[00:00:02] Speaker 04: We'll have our argument next in 15-1640, U.S. [00:00:07] Speaker 04: Ethernet Innovations Against ACER. [00:00:31] Speaker 03: Mr. Herman, whenever you're ready. [00:00:34] Speaker 03: Good morning. [00:00:34] Speaker 03: May it please the court? [00:00:36] Speaker 03: The central issue in this appeal is anticipation. [00:00:40] Speaker 03: Does the sonic reference that's cited in the shared specifications of the patent contain each and every element of the asserted claims? [00:00:47] Speaker 03: The answer to that question is no. [00:00:49] Speaker 04: And that essentially resolves into a claim construction question on these patents or this patent? [00:00:57] Speaker 03: It could be a claim construction question, although the court did not [00:01:01] Speaker 03: offer a construction of the critical claim term. [00:01:04] Speaker 04: But the entire debate is about what is required by this one piece of claim language. [00:01:12] Speaker 03: I believe there is a, as you describe it, the requirement issued to meet the claim language contains a legal claim construction element and a fact question of what a PASIDA would see when they compared the claim language to the sonic device. [00:01:30] Speaker 03: This case can be decided, this issue in this case, can be decided reading the first column and a half of the patent. [00:01:37] Speaker 03: There's two types of prior arc buffers that are referenced. [00:01:40] Speaker 03: The first is a transmit data buffer that's expressly described as a buffer that stores data such that if there's a transmission error, the buffer doesn't have to go back to the host to redownload the data and start all over again. [00:01:58] Speaker 04: I did not, tell me why I'm wrong, I did not read that column as defining all transmit buffers as having that storage capacity, but rather referring to some set of buffers that have that capacity without saying that all of them do. [00:02:21] Speaker 03: Correct. [00:02:22] Speaker 03: I think it's noting that it is an advantage [00:02:24] Speaker 03: of a transmit data buffer, the ability to store the data. [00:02:27] Speaker 04: I have not made myself clear. [00:02:29] Speaker 04: Because of the absence of a definite article, a number quantifying article, it doesn't say in the prior ARC all dedicated transmit buffers have this storage capacity, but they are deficient. [00:02:45] Speaker 04: It says some network adapter interfaces include dedicated transfer buffers into which a frame can be stored. [00:02:53] Speaker 04: That doesn't say that all of the prior art dedicated transfer buffers have that property. [00:03:01] Speaker 03: I'm not sure that it's a distinction that merits a difference here. [00:03:05] Speaker 03: The reason I say that is because the noted advantage of the transmit data buffer was the ability to store the data. [00:03:13] Speaker 03: That was the advantage that the invention was trying to incorporate. [00:03:18] Speaker 03: And that's noted in the summary of invention. [00:03:21] Speaker 03: So it's trying to maximize the advantage of the transmit data buffer, the storage element so if there's an error in transmission, it doesn't have to go back to the host to redownload it. [00:03:31] Speaker 03: That was the express advantage of this invention. [00:03:34] Speaker 03: The disadvantage of the transmit data buffer was that it had to wait until all of the data of the frame was transmitted into the buffer. [00:03:42] Speaker 03: So there was a latency issue. [00:03:44] Speaker 03: The district court confused that downloading of the data so it didn't have to go back to the host to download it again. [00:03:51] Speaker 04: And do I take it that because you're starting with the description of the prior art, you don't really see much room for making an advance toward your position on the claim language? [00:04:06] Speaker 03: I think the claim language is instructive. [00:04:08] Speaker 03: I do. [00:04:08] Speaker 03: But I think before we get to the claim language, I think it's important, as a posita would do, to review what the prior art looked like in the specification. [00:04:17] Speaker 03: So we're starting expressly in the intrinsic