[00:00:00] Speaker 04: The first is number 171570, Imperium IP Holdings versus Samsung Electronics. [00:00:09] Speaker 04: Mr. Edel. [00:00:26] Speaker 02: Go ahead. [00:00:26] Speaker 02: Good morning, Your Honors. [00:00:27] Speaker 02: May it please the Court. [00:00:29] Speaker 02: The board's obviousness determination is erroneous and should be reversed because it's based on incorrect assumptions about the Shimada reference, the only reference relied upon by the board for the 029 patent. [00:00:41] Speaker 02: With our time together, I'll focus on the board's assumption that Shimada teaches its preflash data and its lookup table are stored as separate files. [00:00:51] Speaker 02: This assumption is wrong because while Shimada discusses storing data in its raw memory, [00:00:56] Speaker 02: It never indicates that its preflash data is stored in a file that's separate from its lookup table. [00:01:02] Speaker 02: Without this assumption, the board had no basis to find that modifying Shimada would reduce the number of files stored in its raw memory, which is the key to the board's motivation findings. [00:01:14] Speaker 02: Now, using incorrect assumptions like this and hindsight bias, the board overstepped its authority by crafting its own sui sponte motivations to modify Shimada. [00:01:26] Speaker 02: The harm here is significant because the board's response and motivations are unsupported by the record and are fundamentally flawed as a technical matter. [00:01:35] Speaker 05: Are you saying, because this is a little bit different than the way you presented in your briefs, but are you saying that because of the fact that we're talking about one ROM that even storing these things in a different format doesn't necessarily conserve storage space because [00:01:56] Speaker 05: There's nothing to do with that storage space. [00:01:58] Speaker 02: Exactly, Your Honor. [00:01:59] Speaker 02: And that shows that when the board runs with its own theories and speculations about the technology, that the harm can rise with these sort of suicide motivations. [00:02:07] Speaker 02: Because ROM is static memory. [00:02:09] Speaker 02: It's burned into the chip at the time of manufacture. [00:02:14] Speaker 02: Any unused ROM space can't later be freed up by the computer to save information or delete to make up more space. [00:02:21] Speaker 02: So this shows that the board's speculation [00:02:24] Speaker 02: about the technology led to the erroneous result of finding motivation. [00:02:29] Speaker 02: And this court has explained in cases like Enrey Cowell and in LA Biomedical that the board errs when it speculates about the technology. [00:02:39] Speaker 04: I, like Judge O'Malley, see your argument as you articulated it a moment ago to be somewhat different than it was in the brief. [00:02:49] Speaker 04: Are you saying that Shimada doesn't store [00:02:54] Speaker 04: the information in two separate places? [00:02:58] Speaker 02: Your Honor, as Samsung set out in its arguments and as the evidence that it relies on shows that Shimada when it discusses its lookup table and when it discusses its... But you're not answering my question. [00:03:13] Speaker 04: Are you suggesting that Shimada doesn't show [00:03:18] Speaker 04: the storage of this information in two separate places, part of it in a lookup table and part of it elsewhere. [00:03:25] Speaker 04: Yes, Your Honor, that is an argument that Shimada does not show that these are stored at separate files, which is the assumption that... But the petitioner's expert said they were, and isn't that substantial evidence for the board to conclude that they were stored separately? [00:03:42] Speaker 02: Well, if we look at Mr. Porolsky's opinions, your honor, Mr. Porolsky does not address the issue of whether the preflash data is stored in a file separate from the lookup table. [00:03:54] Speaker 02: That's a conjecture that's brought up by the board on its own. [00:03:57] Speaker 02: There's no evidence from Mr. Porolsky or the only other evidence, the Shimada reference, that these sort of data points are stored in separate files. [00:04:05] Speaker 04: I thought he testified that the motivation would be to avoid separate storage. [00:04:10] Speaker 02: Mr. Prollsky testified in one paragraph, this is the entire basis of Samsung's motivation, in one paragraph, paragraph 