Oral argument — ATI Technologies ULC v. Matal

Okay, and good afternoon, everyone. The argued case this afternoon is Number 16-22-22, ATI Technologies against Matal as acting director. Mr. Farrancroft. May it please the Court. Actual reduction to practice looks at the line between drawings and descriptions on the one hand and actual working systems on the other. And in this case, the inventors created an actual working system created in RTL code that worked to process graphics, and it was a graphics processing system. The RTL code included all of the elements recited in the claims of the patents. Below, LG did not dispute this issue, and accordingly, the Board did not identify any limitation missing from the RTL code in its final written decision. The third feature of Claim 5 of the 053 patent wasn't construed by the PTAB as a memory device? Correct. Is there evidence in the record that a memory device can be construed as software as opposed to hardware? The specification of the 053 patent describes the memory device in functional terms, that it has to be able to store and organize command threads. And indeed, there is no reason that it has to be a specific hardware device. You're not answering my question. Is there anything in the record that is evidence that it can be? You're reading me the description, but to me that sounds like lawyer argument. Did somebody say that? Yes, Dr. Wolf said that the memory device was embodied in the RTL code in his analysis of the RTL. Yes, he did. Where is that? Dr. Wolf's analysis appears. The analysis begins at Appendix 4720. 4720? Yes. Okay. That's good enough. Then point me to any case law or industry evidence that says that. Well, the industry evidence was established again by Dr. Wolf and Mr. Lefebvre, the inventor, who described that the way that these integrated circuits are designed and initially built is in the code evidence, is with the RTL code. And so the case law, there's no case directly on point that a memory device may be embodied in RTL code, but in the industry… Why are you distracting me? Well, Your Honor, in fact, the RTL code was also stored in a computer system, and so it was stored in memory. It was executed in a computer system as well. So when that occurred, there certainly was memory, a physical memory, but the RTL code itself embodied the memory device of these claims as established by Dr. Wolf. But the board made specific findings with respect to Dr. Wolf's testimony, saying that Dr. Wolf testified that the emulator code emulates the behavior of a graphics processing system. And they said, based on his testimony, they believed that what he was saying is that the RTL code is merely a representation of its design. Well, the board did make that finding, and the board pointed first to Dr. Wolf's description of the emulator code, which, as he described in his declaration at 4718 of the appendix, is different than the RTL code. And also, yes, the RTL code is a description of an integrated circuit fabricated in silicon. However, it is within the scope of these claims, and that was where the board went wrong. The test for reduction of practice is whether an embodiment was created that is within the scope of the claims at issue. But the board applied a specific physical limitation, and that's where it went wrong. You argue that the board basically applied the wrong law by saying that you always have to have a physical representation, but didn't really what the board say here is that in this particular case, we read the specification as claiming a physical graphics system. And if you claim a physical system, then you have to have a physical manifestation of it, right? I believe it's unclear from the board's opinion whether it did actually perform a claim construction for the first time in the final written decision that it had not performed earlier, or whether it introduced an additional limitation for the standard for actual reduction of practice of a physical element in addition to an embodiment that works for its intended purpose. The board did not make any claim construction in the proceeding that would disqualify the RTL code from the scope of the claims, and indeed, it did not identify any specific limitation missing. And so however the board reached that conclusion, it was an error of law because it did not apply the proper standard. And when the board pointed to... When the board said that what you're claiming is a hardware system as opposed to... that may have software components, but you're actually claiming a hardware system, you think that that was erroneous? I do think that was erroneous. I don't think the claims are so limited. The board's claim constructions are not so limited. The board did not identify any specific element missing, and the claims encompass the RTL code embodiment created by the inventors here. If you proceeded to a tape-out to produce the chip, at that point, are there other problems that might arise in the production process that would require rewriting the software? No, I don't believe there are, Your Honor. Once the process proceeds to tape-out, it goes to silicon, it goes to a manufacturing process. So at that point, I'm not aware of any additional issues that would