record. [00:04:21] Speaker 03: And the intrinsic record talks about the advantages and disadvantages of the transmit data buffer, column one. [00:04:27] Speaker 03: The advantages and the disadvantages of FIFO buffer. [00:04:31] Speaker 03: The advantage being it was faster. [00:04:34] Speaker 03: It had throughput. [00:04:35] Speaker 03: The data came in and flowed out. [00:04:37] Speaker 03: The disadvantage was that data was not stored locally. [00:04:41] Speaker 03: It was unloaded, is the term that's used in column one of the patent. [00:04:44] Speaker 03: So in other words, if there was a transmission error, the FIFO buffer had to go back to the host [00:04:50] Speaker 03: re-download the data and start all over again. [00:04:52] Speaker 04: Does claim one require the ability to make a second attempt at transmission if the first fails? [00:05:02] Speaker 03: No, but it does require the ability of the buffer to store the data for transmission. [00:05:10] Speaker 02: That claim language... But the claim doesn't talk about storing frames of data. [00:05:17] Speaker 02: That's correct. [00:05:17] Speaker 02: It's talking about storing data. [00:05:20] Speaker 02: Correct. [00:05:21] Speaker 02: Not necessarily a complete frame. [00:05:24] Speaker 03: I wouldn't disagree with you, Your Honor. [00:05:26] Speaker 03: For successful transmission to occur, for this frame of data to be transmitted, it had to go through 64 bytes. [00:05:36] Speaker 03: That's called the collision window. [00:05:38] Speaker 03: There's testimony from all three inventors in our brief that talk about this collision window. [00:05:43] Speaker 03: There's testimony from the experts that talk about the collision window, the first 64 bytes. [00:05:47] Speaker 03: As long as it gets through that, [00:05:49] Speaker 03: the likelihood of a collision goes way down and successful transmission occurs. [00:05:55] Speaker 02: But there's no requirement that the buffer store halt 64. [00:06:02] Speaker 03: I believe the patent language to successfully transmit, it would have to have the first 64 bytes. [00:06:10] Speaker 03: So it didn't have to go back to the host to redownload the data. [00:06:14] Speaker 02: Maybe that's in the descriptive materials. [00:06:17] Speaker 02: But I'm talking about the claims. [00:06:19] Speaker 03: Yeah, our view is the claim language storing data for transmission requires the element of the transmit data buffer to have that data locally to ensure transmission. [00:06:33] Speaker 03: So it doesn't have to go back to the host computer and interrupt it. [00:06:36] Speaker 00: Now, importantly, this is- And for that construction, you rely on isolated places in the specification. [00:06:43] Speaker 03: Is that right? [00:06:43] Speaker 03: There are multiple quotes in the specification [00:06:46] Speaker 03: including the summary of invention, which talks about this being a transmit data buffer, but they talk about having resonant on the adapter the entire frame of data. [00:06:58] Speaker 03: Now, this isn't a case where there's paid experts on both sides coming up with different conclusions. [00:07:05] Speaker 03: There are 10 positas in this record who have come out and said, we read the patent specification, we read the claim language, it's our understanding that [00:07:16] Speaker 03: You have to store a frame of data on the adapter. [00:07:18] Speaker 04: Didn't we say in the remand decision in Teva, basically following what the Supreme Court had said in Teva, that the fact that an expert says, in what by definition is extrinsic evidence because it's newly created, says, here's how I interpret the intrinsic evidence. [00:07:43] Speaker 04: That doesn't make that into a fact question. [00:07:46] Speaker 04: Fact questions are questions about facts, about facts external to the patent. [00:07:53] Speaker 04: Like when we in the MIT or Harvard computer science department use this language, this is what we mean. [00:08:01] Speaker 04: So it probably has that meaning in the patent. [00:08:04] Speaker 04: But is any of the expert testimony here about the interpretive question of that sort? [00:08:12] Speaker 04: Or is it all, I've read all [00:08:15] Speaker 