108, that modifying Shimada would eliminate the separate data structures or memory locations. [00:04:24] Speaker 02: It never discusses the issue that was the basis of the board's conclusion about the different files. [00:04:30] Speaker 04: I understand that's a different issue, but to the extent that if one reads the board as saying it eliminates separate data structures, [00:04:40] Speaker 04: I heard you to suggest that that wasn't supported by substantial evidence, which seems to me difficult because the SAMHSA expert testified that there were separate data structures and that the motivation to combine would be based on the elimination of two separate data structures. [00:05:02] Speaker 02: Yes, but the problem with the board doing it to a specific motivation is that there is no connection in the record [00:05:07] Speaker 02: between data structures being eliminated and the data being different files that being consolidated would lead to a conservation of storage space. [00:05:17] Speaker 05: What you're saying is that Samsung argued one thing and the board relied on something else? [00:05:25] Speaker 02: Yes, Your Honor, and it's compounded here because the board's reliance on a different theory or speculation about the technology led to this incorrect assumption of conflating ROM memory with RAM memory. [00:05:37] Speaker 04: But they, I mean it's a question of how you read the board's opinion because at A-22 they do say that Petitioner argues precisely that Shimada teaches these annotations the person of ordinary skill and they have reason to modify the teaching to achieve these limitations of claims 1, 7, and 14, namely to store the various pieces of data in a single lookup table, which is [00:06:06] Speaker 04: the same, is it not, as Samson had argued? [00:06:10] Speaker 02: Well, that summary is reflected of what Samson's theory is, but when it comes to the actual motivation to do so, which is discussed later in the board's decision, the board's basis is the idea of reducing the number of files stored in ROM to conserve storage space. [00:06:25] Speaker 02: That was the actual motivation, and that's the focus of our argument here, Your Honors, [00:06:30] Speaker 04: Even if we take Samsung's arguments... If they said nothing more than they said at A22, the part that I just read you, what would be wrong with that? [00:06:43] Speaker 01: I'm just reading A22, Your Honor. [00:06:44] Speaker 01: Make sure I give you a full answer. [00:06:51] Speaker 02: I mean, the way I read it, Your Honor, is this statement is simply summarizing what the elements are taught in the prior art. [00:06:57] Speaker 02: It doesn't explain why a person of skill would choose to [00:06:59] Speaker 02: modify Shawana's lookup table to add these preflash values. [00:07:04] Speaker 04: Well, it says namely to store the various pieces of data in a single lookup table. [00:07:09] Speaker 04: That's what they're identifying, at least at that point, as the motivation. [00:07:14] Speaker 02: You're right, Your Honor, that one could interpret that, but that wasn't... If we did interpret it that way, what would be wrong with it? [00:07:21] Speaker 02: Well, what would be wrong is it's not supported by the evidence, Your Honor. [00:07:24] Speaker 02: How so? [00:07:25] Speaker 02: Because there's no evidence in the record. [00:07:27] Speaker 02: that what Samsung argues are the preflash values are stored in a separate file from the lookup table. [00:07:35] Speaker 02: So there's no benefit. [00:07:36] Speaker 05: Is it your position that tables and files are different? [00:07:39] Speaker 02: Absolutely, Your Honor. [00:07:41] Speaker 02: As we explained in our reply brief, that is another speculation that the board does, which doesn't comport with the actual facts. [00:07:49] Speaker 02: Because as we explained, if we do take the board's logic here, then a person of skill would have chose to simply move the preflash values [00:07:57] Speaker 02: and put them in the same file as a lookup table, not to include them in that lookup table, because Shimada teaches that what it wants to do is design a lookup table that uses a geometric sequence to organize main flash values, not main flash values and pre-flash values. [00:08:15] Speaker 02: Shimada's lookup table is focused solely on these main flash values. [00:08:19] Speaker 