require rewriting of the software, which is why on the testing point, Your Honor, when the first triangle test occurred, it did demonstrate that the invention worked for its intended purpose. And again, the board erred on that point also as a matter of law by not reciting, not considering, not applying the invention's intended purpose, which is the appropriate legal standard. Instead, the board again focused on the idea of a physical custom chip as the embodiment required, the error it made in the embodiment aspect of its decision. And so it focused on limited aspects of the test. For example, alleged deficiencies such as the absence of test input or output files, simulation run logs, or input test parameters. But because it did not apply the correct standard or consider the invention's intended use, it did not link any of those things as to the question of whether or not they were relevant to demonstrating that the invention worked for its intended use. Is the triangle test related in any fashion to commercial readiness? It is one step in the process all along a continuous development process, but it is the first time for this invention, as Mr. Lefebvre testified, that the inventors were aware that the unified shader, the ability to process both pixel and vertex operations, worked for its intended use. Is it conceded all around that there's no question here of conception in that date of August 2000 and 2001 that we're talking entirely about reduction of practice? That's correct, Your Honor. Conception was not disputed, and the board found that ATI had established conception prior to all of the prior art references at issue here. So one thought which came to mind, perhaps only in retrospect, because it was clear that your target was an actual reduction of practice, but the idea of constructive reduction to practice, was that separately argued and litigated? Yes, it was, Your Honor. So ATI argued— But it's not in your briefs. You talk about diligence, but only diligence to actual reduction to practice, not diligence to constructive reduction. I believe the briefs— Have I missed it somehow? I believe the briefs do talk about diligence also to constructive reduction to practice. But what about constructive reduction to practice? It's not a section heading, and perhaps in thumbing through it briefly, I've missed the term because it would seem to me, just to go ahead, that all of the issues that have been raised would disappear if we're only talking about constructive reduction to practice. The blue brief at pages 55 and 56, Your Honor. Oh, maybe I didn't get that far. Explains, ATI exercised reasonably continuous diligence from conception through both actual and constructive reduction to practice of the claimed invention. Is that the only word, the only time that constructive reduction appears? Because obviously I missed this one. It appears— It seems as if it was not a focus of the argument. Was it as a matter of confidence in the evidence of actual reduction? So we don't bother with constructive? If actual reduction to practice was found, then the remaining period between the actual reduction and the constructive reduction to practice would no longer be relevant, except for one exception for the 053 patent. Constructive reduction— But if you still provided evidence a fair amount? Yes, Your Honor. I think of diligence all the way to filing. That's correct, Your Honor. The evidence was established that there was continuous work on the R-400 project every business day from conception through constructive reduction to practice as of the filing dates of the priority applications. I'm trying to understand what evidence really does show on this. So one of the things the Board said is you just can't tell. There might be evidence that work was being done, but you can't tell what work was being done. So what evidence did you present? I mean, I read that affidavit from Mr. Lafarve, and I don't see him other than saying we worked on the project. So the Board pointed it might have been for commercialization purposes and not necessarily for continuing to reduce it to practice. Mr. Lafarve testified as to the work was done on the sequencer and shader pipe blocks that correspond to the elements of the claimed invention, as described by Dr. Wolf. He also described that the work that he cataloged in the calendars was work related to being able to produce those two blocks because those blocks could not be produced and worked on in isolation. They were very cryptic sentences, though. In other words, he said we just kept doing this stuff. And I guess my point is I looked at all those exhibits and the references to the metadata. How can you tell from the metadata what exactly was being done? The metadata corresponds to the folder histories and logs, and Mr. Lafarve explained what those were and the work that was being done. And I think his testimony establishes that there was continuous work, reasonably continuous, because, again, we believe the Board applied the incorrect legal standard on diligence. And then the metadata and specifications that he cited corroborate that testimony. Even if we agree with you that under perfect surgical the Board misapplied because they were citing