require that. That's why… No, no, they didn't require the synchronization element, but they did require a random number generation, correct? I believe that they did. And they required a logic element to perform that random number generation, correct? Well, they were… I'm not sure, but continue, please. Well, I guess because the argument goes, if the only way the logic element can generate is by synchronizing two transistors to operate to cause the generation, then they do have the same scope. Well, then that's a different question. We've got to remember here, we're jumping ahead here. The sole question in the proceedings below is what do these claims mean? Specifically, what does claim one of the 422 patent mean? Right, but we're supposed to look at the whole history, we're supposed to look at the prosecution history and the specification. We can't just say, I mean, you have some pretty good arguments from a grammatical standpoint, but that can't be the end of the inquiry. Well, again, Your Honor, at the risk of… I don't mean to be flippant, but we're getting a little bit ahead. Ahead of what? These are the questions that we want answers to. No, what I'm saying is the proceedings below did not get into questions of what the patents or whether the specification supported the claims. The validity, in essence. These are questions of validity. These are questions of claim construction. We've got to figure out, I mean, whether or not you can have a single logic element in this claim. And you've got somewhat contradictory language in this claim, some of which was added during prosecution. So we have to be sure that the language added during prosecution has meaning. We can't write it out of the claim because it is what allowed the patentability of your claim over the 101 rejection. And so we have to, as part of the construction, consider that language. And so that's exactly what we're doing. Whether the district court did a thorough job of considering it and explaining it is irrelevant. That's what we have to do in this case. I understand, Your Honor. And, again, the only way I can answer that question is to say our position is the claim has meaning if you view synchronization means as meaning synchronizing the inverters of the logic element, a single logic element. I don't believe that that was claimed in the 518 patent. There's no claim in the 518 patent that goes directly to that. So even if you construe the claim as I propose to construe it, obviously I'm biased, but our proposed construction of the claim, meaning the synchronization, refers to synchronization of the inverter circuits within a single logic element. But nothing says that, right? Is there anything that you can point me to in the prosecution history that ties the synchronization to the inverter elements as opposed to tying the synchronization to a plural number of logic elements? Is there anything? Give me your best statement in the prosecution history. The best statement I can say to that is for that to apply, there has to be a clear… In the prosecution history referring specifically to that amendment, there is nothing in that that discusses that because the discussion of that element was… There was no discussion, period, on it. There was very, very little discussion. So what I'm saying is I can't point to something for the simple matter that I don't believe there's anything in there. But the language of the claim itself says synchronization for synchronizing operation of the nondeterministic logic elements. And you're saying that, and it talks about the common synchronization means coupled to all of the nondeterministic logic elements. And you're saying that you're not synchronizing those elements? You're synchronizing the other aspects of each one? No, what we're saying, the consistent interpretation we have is this. If you have a single, and I think there's no dispute that the prior part of the patent claim contemplates having a single nondeterministic logic element. I don't believe Intel even disputes that. Our focus is on the last paragraph. Our position, very simply, is if you have one, and the claim in the preamble, or the first part of the claim says you can have one, if you have one, you're synchronizing the inverters in that logic… Yeah, but the problem is that you added the synchronization language to overcome the prior act. But the synchronization language does not require two or more. Well, but in every reference that I can find throughout the history of this claim, you talk about synchronization in terms of two or more logic elements. I can look at the appellant's opening brief before the board at page 220, where you explain the synchronization element. The circuit includes essentially two elements, one nondeterministic logic element for each variable to generate a random Boolean value, and B, a common synchronization mean. That synchronizes operation of the nondeterministic logic elements, and then you explain. They have to be synchronized because each guess at a solution to the problem requires a random value be picked for every variable in the problem. And all of that language, I understand them to argue, suggests multiple elements have to operate simultaneously so you get different values for each variable at the same time. Well, that is not inconsistent with a single – if you have a single variable and a single nondeterministic element, that language is equally applicable. The attorney saying that is thinking the general case of one to many. So instead of saying in the case of one, I'm going to talk about one, and now we're going to talk about two or more, it's the way the language works. When you're referring to one or more, it is common to thereafter refer to that as plural form. One or more people, the people in the car. One or more ice cubes, you know, ice cubes. When we pointed this out in our brief, this is how language – But your only articulation in the specification of one to many is iterative. It isn't simultaneous. Your only embodiment described in this spec of one to many, a single logic element that produces multiple variables, multiple random numbers for multiple variables, does it iteratively. Does it do it synchronized in a simultaneous fashion? So all the other embodiments are