04: many columns of the patent and the claims, and this is how I would interpret it. [00:08:19] Speaker 04: That's not the kind of expert testimony that turns it into a fact question. [00:08:25] Speaker 03: We don't disagree with Your Honor's reading of Teva. [00:08:28] Speaker 03: The point we would make is it's the second prong of the anticipation analysis. [00:08:33] Speaker 03: What evidence is there that a posita would understand this sonic device to read on the claim language? [00:08:40] Speaker 03: So that is a quintessential fact question. [00:08:44] Speaker 03: And the facts here are we have ten positas. [00:08:47] Speaker 00: But before we get that, we have to determine whether the claim language means what it says you say it means or what your friend says it means. [00:08:55] Speaker 00: And the expert testimony about whether this prior art reads on it or not is irrelevant to that question. [00:09:02] Speaker 03: Isn't it? [00:09:03] Speaker 03: I agree that it is less relevant to the legal question. [00:09:07] Speaker 03: Yes, it's less probative, I believe is what the court's holdings are. [00:09:11] Speaker 03: But despite [00:09:13] Speaker 03: Despite that, I think the fact question here, anticipation was granted on summary judgment in California. [00:09:19] Speaker 03: What we're saying is there was demonstrative weight of evidence showing that the fact question should be resolved in our favor. [00:09:26] Speaker 03: The only witness to come forward and offer a different conclusion was Dr. Wicker, a paid expert for the defense. [00:09:34] Speaker 03: And Dr. Wicker didn't review, in his opinion, in detail. [00:09:39] Speaker 03: the language of the specification, the requirements of the transmit data buffer, the advantages of the transmit data buffer. [00:09:46] Speaker 03: That was given very cursory review when it was compared to the storing data frames requirement in the claims in comparison to the sonic reference. [00:09:54] Speaker 03: So we think given the weight of that evidence. [00:09:58] Speaker 04: Can I ask a clarification of, and the question itself may be confused, but is your view that [00:10:08] Speaker 04: the claim language has to be read so that the buffer can hold an entire frame full stop or, more aggressively, an entire ethernet frame. [00:10:27] Speaker 04: So that if you didn't have an ethernet, am I right? [00:10:29] Speaker 04: The claim doesn't refer to ethernet, is that right? [00:10:31] Speaker 04: Not directly in the claim language. [00:10:32] Speaker 04: Right. [00:10:33] Speaker 04: So if you had a protocol that had 32-byte [00:10:38] Speaker 04: frames, it still might require a storage that the buffer be at least 32 bytes big plus whatever else it needs to hold the 32 bytes. [00:10:53] Speaker 04: But that still wouldn't be enough for you. [00:10:55] Speaker 04: You need to get it into the 64-byte range. [00:10:57] Speaker 04: Is that right? [00:10:58] Speaker 03: The language in this column one is it requires storage for successful transmission. [00:11:05] Speaker 03: Under the Ethernet standard, that's 64 bytes. [00:11:07] Speaker 04: Do you need to read Ethernet standards into this claim? [00:11:13] Speaker 03: We think that the patent is focused squarely on Ethernet, although it doesn't appear directly in the claim language. [00:11:20] Speaker 03: The specification is replete with Ethernet. [00:11:23] Speaker 03: But the important point from our point, if you want anticipation, is there's no dispute that the Sonic doesn't store at all. [00:11:31] Speaker 03: The Sonic unloads data. [00:11:33] Speaker 03: So if there's an error, [00:11:35] Speaker 03: It has to go back to the host where the memory is to re-download the data and start all over again. [00:11:41] Speaker 03: So this ephemeral temporary storage the defendants call about, in reality, the data is not there. [00:11:47] Speaker 03: It's unloaded out of the sonic. [00:11:50] Speaker 03: And that's described pretty clearly as a disadvantage of a FIFO in column one of the specification. [00:11:56] Speaker 03: And that concept never made it into the district court. [00:11:59] Speaker 03: The glaring error, and it was a misbetrayal of