02: The only reason to add these pre-flash values is hindsight, your honor, because there is no benefit [00:08:25] Speaker 02: to, based upon the paragraph that we discussed on A22, or this idea of conserving storage space to move these together other than hindsight. [00:08:36] Speaker 05: Well, when you say other than hindsight, so what are you saying? [00:08:38] Speaker 05: There's no value in putting them together? [00:08:41] Speaker 02: Absolutely, Your Honor. [00:08:42] Speaker 02: It actually undermines Shimada's goals of designing a lookup table that includes just main flash values and uses a geometric sequence to organize those values. [00:08:51] Speaker 02: Now Samsung is trying to squeeze in pre-flash values, which are unrelated to the main flash values and are not based upon a geometric sequence, into the lookup table. [00:09:01] Speaker 02: It wants to tack those on to the end. [00:09:03] Speaker 02: Which brings us to the board's other Suasponti motivation. [00:09:06] Speaker 02: The board also said that a person would be motivated to change the data intervals in the entries for the lookup table to add the pre-flash values in without increasing the memory capacity used for the lookup table. [00:09:18] Speaker 02: Now this, again, is Suasponti. [00:09:20] Speaker 02: It wasn't [00:09:21] Speaker 02: raised or addressed or wasn't raised by... But you certainly had notice of the board's theory. [00:09:27] Speaker 05: I mean, that was in their institution decision, right? [00:09:30] Speaker 02: This latter motivation was not, Your Honor. [00:09:32] Speaker 02: This is a totally new motivation about changing the data intervals in order to add the pre-slash values. [00:09:37] Speaker 02: This was not an institution decision. [00:09:39] Speaker 02: It was put forward for the first time in the final written decision. [00:09:43] Speaker 00: Didn't you say in your... I guess I'm looking at at least a quote in the board opinion at 25. [00:09:50] Speaker 00: a quote from your patent owner response that seems to have pretty clearly understood the saving of storage space point being made. [00:10:01] Speaker 00: It says, patent owner contends that, quote, the institution decision presumes that Shimada must store a pre-flash preparatory power value and storage in one table rather than two may conserve storage space. [00:10:15] Speaker 00: I mean, doesn't that say we understand what the point is that's being made? [00:10:19] Speaker 02: Your Honor, this statement is accurate, but there's two motivations that the board relied on in its final written decision. [00:10:24] Speaker 02: The one that we've discussed, the idea of conserving storage space, which is what this sentence refers to, and the latter motivation, which is the idea that personally motivated to change the data intervals and the lookup table in order to add in these preflash values without making the table longer. [00:10:41] Speaker 05: But wasn't the reference to data integrals really almost identical to the reference to storage space? [00:10:49] Speaker 05: discussion of how to conserve storage space. [00:10:52] Speaker 05: Well, Your Honor, that's why the harm here is... I don't see them as two different motivations. [00:10:56] Speaker 02: Well, Your Honor, respectfully, I mean, this shows why the harm in making these sort of... the board going on its own with its own motivation theories is so problematic because we're left with guessing about what the board speculated and its motivation to modify. [00:11:10] Speaker 02: If we look at what it actually said, though, in its determination, it makes a distinction between [00:11:15] Speaker 02: Adding on a piece of preflash to the lookup table, which will make the table longer, which actually undermines Shimada's goals. [00:11:24] Speaker 02: That's the conservation of motivation theory. [00:11:27] Speaker 02: And then on the other hand, the board later explains that, wait, another motivation would be changing the data intervals so that the table isn't even longer when we add the preflash data values. [00:11:36] Speaker 02: So these two sort of response to motivations are not actually supported by anything other than [00:11:42] Speaker 02: Mr. Porolsky's conclusory opinions, because it's not in Shimada. [00:11:46] Speaker 02: And it's also inconsistent with the fundamental teachings of the