the same cases that were cited in perfect surgical that we said was an incorrect understanding of the law. But even if that's true, even if you don't have to show diligence on every single day, you do have to show that what you were doing related to the development of what ultimately matured into the claims, right? That's right, but I don't believe the Board considered that analysis because it looked for unexplained lapses instead of looking at whether what Mr. Lafarve said was... But it did more than that, didn't it? Didn't it say that there's no way to tell what was being done? Yeah, I believe it did say that, but I don't think that finding is supported by substantial evidence in view of Mr. Lafarve's testimony and the metadata. Well, that's a good portion of his testimony because, like I said, I didn't find it to be as persuasive as I think in perfect case, or as thorough anyway. His testimony is from pages 2648 through 2654 in his declaration in the appendix, Your Honor. I believe I've used my initial time, and I'd like to reserve some or full rebuttal if I may. Yeah, well, that's answered your question. Sure. So what's the... I know where it is, but what do you think is the place where he says this is what we were working on? Paragraphs 41 and 42 on appendix 2654 where he says, My analysis shows that at least one person on the team worked on the design every non-holiday business day. Paragraph 42, he describes, This metadata shows work that was necessary for implementing the R400 design. He describes the metadata for the sequencer and graphics blocks in that paragraph, and then describes that the design and development of the R400 more generally was necessary to continue to work on the sequencer and shader pipe block, Your Honor. So... But you don't have any documents to support that conclusion. The supporting documents were the metadata itself, Your Honor. You reference exhibits. Within this, you reference exhibits. Yes, that's right, Your Honor. Let me ask you this. This is sort of my idle curiosity. I don't know. If you were demonstrating the RTL model to a potential customer, could you just run that as if you were running a chip? Yes. Thank you. Okay, all right. We'll save you for rebuttal, and let's hear from you. Thank you. Thank you. Mr. Krause? It may please the Court. Mr. Krause, how can you distinguish the claimed elements in this case from cases like Fantasy Football, where we held that software claims in a patent could infringe a claim for computer-playing football? Certainly software running on a general-purpose computer constitutes an apparatus, and therefore can be an embodiment of a claimed invention. In this case, that's not the way this case was presented to the Board at all. Suppose the chip exists in a cloud computer. Is it ever embodied? It may... One might use the word embodied to describe it, but I would not say that it's covered by the claims at issue here. What I want to emphasize is the claim interpretation that's being advanced now most prominently in the reply brief is an entirely new claim construction that was not presented to the Board. As the Board says on pages 8, appendix 26 and 27, also appendix 30, they never made the argument that the computer software code running on a general-purpose computer satisfies the claim limitation. The idea was always that the RTL code is some kind of a representation that shows that the chip would work for its intended practice. They say they're not in the chip-making industry, they're in the chip-design industry, so they thought that all they needed to do was present this code that would show that this unified shader component of the invention would work. So they made the argument that at least when it was being tested, it was being run on a general-purpose computer? They've made that argument now. I see it starting in the blue brief and then... But the Board addresses the argument, so how could they have not made it to the Board? No, the Board addresses the primary argument that they made, which was that there's no physical embodiment requirement. They said, they used the words, this invention is, quote, embodied in the code itself. They weren't making the argument that a general-purpose computer running this software would constitute an embodiment of the invention. And that's actually not a very persuasive argument either. If you look at the specifications in all of these cases... I don't know. Well, if you look at all the specifications, they are all about chip design. The whole purpose of a unified shader is to speed up the processing and save real estate on a physical chip. The idea that a software embodiment running this kind of unified shader embodiment on a general-purpose CPU as well as a general-purpose graphics processor doesn't make any sense. It doesn't speed up... One of the arguments you make is that on page 51 of your brief, that Mr. Lefebvre was working on a second chip design to include an optional feature that one of the customers wanted so that he can't demonstrate reasonable diligence. What shows that he wasn't working as well at the same time on the primary design? As Judge O'Malley pointed out, there's very scanty evidence on diligence to begin with. There are a