one to one, right? I mean, you're looking at me like I'm crazy. Am I misunderstanding the spec? I believe you're misstating what the spec says. I think the specification in the inventory has always talked in terms of one or more variables. And they have to talk about the general case of when there's more than one, but they do point out – Then why did you need to add the word synchronization? If all it means is that each one of the logic elements is functioning properly, then why use the word synchronization and why even add that language? Overcome the double patenting rejection. That was not to overcome any form of prior art. So are you saying in your prior art – All right, I refer to it as prior art, but in the earlier patent, are you saying the logic elements didn't have to operate properly? No, we're not saying that at all. What we're saying is in the absence of the synchronization means the elements that were previously stated in claim one were, in the examiner's view, the same as what was in claim in the 518 patent. What they were doing simply is adding additional structure in order to overcome a double patenting rejection. Pure and simple. So you're saying in the first patent, those logic elements did not have to operate properly because they weren't synchronized, as you say. No, I'm not saying they were. It's a different rejection here. The examiner was saying that based on what you've got in claim one of the 422 patent as initially drafted, that is too close to what you're claiming in the 518 patent with something else in there. They're not saying it doesn't work. They're not saying this is what you have to put in to make it work. They're saying give us something so the claims are not identical. Well, you can't just give them a word. There has to be something that actually changes the way the claims operate. No, the way the claims are specified. What they did, they did add. They added several words about the synchronization means. What they did not say is we're going to go back and rewrite the first part of the claim to where it says one or more, we really mean two or more. That did not happen. Intel is trying to argue that that's what happened by the addition of the synchronization language, but that is not what happened. What happened was they simply added synchronization means, that in its purest form talks about synchronizing the inverters. This happens in all. In this record, we don't have the examiner's actual rejection on the 101 grounds. What were his comments consistent with? I can't answer that, Your Honor. I don't have that at my fingertips. Is it in the record, not in our appendix, but is it part of the record of this case? I assume it was. I did not prosecute this patent application. I don't know. My assumption is that it is, but I can't tell you certainly that it is. Just in the district court, was either the entire file wrapper put into the record in the district court or even the double patenting rejection to which the response was a response? Again, my assumption is that it was. I'd be surprised if it wasn't, but again, I can't tell the court absolutely that it was. It would be my assumption, but I don't know for a fact. Okay. All right. Well, we better hear from Mr. Berman, and we'll restore a few minutes of rebuttal time for you, Mr. Mann. Okay. Thank you, Your Honor. Mr. Berman. May it please the Court, David Berman for Intel. Intel has tried to simplify its response on appeal in terms of grounds for affirming the judgment below, and we have focused on the fact that Dr. Hangardner, as the Court's questions have suggested, chose to add the common synchronization limitation to the rest of what is now Claim 1. Page 11 of our brief sets out Claim 1, and you'll see that the bulk of it is identical to a claim in the parent patent, in the 518 patent. By the way, is the 518 – were we given the 518 patent? A lot of your material on pages 11 or 12 and 13 come without citation to anything in the joint appendix. I mean, I'm kind of – I don't believe that we put the 518 patent in the joint appendix. Is the entire prosecution – I know that the prosecution history, the actual rejection is not in the appendix. Is it in the record in the district court? It is not in the record in the district court. Portions of the prosecution history are in there, including his response to the rejection, from which you can infer – Suppose I thought that the single most important question to answer is, if the synchronization limit meant to require two logic elements in order to overcome the double patenting rejection, I need to be persuaded that that was actually necessary to overcome the double patenting rejection. How am I supposed to figure that out without even the material in the record? I think you need to know that something was necessary, something additional that wasn't already in the 518 claim or the rest of Claim 1 and the 422 patent. And what that must be is what's described in the common synchronization limitation, which is that multiple elements – it uses elements in the plural twice – are coupled in common synchronization. Suppose I thought that just looking at the language, it is completely ordinary to put in a requirement that says, synchronize any of the logic elements, all the logic elements that you have. And sometimes you have one and so there's nothing to synchronize. That's a perfectly fine claim. So the only way that I can get, suppose for these purposes, that the only way that I can get to an inference that there must be two is that there had to be two for this element to overcome the double patenting rejection. How do I figure that out? Because common synchronization must add something to the prior part of Claim 1. And for that to be true, there must be more than one element. Even if you toggle one of those elements off, as the specification does suggest you can do, I'm sorry, it does add something to the pre-amended claim elements. It says when you have more than one, synchronize them. Which means it must be capable of dealing with more than one variable with more than one element. It has to be capable of that for the common synchronization limitation to add anything. As he's admitted, the 518 had to operate. It had to work. So synchronization must do