technology, that in raw memory there is no benefit to this sort of conserving storage space. [00:11:56] Speaker 02: It wasn't recognized. [00:11:57] Speaker 05: I mean, the board really said that Shimada and 029 do basically the same thing, using basically the same equation, and it's just a slight difference in how they treat the variables, right? [00:12:09] Speaker 02: That was the board summary, but that is factually incorrect, Your Honor, because 029, and it's claimed in Claim 1 and the other independent claims, specifies that a lookup table has to include these preparatory values, which sometimes we refer to as preflash. [00:12:26] Speaker 02: The table has to include the preparatory strobe duration that's associated with a preparatory power value. [00:12:33] Speaker 02: Now, the board and Samsung concedes that [00:12:35] Speaker 02: Nowhere in the prior art has anyone ever done this where it uses these sort of preparatory values to organize a lookup table. [00:12:43] Speaker 02: Shimada points to using solely main flash values to organize this lookup table. [00:12:49] Speaker 02: And it does that because the main flash is being used with a lookup table. [00:12:56] Speaker 02: Shimada doesn't teach using a lookup table for its pre-flash procedures. [00:13:01] Speaker 02: So the 029 is starkly different than the Shimada reference. [00:13:05] Speaker 02: While they obviously do share some characteristics in common, namely predetermining preflash values and generating preflashes, the actual lookup table is organized extremely different and achieves different results. [00:13:20] Speaker 04: You're into your rebuttal time. [00:13:21] Speaker 04: Do you want to save the rest of it? [00:13:22] Speaker 04: Yes, Your Honor. [00:13:23] Speaker 04: Thank you. [00:13:36] Speaker 04: Mr. Pepe, is that how you pronounce it? [00:13:39] Speaker 03: Mr. Post, actually. [00:13:40] Speaker 03: Mr. Pepe is up next. [00:13:42] Speaker 04: Oh, really? [00:13:43] Speaker 04: OK. [00:13:44] Speaker 04: My sheet is wrong. [00:13:47] Speaker 03: OK. [00:13:47] Speaker 03: May I proceed, Your Honor? [00:13:48] Speaker 03: Yes. [00:13:50] Speaker 03: Good morning, Your Honors. [00:13:50] Speaker 03: May it please the Court? [00:13:52] Speaker 03: The issue in this appeal is a fairly narrow one. [00:13:56] Speaker 03: And the pivotal issue is whether the board's determination that these four pieces of data that we've been discussing so far this morning, namely preflash power, [00:14:06] Speaker 03: pre-flash duration, main flash power, and main flash duration, whether they be stored in a single table or stored in multiple locations within the same memory of Shimada, which is ROM 16. [00:14:19] Speaker 05: Let's then go to what I think, as your friend on the other side, what their strongest argument is, is the board found that conserving storage space in the ROM was really the motivation to put the two together. [00:14:34] Speaker 05: That's correct. [00:14:35] Speaker 05: It's not exactly how I understood you argued it, but that's what the board found. [00:14:40] Speaker 05: So if it is true that conserving storage space has zero value when there's no way you could repurpose that conserved storage space, then why would someone be motivated to even do that? [00:14:54] Speaker 03: So that I think what's underlying that point is the idea that that Ram and Ram that there's a difference that matters in this context. [00:15:01] Speaker 03: And we, we certainly disagree that that's the case. [00:15:04] Speaker 03: In either case, whether RAM or ROM, when the system is designed. [00:15:10] Speaker 05: So if... Just to answer the basic question, is there any value to conserving storage space? [00:15:16] Speaker 03: Absolutely, yes. [00:15:17] Speaker 05: All right. [00:15:17] Speaker 05: So what is the value to conserving storage space in ROM? [00:15:22] Speaker 03: So in that case, there are two potential benefits. [00:15:25] Speaker 03: One would be smaller storage space requires a smaller ROM at the outset. [00:15:29] Speaker 03: The other benefit is Shimada doesn't say, this is everything that's in the ROM. [00:15:34] Speaker 03: So there may be, in fact, other things. [00:15:36] Speaker 03: What we know from Shimada expressly is that the lookup table is stored in a file. [00:15:41] Speaker 03: So Shimada tells us that in paragraph 