lot of logs that indicate very cursory indications that work was being done. And then we have the inventor, Mr. Lefebvre, trying to tie it all together. Sort of like a testimony about business records. Right. Well, he's sort of trying to corroborate the business records. There's nothing corroborating him. The requirement is that the inventor's testimony needs to corroborate it. He's trying to explain what these very sketchy diligence records show. And the bottom line is... What do you mean by that? It wasn't able to glean anything? Well, I mean, he derived calendars from it and logs from it. And he can come up, and I believe this is probably the case, that at least one engineer on every day during the relevant critical periods did some work on the R400 project. But that's not necessarily the invention here. The R400 project was never completed. That's the chip from Microsoft. We have from their blue brief that they only first reduced a satisfactory chip, or they only first commercialized a chip, in November 2005. That's more than four years after conception. But you're here representing the office, right, and advising us on what the correct law should be. Here we have tens of thousands of patents that have issued without... and have gone through assorted litigation processes and so on, without being challenged that you don't have any reduction to practice that you can rely on because you don't have it embodied in a chip. You just have the drawing of the various steps of either the system or the method. So is this something new for the office, that unless you have your software embodied in a physical chip which you can produce by some critical date, that you haven't established invention? No, not at all. This is standard reduction to practice. But that's what you're saying. I'm not saying that anything has changed. You're saying there's no constructive reduction to practice, and there's no actual reduction to practice, although the weight of the evidence... there's no dispute in terms of conception. And then we have at least evidence that something was done every day. We might get to argue about how much diligence is diligence when your only problem is a swearing-back affidavit. And that's another question that I wanted to raise with you. We're not talking here about interference priority anymore. We're talking about getting behind a reference. And all you have to show, as I always understood it, to get behind a reference is that you possessed what was in the reference before whatever effective date of the reference. So I'm trying to put all of this together with what looks to me like something brand new, actual reduction to practice, a physical chip, a commercial embodiment, nothing further to be done to correct errors, even though from what I read, any routine software that you buy might have a billion errors into it that may or may not make a difference. So to try to understand, this seems to me that there's something very different here in the first place by imposing interference standards rather than the traditional Rule 131 standards. And then the apparent insistence on a physical embodiment of an actual reduction to practice rather than diligence to constructive reduction to practice. So those are a lot of questions all at once. Let's start with the constructive reduction to practice and the position of the office on that. There's no question there was a constructive reduction to practice, the 053 patent application. Well, then why doesn't that end it? It wasn't started before the earliest date. That's not disputed. It was not reasonably continuous throughout the reduction to practice date. The reduction to practice date... If we think it was reasonably continuous, does that end it? I believe it would, but I think the testimony is crystal clear that it was not reasonably continuous. And let me just state to be clear... I think that the law the board applied was wrong because I think that Perfect Surgical explains that the board misunderstood the law, and I know that this board didn't have the benefit of Perfect Surgical at the time. But assuming that we say that requiring daily activity every single day with no gaps is wrong, the question is whether or not that could be harmless because the board separately said that the evidence didn't tell us what any of the activity really was. The board fully acknowledged the testimony that there was continuous activity on every single day, as submitted by ATI. The problem the board had was not the problem in Perfect Surgical where the board had identified certain gaps and said that's not continuously reasonable. The board here found continuous conduct but said that's not diligent reduction to practice. But I'm trying to determine whether one error could have affected another. So in other words, if the board thought it had to be every day, then is what the board's saying because I don't know what you did every single day, then it doesn't matter if you did something throughout the whole period because any one of those days could have been just on commercialization or just on marketing. I don't think the board imposed a requirement that something had to be done every single day. I think it was a given that something