the technology that the district court accepted, [00:12:03] Speaker 03: was that the sonic FIFO unloads, stores data in the same way that the transmit data buffer stores data. [00:12:12] Speaker 03: And that is incorrect. [00:12:14] Speaker 03: The transmit data buffer has it resident locally on the adapter so that it didn't have to go back to the host. [00:12:21] Speaker 03: The FIFO unloads the data so it does have to go back to a host. [00:12:26] Speaker 03: And that created enormous performance problems in the FIFO. [00:12:30] Speaker 03: And the record has numerous testing [00:12:33] Speaker 03: both by defendants here and by 3Com showing the superior performance of this product, it increased by well over 100% compared to the prior order. [00:12:43] Speaker 03: And it sold as a result. [00:12:45] Speaker 03: Over $2 billion worth of this product was sold. [00:12:48] Speaker 04: You're into your rebuttal time. [00:12:50] Speaker 04: I don't know whether you want to use it now or save it. [00:12:53] Speaker 04: I'll reserve it. [00:12:54] Speaker 04: Thank you. [00:12:59] Speaker 04: Mr. Stevens? [00:13:00] Speaker 01: Yes, Your Honor. [00:13:01] Speaker 01: Good morning. [00:13:03] Speaker 01: We heard a lot about various advantages that are described in the description of the prior art. [00:13:10] Speaker 04: Can you actually start there? [00:13:12] Speaker 04: Sure. [00:13:13] Speaker 04: How, in your view, in a way that I might be able to understand, does this claim identify some advance over the particular prior art described in the background section? [00:13:32] Speaker 04: as having disadvantages, if you're right in your view about the claims scope. [00:13:37] Speaker 01: Certainly, Your Honor. [00:13:39] Speaker 01: So the claims refer to monitoring the transfer of data into a buffer memory in order to make a threshold determination based on an amount of data that's been transferred. [00:13:51] Speaker 04: When we get this amount, we're going to start sending it along. [00:13:53] Speaker 01: That's correct, Your Honor. [00:13:55] Speaker 01: So in response to that threshold, it initiates transmission early. [00:13:58] Speaker 01: That's why they call them the early transmit patents. [00:14:00] Speaker 01: That is not described in the background section that discusses the prior art in column one at all. [00:14:07] Speaker 01: But it is disclosed in the sonic data sheet, which is not cited prior art. [00:14:10] Speaker 04: And the reason it's not disclosed is that the FIFO system uses no thresholds at all? [00:14:17] Speaker 04: Is that right? [00:14:18] Speaker 04: As described in the... I mean, is that kind of a, I don't know, a superconducting buffer? [00:14:22] Speaker 01: Which there's no resistance? [00:14:25] Speaker 01: Not at all, Your Honor. [00:14:25] Speaker 01: So it goes in one end, it comes out the other. [00:14:28] Speaker 01: First in, first out. [00:14:30] Speaker 01: And no threshold is required to use a FIFO to buffer in that way. [00:14:35] Speaker 01: So the invention, as described in the summary of the invention and as claimed in all the claims in the 872 patent, include this threshold determination. [00:14:44] Speaker 01: And that isn't discussed anywhere in the prior art section of the patent. [00:14:49] Speaker 01: However, it is disclosed in the data sheet for the Sonic. [00:14:53] Speaker 01: And we know that the data sheet for the Sonic was not in [00:14:57] Speaker 01: not before the examiner because it's not cited on the front of the patent, nor does it appear anywhere in the file history. [00:15:03] Speaker 01: So it appears that whoever wrote the patent just didn't understand because they didn't have the description of the sonic that discloses the threshold determination and the early initiation of transmission that is clearly set forth in the sonic data sheet. [00:15:19] Speaker 01: So they claimed that, even though that doesn't distinguish the sonic. [00:15:23] Speaker 01: And what they described in the prior arc for FIFOs [00:15:27] Speaker 01: was a more conventional system where data just went in one end and came out the other, and you began transmitting when the FIFO was full, instead of before the