34. [00:15:44] Speaker 03: That was one point raised in the opening argument is that there's no connection between a lookup table and a file. [00:15:51] Speaker 03: That's incorrect. [00:15:52] Speaker 03: Shimada says. [00:15:52] Speaker 04: There's confusion here that what we're talking about is designing the ROM in the first place rather than rewriting what's already on the ROM. [00:16:04] Speaker 03: I think that is part of the confusion, your honor. [00:16:09] Speaker 03: What Shimada talks about in ROM is you have this lookup table. [00:16:13] Speaker 03: It is stored. [00:16:14] Speaker 03: You have k. That's a predetermined value stored through experiment beforehand. [00:16:19] Speaker 03: Those are certainly fixed values. [00:16:20] Speaker 03: That's what Shimada tells us. [00:16:23] Speaker 03: Once that's known, you have some size for these data structures, the lookup table in a file, k in some other structure. [00:16:32] Speaker 03: And if those are combined in one, [00:16:34] Speaker 03: like Mr. Parolsky set forth in his testimony, there would be a benefit. [00:16:38] Speaker 03: You would need less storage space for that one data structure than multiple data structures. [00:16:43] Speaker 05: Can you show me where your expert ever referred to storage space? [00:16:49] Speaker 03: So in the petition at page 33, note 20 was where this motivation to modify comes from. [00:16:56] Speaker 03: Mr. Parolsky's paragraph 108 was cited as support along with Shimada itself. [00:17:02] Speaker 00: I'm sorry to bother you. [00:17:04] Speaker 00: 108s and 33s don't do it for me when I have a joint appendix. [00:17:08] Speaker 03: So the joint appendix, so for response to Judge O'Malley's question, the Parulski Declaration paragraph 108 appears at appendix 1958. [00:17:22] Speaker 03: And it carries over to 1960. [00:17:25] Speaker 03: And you're on to the relevant language appears three lines from the bottom of 1958 where Mr. Parulski says, alternatively, [00:17:34] Speaker 03: A posita would also have found it obvious and advantageous to include the predetermined duration and corresponding relative amount of emitted light as for the preliminary emission in the lookup table. [00:17:46] Speaker 03: So he's saying, take that pre-flash power. [00:17:49] Speaker 05: Where are you? [00:17:50] Speaker 05: 1958. [00:17:51] Speaker 05: 1959. [00:17:51] Speaker 03: 1959. [00:17:53] Speaker 03: Yes, the paragraph begins. [00:17:55] Speaker 05: Very bottom. [00:17:56] Speaker 03: Very bottom. [00:17:56] Speaker 03: Last three lines, Your Honor, beginning with alternatively. [00:18:03] Speaker 03: So he's saying in those lines, take that preliminary, that pre-flash power and duration, those things we know are correlated from Shimada, put that in the lookup table, quote, because it would eliminate the need for a separate data structure or memory location for storing. [00:18:18] Speaker 05: So what I'm trying to understand is that in order for us to conclude that the board's reference to motivation to combine or discussion of motivation to combine was accurate in light of the record, [00:18:32] Speaker 05: we'd have to conclude A, that there was some value to conserving storage space, and that B, there was some relationship between your data structure position, which is all your expert talked about, and conserving storage space. [00:18:47] Speaker 05: And I'm not sure that I see where either is the case based on this record. [00:18:54] Speaker 03: I just read the portion of Mr. Parulski's declaration. [00:18:56] Speaker 05: But he doesn't talk about conserving storage space. [00:18:58] Speaker 03: He does talk about storage space. [00:19:00] Speaker 03: He doesn't use conservation as his language. [00:19:02] Speaker 03: That language first. [00:19:03] Speaker 05: He talks about data structures. [00:19:06] Speaker 03: He talks about eliminating the need for a separate data structure, correct? [00:19:09] Speaker 03: He's saying you have two or more, you only need one. [00:19:12] Speaker 05: How is data structure, the form of structure, the same as conserving storage space? [00:19:18] Speaker 03: So by going from two to one, two data structures of some form down to one, you would necessarily save storage space. [00:19:24] Speaker 03: What we know from Imperium's reply brief is that things like data structures