was done every single day. Their question had to do more with what was done. You haven't shown me that you have actually been diligently reducing the claimed invention here, this unified shader, to practice. And one way we can see that is that they never did reduce it to practice. Now, again, we're assuming that the July 2002 reduction to practice was not valid. That's how we get to the diligence argument. You have inventor testimony and it's corroborated with something that you say is insufficient. But the problem I have is that the inventor testimony then goes back and says, well, this is what these numbers and letters mean, and you don't have anything that says that the inventor is incorrect in his analysis of that. You just say, well, you're not showing me enough. His analysis doesn't show that work on the unified shader was reasonably continuous from the relevant period, and that the three critical dates at issue here, and just returning to Judge Newman's question, I've always wanted to say the constructive reduction to practice date on the 053 application was September 29, 2003, and constructive reduction of practice for the other ones was a date in November 2003. So we don't deny that patent applications were filed. What we don't have is any indication for the critical periods, and that's before the date of the Lindholm reference is the one that applies to all of them. So we can just look at that from June 25, 2003, to September 29, 2003. There's no evidence that they were diligently working on the unified shader during that part. They made a statement. Wasn't it the opponent's burden either to say, let me see your notebooks, or this or that? At what point, when there is no opposition, then what kind of additional burden after burden after burden? Again, we're not talking about an interference or competing inventions. All they have to do is get behind a reference date with as much information as is in the reference, not with everything else, but to show that they had what the reference had before the publication of the reference. And I didn't see anything to say that that burden was not met. I don't think that's entirely accurate. They had to show from the time period after. These are 102e references. It's not publications. These are 102e references. They had to show that from the period after the filing date of the 102e reference until they got around to filing their patent application. That's the only real reduction of practice they're relying on. They were reasonably continuous. And yes, they had engineers working on this, but there's no clear indication that they were working to bring this embodiment to practice. And let me tell you one reason why I think... Go ahead. You say there's no clear indication. There is a clear indication. That is the testimony of Mr. Lefebvre. The uncorroborated testimony of Mr. Lefebvre about logs which refer to generalized work on a commercial embodiment that ultimately would have embodied this technology but didn't have, didn't actually get embodied. It is corroborated. You say it's uncorroborated, but it's corroborated by this documentation, which you then say is inexplicable. And the answer to that is he says, no, this is what it means. And you say, no, that's not corroborated. Corroboration is like what we had in the Perfect Surgical case. There the inventor had his own testimony. The board thought it was uncorroborated, but the federal circuit, this court, looked at it and looked at the patent attorney's testimony. The patent attorney filed a separate affidavit corroborating what the inventor said. If Mr. Lefebvre or the applicant here, the patentee here, had shown emails or had shown testimony of anybody else backing up what Mr. Lefebvre said all this meant, that would have been helpful. But now all we've got is him explaining some otherwise hard-to-explain logs. So logs aren't corroborated. When you say corroboration, it doesn't mean that it has to exist independently of the inventor's testimony. So you had PowerPoints, you had all this metadata, and so there is corroboration. Again, the PowerPoints are a good example. That related more to the actual reduction to practice, but there it would have been easy enough for him to get, for ATI to get the presenter of those PowerPoints to explain what those PowerPoints really meant instead of having Lefebvre secondhand saying what he thought they meant. Let me give you one piece of evidence why I think... That takes this notion of inventor testimony having to be corroborated to a sort of absurd level. In other words, you're saying that even if he's got documents and everything else that support what he's saying, that we have to ignore everything because he said it? Again, the documents don't support what he's saying unless they focus on the particular embodiment. In this case, they focus on a commercial embodiment. This is inter-parties. Where is the burden on the other side to inquire? After all, again, going back to the defunct interference practice, all that you need is somebody to sign read and understood on the bottom of every notebook. They don't have to show that they looked over the inventor's shoulder at every manipulation. That has never been the rule of