FIFO was full. [00:15:39] Speaker 01: And in fact, if you look at the claim language of claim one of the 872 pattern, which is reproduced in our brief, you'll see that the phrase data of frames as used in the claim, and in the particular phrase that Mr. Herman refers to, [00:15:57] Speaker 01: must refer to less than all of the data of the frame precisely because of this threshold determination that happens. [00:16:05] Speaker 01: So we see first the buffer memory for storing data of frames in the first clause, and of course there was an agreed construction below that buffer memory means a memory for temporary storage of data. [00:16:16] Speaker 01: So there's no dispute that that buffer memory stores data. [00:16:20] Speaker 01: And there's also no real dispute that all the data that goes through that FIFO is frame data that's to be transmitted. [00:16:27] Speaker 01: That's clearly disclosed in the Sonic as explained by the defendant's expert below. [00:16:34] Speaker 01: So then you look further down the claim and you see the means for monitoring the transfer of data of a frame to the buffer memory to make a threshold determination of an amount of data of the frame transferred to the buffer memory. [00:16:48] Speaker 01: That's at line 15. [00:16:50] Speaker 01: of column 30. [00:16:52] Speaker 01: The next phrase refers to a means that's responsive to the threshold determination of the means for monitoring for initiating transmission of the frame prior to transfer of all of the data of the frame. [00:17:05] Speaker 01: So we know that when the claim uses the phrase data of the frame, it's referring to less than all of the data of the frame. [00:17:14] Speaker 01: And when the claim means all of the data of the frame, those are the words that it uses. [00:17:19] Speaker 01: That same kind of distinction is also found in the description of the prior art itself. [00:17:29] Speaker 01: So if you look at column one, you see around line 49 a description of downloading data of a frame into the FIFO. [00:17:40] Speaker 01: So as Mr. Herman says, these FIFOs could not hold an entire frame. [00:17:45] Speaker 01: So when the patent talks about data of a frame, [00:17:49] Speaker 01: being stored in a buffer memory, it's talking about less than all of the data of a frame. [00:17:55] Speaker 01: And when it's referring to an entire frame of data, as you see around line 40 in column one, that's what it says. [00:18:01] Speaker 01: It says the frame is stored in the transmit data buffer. [00:18:06] Speaker 01: So again, there's no real issue here. [00:18:09] Speaker 04: So just say again, what do you think, under your view of the claim, the [00:18:16] Speaker 04: Advantage would be over a FIFO system? [00:18:20] Speaker 01: The advantage, Your Honor, is it doesn't have to wait until the FIFO is full to begin transmission. [00:18:24] Speaker 04: Is there something in column one that says FIFOs wait until the buffer is full before? [00:18:33] Speaker 01: That was the other one. [00:18:36] Speaker 01: This is where it starts talking about unloading the FIFO during a transmission, right? [00:18:40] Speaker 01: So this again is around line 47, I guess. [00:18:46] Speaker 01: First in, first out FIFO systems in which the sending system downloads data of a frame into the FIFO while the network adapter unloads the FIFO during a transmission. [00:18:57] Speaker 01: So it goes in one end and it comes out the other. [00:19:00] Speaker 01: Whereas the Sonic, like the claims, monitors the downloading and when a threshold has been met, it begins transmitting. [00:19:09] Speaker 04: I think I'm just going to go over old ground, but I'm still confused. [00:19:15] Speaker 04: background section, this two-thirds of a column or something, describes two different kinds of prior art. [00:19:22] Speaker 04: Yes. [00:19:23] Speaker 04: One, not the FIFO, says you put stuff into a buffer, you have to wait until it's all there before you send it along, and the problem with that is it delays transmission. [00:19:32] Speaker 04: Now let's talk about FIFO. [00:19:35] Speaker 04: Forget about the first thing for a minute. [00:19:38] Speaker 04: Under your view of the claim, what is