something more than make one element work. And as the record in Joint Appendix 220, as Judge Moore pointed out, indicates, he explained that the way it does that is to synchronize plural elements. So for it to add something, it has to do it in that way. The plain language of that limitation, which has to have independent meaning, is what we base our case on. And that is not as far as the district court went, however. The district court pointed out that that's consistent with the intrinsic evidence. When you say the plain language of that, you don't really mean that, right? You said we base our case on the plain language of that claim element. You mean, don't you, that we base our case on the disclaimer that occurred when that claim element was added for patentability purposes. Isn't that what you base your case on? Because if you base your case on the plain language of the claim in the absence of the prosecution history, I promise you, you lose. There might be ambiguity. So what do you base your case on? There might be ambiguity absent the disclaimer. The disclaimer clarifies the ambiguity completely. But we would say that the language of the common synchronization limitation is plain. It refers to elements twice in the plural. That element might be plain, but the language of that element in the context of this claim is not plain. But that limitation has to have some meaning. It is the limitation that we are construing in what's most relevant here. And the plain language, if you look just at that limitation, is plural. It's common synchronization coupled to all of those. So if you say in the claim, you have to synchronize all the logic elements you have, why does that imply that you have to have at least two logic elements? It seems to me it doesn't. It states that you have to have plural elements. It doesn't, though. It says you have to synchronize whatever number of logic elements you have. And if you have one logic element, there's nothing to synchronize. With all due respect, Your Honor, that's not the plain language. That's where the appellant gets by going back to the first part of Claim 1 and saying you have to read in the antecedent one or more. But the antecedent one or more applies only to variables. And there is no need to go back to the prior part of Claim 1 to determine numerosity. But does it apply only to variables? Because it says that you have one logic element for each one or more variable. That's the next step of the appellant's argument. So if you've got more than one variable, you necessarily have more than one logic element. But if you only have one, you have one. And the invention has to be capable of solving multivariable problems. That is the problem that was identified. That is certainly relevant, intrinsic evidence to how to read the common synchronization limitation. And the specification, as has been pointed out, never describes a single logic element. What appellant belatedly points to is that in terms of the evaluation circuit, which page 6 of his brief admits is separate from the generation circuit, the generation of the random variables or random numbers to match up with the variables, the testing or evaluation circuit can be run serially. But there is nothing that suggests that the generation is anything but parallel generation using multiple logic elements toggling one or more. Can you try to explain this to me? Suppose that I thought that the last element, just taking as its language, read with the rest, simply means on its face, synchronize any logic elements, non-deterministic logic elements that you have. Tell me why, if that were read to cover a circuit that had just one non-deterministic logic element, it would not have overcome the double patenting rejection. It wouldn't have added anything as to the situation where there is one variable, one element, because it would still have to operate in that situation. So it would not be a proper construction consistent with that being added in to say in that situation the patent should be construed to apply. Is that because, this is what I'm trying to get at, when you look at the 518 claim 1 as set out in your brief, that whatever the proper statutory double patenting analysis is, that that would still be covered by the interpretation of the amended claim? Exactly, by appellant's interpretation. Even though synchronization of no sort is discussed in the 518? If synchronization only applies when you have multiple elements, or if all it means is that it causes elements to operate, then it adds nothing. It adds nothing because in the 518 it said that the elements had to be able to generate a random Boolean value. Correct. So they had to operate. Correct. I find your argument at the podium today to be surprising because you're basically arguing the district court got at least a portion of its construction wrong. Right? You're arguing the district court got a portion of this construction wrong. There's no other way for me to interpret your argument. We believe the district court should have accepted all our arguments. Wait, no. No. You agree, because you are arguing to Judge Toronto and Judge O'Malley a one-to-one correspondence and the district court rejected that and said it's a one-to-many. Actually, I think it's the appellant who's rejecting that. You just stood here and told Judge O'Malley there's a one-to-one correspondence between logic elements and variables, and the district court expressly rejected that. I'm trying to take the appellant's argument, which is that if you... What is your argument? Is your argument, do you agree with the district court that found one logic element can actually generate one or more variables in series, and that that is the way to interpret the first element of Claim 1? We disagree with the conclusion that that's the way to interpret the first part. So you are standing here telling us the district court's construction is wrong. On that part, we believe the district court's judgment should be affirmed by rejecting that part of the district court's logic. We believe that in the common synchronization limitation that it is... Both parties agree that it's a means plus function limitation, and it describes at the top of Column 7 in the specification in Figure 4 and in