corroboration. The corroboration has to do with that work was done on such a day. And if there's nothing on his face to cause the testimony to be suspect, you're saying there's no burden on the other side. It's just keep on producing page after page. This is quite a record as it is for a board proceeding. And never mind if we don't, whatever, I don't know if what. That's not enough. In this case, again, they never... From the position of the board, of the office, for the objective advice that applicants can rely on, what else was needed here for corroboration? Let me say one thing is clear. The applicant, the patentee in this case, has the burden of proof. He is trying to take advantage of the exception that allows him to get around the filing date of the otherwise 102E prior art. I think the law is very clear on that. There's a burden going forward which shifts when you make your prima facie case. And here we are. You're saying you never even had any prima facie case of continuous... I think the word prima facie case wasn't used, but I think that's an accurate representation of what the board's findings were. And those findings have to be upheld if substantial evidence supports them. And I think one of the strongest pieces... I can end with this, but I'd love to keep talking about it. One of the strongest pieces is the November 2002 PowerPoint presentation by ATI which said we're going to have prototypes out by June 25, 2003. So that's a half-year time frame in which they themselves said, exercising presumably reasonable diligence, they should be able to get prototypes of the fully developed R400 chip. But instead, from November 2002, that's part of the critical period, all the way until November of 2005, they still had not produced a commercial embodiment. From the blue brief, we understand that a commercial embodiment first came out in 2005. We don't know when they reduced it to practice, but it was certainly after the constructive reductions to practice we're debating here. And the point is the diligence period, one can measure it by that six-month period that I just mentioned coming from ATI itself. And the board, I think, reasonably said, you have to show where the diligence is. All they showed was that they were working very hard on a commercial embodiment for Microsoft that they ended up not producing. And that's engineering work, certainly. It justifies why they put this into logs, but it's not diligence to reducing this invention to practice. They could have. They had the conception, remember, back in August of 2001. They could have filed a patent application that date and constructively reduced it to practice. They took a risk by, I guess, not filing a patent application, finally getting around to it years later. But there are a lot of periods here where the diligence is simply not explained by these logs. And a more stringent requirement than they are suggesting must be applied. And I also want to be clear on the reduction to practice issue. So there's no debate. There was a complete waiver of that issue before the board. The board, again, at A26 and A30 said that. If they actually wanted to present the argument that the RTL code running on a machine was an embodiment of this claim, they should have made it then. I'll make one more point on that. They point to the board's claim constructions, which seem to contemplate that there might have been a software aspect for each of the elements. The board never said all of the elements would be software. And as the court has noted, all of the elements certainly were not software. The board's whole premise was the elements are hardware. But even beyond that, I don't think it would make the board's claim construction that they've pointed to was under the broadest reasonable interpretation, which is just an examination expedient to ensure that the claims that ultimately issue are appropriately cabined. It's not a claim construction to be applied in the reduction to practice context. And the reason for that is quite simple. If we applied the broadest reasonable construction rule for reduction to practice, we would grant patents on things based on embodiments that were outside the true scope of the claim. So then the patent would issue would be instantly invalid under the Phillips standard. So that is not the claim construction standard for reduction to practice. And I'm happy to answer any questions about any of that. I see my time is up. Any more questions for Mr. Krause? No, just a comment, Mr. Krause. And that is I know what we say up here is incredibly exciting. But you still have to wait for us to finish before you can talk. Thank you, Your Honor. Mr. Farron-Krug. May it please the Court. Before you do anything else, I want you to respond to the point that saying that you're working on R-400 generally is somewhat meaningless when that is simply a particular commercial embodiment that was being designed for a particular customer. And it doesn't necessarily tie that testimony to actually working on the elements of the claims. Your Honor, I respectfully disagree. The portions we referred to earlier, Mr. Lefebvre's testimony, points to the sequencer and graphics blocks as the metadata corresponding to that. All