the advantage over the FIFO system? [00:19:45] Speaker 04: as described in column one. [00:19:47] Speaker 01: So the advantage of the sonic system and what's claimed is that it does not need to wait until the FIFO is full. [00:19:54] Speaker 04: Where in column one does it describe the FIFO system as delaying its sending along until the FIFO buffer is full? [00:20:07] Speaker 01: It does not say that, Your Honor. [00:20:08] Speaker 01: But it also does not mention this threshold ability, which is claimed and described in the summary of the invention and also in the sonic data sheet. [00:20:17] Speaker 01: So it's not described in the description of the prior art, I believe, because the prosecuting attorney simply didn't realize that the advantage that it was talking about in the summary of the invention was actually present in the sonic. [00:20:32] Speaker 01: If they'd known that, presumably they would have cited the data sheet, for example, to the examiner and would have claimed something different. [00:20:39] Speaker 04: Was any of the expert testimony or was there other evidence that described how FIFO systems worked? [00:20:50] Speaker 01: Oh yes. [00:20:52] Speaker 01: The experts for the defendants below put in a very extensive report [00:20:57] Speaker 01: describing in detail how FIFO systems work and explaining how the Sonic did something different than... Can you point me? [00:21:03] Speaker 01: I'm afraid I can't, Your Honor. [00:21:05] Speaker 01: I do believe that the... What's the name of the expert? [00:21:09] Speaker 01: Wicker. [00:21:10] Speaker 01: That's Professor Wicker. [00:21:11] Speaker 01: This is the so-called paid expert. [00:21:13] Speaker 01: Well, he was hired by the defendants as experts typically are hired by one side or the other. [00:21:18] Speaker 01: I would like to mention briefly the discussion that Mr. Herman had about all the various advantages [00:21:26] Speaker 01: that are described in the prior art. [00:21:28] Speaker 01: And, of course, this court has held in Phillips that the fact that the written description sets forth multiple advantages doesn't mean that the claim has to have all of those. [00:21:38] Speaker 01: And, of course, the claim language here is conspicuously missing. [00:21:42] Speaker 01: Most of the things that Mr. Herman describes as advantages, none of the claims recite the word Ethernet. [00:21:48] Speaker 01: None of them recite a transmit data buffer or a dedicated transmit buffer. [00:21:54] Speaker 01: which is discussed here in the background section. [00:21:57] Speaker 01: And in fact, when the patent prosecutor wanted to claim various different features of the buffer memory, he did so in various other claims. [00:22:07] Speaker 01: So for example, claim seven says wherein the buffer includes a transmit descriptor ring buffer and other characteristics. [00:22:16] Speaker 01: Those are not found in claim one. [00:22:18] Speaker 01: So the examiner, excuse me, the prosecuting attorney was perfectly capable here of [00:22:24] Speaker 01: including various structural limitations of the kind that Mr. Herman would have the court read in, but chose not to include those in the claims that are at issue before the court today. [00:22:38] Speaker 01: Your Honor, the file history also includes a relevant portion here. [00:22:44] Speaker 01: There was a Rule 131 Inventor Declaration filed, and this was in the record below. [00:22:50] Speaker 01: and it's referenced at 9376 in the joint appendix, the inventor said to the patent office, of course, to overcome a prior art reference, that a buffer memory storing a number of bytes of a packet that is storing data of frames as recited by the claim. [00:23:08] Speaker 01: So the inventor said to the patent office during prosecution that storing data of frames as recited by claim one [00:23:17] Speaker 01: was simply storing a number of bytes of a packet. [00:23:21] Speaker 01: And so we would suggest that the intrinsic evidence here is very clear, and it does not require storage of an entire frame or any of the other benefits that Mr. Herman suggested like retransmission. [00:23:40] Speaker 01: If the court has no other questions, I'll sit