common sense reading of the common synchronization limitation the signal fanning out to more than one element. We believe that that is the proper construction of the patent. And we believe that the district court was correct when it looked beyond the language and the history in terms of the disclaimer type approach of the common synchronization limitation to the remainder of the intrinsic evidence, and concluded that when you look at the problem that the inventor said he was addressing, complex, combinatorial, multivariable issues... But you understand perfectly well that the claims can have different breadth and some of them can address various things. Until you try to understand what the addition of the synchronization was doing, it seems to me actually quite plain on the face of the claim language that it covers single logical element claims. If there had not been a prior history of his own patent and a prior art, that may be true. So I guess I'm... We would say it's at most ambiguous, and in fact if you look at what... And we don't even have in the district court record the full prosecution history, right? That's my recollection, yes. So your argument is D-Delay is basically the structure for performing... I call it the structure, but it's a signal A. It's a signal, and it fans out through the inverters, 80, as you see on figure 4 and described at the top of column 7. We believe all of that structure is necessary, is what's specified by the common synchronization limitation, and that that is admitted not part of... So the structure for producing D-Delay is the structure for the synchronization means? It's actually the structure that causes D-Delay signal to flow through to all of the logic elements, all of them simultaneously, and that's that split if you look at figure 4 that's shown in the inverter 80 box. That structure, that fanning out and those inverters, we believe is a necessary part of that means plus function limitation, and it's admitted that that's not present in the Intel technology. So we also believe that it is important to look at what the inventor was trying to do to understand why it is logical that he would choose to add the common synchronization limitation when he faced the double patenting rejection. That was his way of addressing, as he makes very clear in the specification, of dealing with these complex problems, and the task that he was discussing was addressed by common synchronization coupled to all of plural elements. Up above, in the remainder of Claim 1, it does say each element, but each is not an indefinite article. Each is a determiner, which means every one of two or more. It says one non-deterministic logic element. It's not each, it's the word one. One non-deterministic logic element. It doesn't say A, it doesn't say logic elements. It says one, the number one. It's actually their argument that it says A, and what it says, it starts off with one non-deterministic logic element for generating a random Boolean value for each one, and then it says and each non-deterministic logic element comprising. The only antecedent that is needed to understand the common synchronization limitation is what is the nature, not the numerosity, of the logic elements. Common synchronization limitation determines numerosity. It indicates that they are plural, and that they need to be commonly synchronized by this fanning out of the signal, coupling all of them together. For those reasons, because it is undisputed that Intel does not infringe on that construction, Intel asks you to affirm. Thank you. Thank you, Mr. Berman. Mr. Mann, how much time do we have? We have two minutes for rebuttals. Very brief, Your Honor. Sitting here listening to this, what struck me in this is, I think the court's precedent is clear that if a patent applicant is going to disclaim certain subject matter, it has to be done clearly. It has to be done intentionally. There's got to be a clear indication. Now, if the prosecution history in this case had included something where Dr. Hangartner comes in and says, hey, I am distinguishing the prior art. That only has one logic element. I have two. That is why my patent is different, is because I have two and prior art only has one. We'd be clear. Then what else is different? We don't have that. What we have here – What is different between the 518 and this patent? In the claims, this one has synchronization. In the independent claim, it has the synchronization means. That is what is different. But that goes back to your argument that all you're doing is synchronizing internally to make it operable. Or consistent with even Judge Cronk. Didn't the 518 one logic element have to be operable in order to generate a random Boolean value? I don't know whether it does or does not, but it's not in the claim. So you're not adding anything. It's not in the claim. What we have to say is it's double patenting. It's not a double disclosure. It's double patenting. The claim did not include the synchronization means. So when you add the synchronization means, even if it's only synchronizing a single logic element, that distinguishes the claim. That obviates the double patenting rejection. Let me see if I'm understanding right. It's really quite wrong to think about this double patenting rejection and 518 in a prior art sense, which covered single logic element situations. That's not what statutory double patenting doctrine says. Statutory double patenting doctrine requires that the objection is overcome as long as the claim language substantively is different, even if the earlier one would cover some of the same stuff. Absolutely, Your Honor. That's our interpretation of it. And by adding the synchronization means, that is enough to say the claim is different. Mechanically it's different. We have elements A, B, and C. Now we have elements A, B, C, D. However trivial D is, that may get into other issues, but to overcome the double patenting, same invention double patenting, that does it. Absolutely, Your Honor. That's how we interpret this. Okay, thank you. Thank you, Your Honor. Thank both counsel. The case is taken under submission. --- Source: Frix Law Library, https://www.frixlaw.com/law-library/documents/argument%3A16193. Public record. Not legal advice.