right. All he says is the metadata is not exhaustive of all sequencer shader pipe files that were edited during this time frame. But it shows that work was necessary for implementing the R-400 design. Isn't that a general reference to a particular chip that never materialized? The R-400 project was the entire context in which this invention arose and in which they were building the invention. So, yes, ATI is a company that produces, ultimately, commercial products, but the invention arose in the R-400, and hence Mr. Lefebvre's testimony that the metadata shows the work on the sequencer and graphics blocks, which Dr. Wolf elsewhere compared to the RTL or compared the claims to the RTL and determined all the elements were present. Mr. Lefebvre also testifies that the design and development of the R-400 generally was necessary to make progress on the sequencer block and the shader pipe block. So what he's testifying to here is that work was being done continuously around this invention to reduce it to practice, and it was constructively reduced to practice near the end of 2003 upon filing. And the evidence that he relies on does establish that everyday work was going on in the project. And because the board applied the incorrect standard of continuous diligence instead of reasonably continuous diligence, the board was looking for specific gaps on a day-to-day basis. It was looking for unexplained lapses. And indeed, it didn't even consider, had there been an actual reduction to practice, the only diligence period that ATI would have needed to establish was through August 2002 and only for the 053 patent and the stuttered reference. But on this constructive reduction, doesn't it mean that the best you could ever hope for would be a remand to the board to apply the correct standard? Because there would be an awful lot of fact-finding for us to have to say what these metadata showed. I believe a remand would be appropriate, Your Honor, to apply the correct standard to this particular factual record because the board did not apply that standard, and indeed it also didn't apply the correct standard for corroboration by saying on the one hand that the documents are not self-explanatory, and on the other hand, Mr. Lefebvre's testimony is uncorroborated. His testimony established what happened, and the documents corroborated that testimony. And so it is appropriate for a remand on the issue of applying the correct standard because that continuous diligence, looking for unexplained lapses, and articulating that ATI bore the burden. Your friend on the other side would say, We didn't. We assumed you worked every day. What we're saying is you didn't prove to us what you worked on. And I think even the application of the incorrect standard affects that analysis as well. The evidence as a whole is to be evaluated. What bothers me about that whole discussion is that you have a witness, admittedly he's the inventor, but he says, Look, here's all the record evidence, and this is what it means. And nobody says, No, that's not what it means. That's what bothers me. Yeah, I agree with that, Your Honor. I still think you're being, and I don't know whether you're being vague on purpose or whether it's because you can't answer my question. But the problem, again, is assuming we believe everything that Mr. Lefebvre says, assuming that it's corroborated by the metadata, we can't tell, and the Board couldn't tell, what was being worked on. That just general reference to a project is not the same as working toward the claim limitations. And that's what the problem is. There's just the Board found there was insufficient evidence that you were working toward the claim limitations, right? And so where's your evidence that that's what was going on? If it's only those two paragraphs in Mr. Lefebvre's testimony, then I've got a problem with that. That's not enough in my views. Well, Your Honor, there's the evidence in Mr. Lefebvre's testimony. There's the corresponding metadata. There is the RTL code. Tell me what the corresponding metadata shows you. I can't tell. If it's just calendars and the fact that people were doing stuff, that to me doesn't seem like enough. The calendars refer to the sequencer folder, sequencer parts folder, and the sequencer is established in the evidence as corresponding to elements of the claims by Dr. Wolfe and is explained the sequencer corresponds to the inventions in Dr. Lefebvre's testimony. There is also evidence in the record, some of the PowerPoint slides that are in the record, that refer to specific blocks being tested, for example, in the context of the first triangle test, like the sequencer and the shader pipe, and Dr. Wolfe examined those blocks in the RTL code and described exactly how those each correspond to the claim limitations. So there is evidence in the record that the claim elements were being worked on during this period, and at a minimum, a remand is appropriate for consideration of that evidence under the proper standard. I see I have exhausted my time. I'll accept any further questions. Any more questions? Okay, thank you both. The case is taken under submission. And that concludes this afternoon's argument for this panel. All rise.