files, Shimada again uses files. [00:19:31] Speaker 03: They have inherent properties, unique properties that go along with them overhead, if you will. [00:19:36] Speaker 03: And by eliminating the need for two of something and going to one, you have a savings there. [00:19:42] Speaker 03: What Mr. Parulski talked about throughout the trial proceedings. [00:19:45] Speaker 00: What were you? [00:19:48] Speaker 00: citing to or pointing to as establishing that a data structure has a certain amount of overhead that can be shared instead of duplicated. [00:19:58] Speaker 03: So that concept was one that actually Imperium set forth in its reply brief in the gray brief at page 16. [00:20:11] Speaker 03: The quote is, a file after all is itself a unique structure. [00:20:16] Speaker 03: with unique properties in computer science, something we agree with. [00:20:20] Speaker 05: That's it? [00:20:21] Speaker 03: That's correct, Your Honor. [00:20:23] Speaker 05: So wait, their reply brief is establishing the evidence that you needed to support the board's conclusions? [00:20:30] Speaker 03: No, no. [00:20:31] Speaker 03: To be clear, what they're talking about is they're acknowledging that files have properties, that files take space, that data structures have some space element. [00:20:42] Speaker 03: We think that that's undisputed. [00:20:44] Speaker 03: The evidence that the board relied upon was, again, Mr. Parulski's initial testimonies declaration, along with... But this whole point of the gray brick is to say that a file and a table can't be treated synonymously. [00:20:57] Speaker 05: Right, and we know from Shimada... And so if you're altering the table, and that's the data structure you're altering, you're not altering a file, right? [00:21:05] Speaker 03: Wrong, because Shimada, paragraph 34 says, our lookup table is stored in a file. [00:21:11] Speaker 05: Well, it's stored in a file, but... [00:21:14] Speaker 05: So I don't understand what you're, what you're benefiting, how the file is benefiting from altering the, from making the table shorter. [00:21:21] Speaker 03: So what's the benefit is, is more broadly to the memory itself. [00:21:25] Speaker 03: So ROM16 we know has this lookup table in a file. [00:21:28] Speaker 03: It has in Shimada a separate K value. [00:21:30] Speaker 03: And we know again, undisputed, that there is a duration that's associated with that file, with that K value, my apologies. [00:21:37] Speaker 03: The duration is what's used to control [00:21:40] Speaker 00: You say there is a duration associated with the K. Are you saying that somewhere in that ROM is recorded a duration associated with the K value, which is also recorded? [00:21:52] Speaker 03: Shimada doesn't say expressly. [00:21:54] Speaker 03: But what we know is that in accordance with Shimada paragraph 45, I believe, that appears at appendix 604, [00:22:05] Speaker 03: The amount of emitted light is controlled by duration. [00:22:08] Speaker 03: That's the variable that's used to ultimately fire the strobe. [00:22:12] Speaker 03: And Samsung established in the trial proceedings that there would be this correlation. [00:22:17] Speaker 03: That's something that's not disputed here on appeal. [00:22:21] Speaker 03: So what we don't know from Shimada expressly is where that duration appears. [00:22:26] Speaker 03: It could be in a table along with K. It could be in a third data structure. [00:22:30] Speaker 03: What we do know is by eliminating those separate structure or structures and including that information, [00:22:35] Speaker 03: in the lookup table, that file, that there would be a savings. [00:22:40] Speaker 05: I still don't understand. [00:22:42] Speaker 05: So you're restructuring the file, but that doesn't mean there's anything more you can do with that file or that there's anything more you can do with the ROM. [00:22:53] Speaker 05: So what's the value of restructuring the file? [00:22:56] Speaker 03: So there need not necessarily be a restructuring of the file itself, but we know from the 029 patent [00:23:02] Speaker 03: In Figure 6, this is the lookup table. [00:23:05] Speaker 03: It's at Appendix 58. [00:23:06] Speaker 05: You just told me that restructuring the table restructures the file. [00:23:11] Speaker 05: Now, are you backing off of that? [00:23:13] Speaker 03: Restructuring the... All I'm