down. [00:23:49] Speaker 02: Mr. Herman. [00:23:49] Speaker 03: Thank you, Your Honor. [00:23:51] Speaker 03: I think, Your Honor, put your finger on the exact issue. [00:23:55] Speaker 03: There were two advantages, one for a FIFO, one for a transmit data buffer. [00:24:00] Speaker 03: The advantage of the FIFO was the throughput. [00:24:02] Speaker 03: It hit the buffer and went immediately out. [00:24:04] Speaker 03: It was not stored locally on the buffer, so it was fast. [00:24:08] Speaker 04: What does the record tell us about, I mean, I guess as a non-expert here, I'm imagining sort of two, and I think maybe I've even heard two different descriptions. [00:24:17] Speaker 04: One is first in, first out, even before the FIFO is filled, FIFO buffer is filled, and the other is wait until it's filled and then it just goes out the back end and it's never to be heard from again. [00:24:31] Speaker 04: What does the record tell us about those two possibilities being under the FIFO label as it's used in column one or anything else? [00:24:41] Speaker 03: It's a highly technical issue that I'm not sure bears on the outcome, but my understanding is [00:24:47] Speaker 03: that the FIFO would speed match the speed of the output. [00:24:51] Speaker 03: So the data would flow through based on how quickly it could leave the buffer. [00:24:56] Speaker 03: Yeah, so it doesn't get blocked at the entrance. [00:24:58] Speaker 03: Correct. [00:24:58] Speaker 03: So the point is it was not stored. [00:25:01] Speaker 03: The patent describes a column one as being unloaded. [00:25:05] Speaker 03: So if there's a transmission error, it's not there in the FIFO. [00:25:08] Speaker 03: That was the disadvantage. [00:25:09] Speaker 03: The advantage of the transmit data buffer, the advantage they don't want to talk about, and Jorana correctly pointed this out, was that it did store. [00:25:16] Speaker 03: So that if there was a transmission error, it didn't have to go back to the host. [00:25:22] Speaker 03: It was stored locally. [00:25:23] Speaker 03: It could go retransmit out of the buffer. [00:25:27] Speaker 04: Was there any evidence in the record about the range of technical functionalities of FIFO that said some of them start the exiting when, I mean, I assume in the FIFO there is a buffer, right? [00:25:46] Speaker 04: Is it some kind of buffer or no? [00:25:49] Speaker 03: It's a, it is a buffer. [00:25:50] Speaker 04: Something. [00:25:50] Speaker 04: So, so when it is, you know, half full, when it is completely full, regardless, it just, you know, the soon as the first bite comes in, that first bite is going out the exit. [00:26:03] Speaker 04: Um, or some FIFO systems, this is, I guess I'm imagining the possibility of what feels at least conceptually, like it might be helpful evidence to you. [00:26:13] Speaker 04: Some FIFO systems, called FIFO systems, allowed the user or somebody to set when the exit started. [00:26:22] Speaker 03: And I think that's right. [00:26:23] Speaker 03: I think there was an ability to say, OK, when you get eight bytes. [00:26:26] Speaker 04: My question specifically, is there something I can look at in the record that says that? [00:26:31] Speaker 03: I believe the sonic data sheet talks about being able to set a number when the bytes start to leap. [00:26:38] Speaker 03: But the point is they're not stored there for successful transmission. [00:26:42] Speaker 03: They don't meet the claim language. [00:26:44] Speaker 03: If there is a problem in the transmission, it doesn't have the advantage of the transmit data buffer being stored locally. [00:26:52] Speaker 03: So even though they may be there, they're not in a memory format like a transmit data buffer that allows the adapter to go ahead and retransmit it. [00:27:03] Speaker 03: The advantage was the storage. [00:27:05] Speaker 03: The claim element was storing the data, the frames. [00:27:07] Speaker 03: This issue was considered by two previous courts and two previous juries, and they all found our way. [00:27:13] Speaker 04: Speaking of two courts, why don't we wrap this argument up and move to the next one. [00:27:18] Speaker 04: And if you want to stay at the podium or...