saying is that the lookup table is in a file. [00:23:18] Speaker 03: And what Mr. Parulski testified to, again, during the proceedings in his second deposition, this was something that he was examined on, was, well, you have this 64-row table. [00:23:28] Speaker 03: That's the format of Shimada's lookup table to file. [00:23:32] Speaker 03: If you're adding a row, is that a reformulation? [00:23:35] Speaker 03: Is that adding more space? [00:23:36] Speaker 03: He says not necessarily. [00:23:37] Speaker 03: It may be a small incremental addition. [00:23:39] Speaker 03: Or you can accomplish the same thing using that same 64 row table. [00:23:45] Speaker 03: And that was part of the discussion. [00:23:46] Speaker 03: That's where this idea of interval adjustment, I think, first became part of the proceedings. [00:23:51] Speaker 03: They were debating, how are we going to accomplish this conservation? [00:23:54] Speaker 03: How are we going to accomplish including these four pieces of information in that single table? [00:24:00] Speaker 03: And what he discussed, and this comes again from Shimada, is that if you do increase intervals, you have the potential of creating a coarser operating strobe. [00:24:11] Speaker 03: But what Shimada also says in that same paragraph 35. [00:24:14] Speaker 05: But that's a totally different potential motivation. [00:24:18] Speaker 05: It may have been one that you contemplated when you filed your petition, but it's not one that the board accepted. [00:24:25] Speaker 03: Respectfully, we think it is, that the idea of interval adjustment is an implementation [00:24:30] Speaker 03: question, if you're going to go from two or more data structures to one, what happens? [00:24:35] Speaker 03: How do you do that? [00:24:36] Speaker 05: What Mr. Parulski testified to is- But you still haven't told me how going to different data structures conserves any storage space and if there's any value in conserving storage space. [00:24:50] Speaker 05: Because the whole point is, is somebody going to be motivated to do this in order to save storage space? [00:24:57] Speaker 05: And A, if you're not really saving storage space, [00:24:59] Speaker 05: Or B, if saving storage space is useless, then you've got a problem with that motivation, right? [00:25:06] Speaker 03: So it's not useless. [00:25:08] Speaker 03: By having a smaller amount of memory, you either free up memory for some other use during programming, or you use a smaller ROM. [00:25:14] Speaker 03: Those are two benefits. [00:25:14] Speaker 05: So what's your response to your friend's argument that it actually does not free up memory for some other use? [00:25:20] Speaker 05: That would only be true if you have- Of course, didn't make that specific finding, by the way. [00:25:24] Speaker 03: Correct. [00:25:25] Speaker 05: But how do you free it up if it actually doesn't work that way? [00:25:29] Speaker 03: So only if you have some fixed-sized ROM and you have now less use, you have some unused space, in that circumstance, you wouldn't have a savings. [00:25:37] Speaker 03: But that's not what's contemplated here, where Shimada is, in fact, describing this information being determined experimentally and stored beforehand. [00:25:45] Speaker 03: So we do think you would get that savings. [00:25:47] Speaker 03: You could use a smaller ROM. [00:25:50] Speaker 03: As to Mr. Parulski's testimony about where that conservation comes from, again, [00:25:55] Speaker 03: He testified you have right now two or more. [00:25:58] Speaker 03: That's what Shimada says. [00:25:59] Speaker 03: You have a lookup table file. [00:26:00] Speaker 03: You have k. And you have this duration value. [00:26:03] Speaker 03: Again, the parties don't dispute. [00:26:04] Speaker 03: It's at least accessible to. [00:26:06] Speaker 03: This is the only memory in the Shimada system. [00:26:09] Speaker 03: So it's either there or accessible to that memory. [00:26:13] Speaker 03: That requires multiple. [00:26:16] Speaker 05: Am I correct that you need to account for adding k by increasing integrals, right? [00:26:25] Speaker 03: There are two ways that Mr. Perulsky testified you could do that. [00:26:28] Speaker 03: You could add a row. [00:26:29] Speaker 03: That would make a 64-row table a 65-row table. [00:26:33] Speaker 03: His testimony was that really wouldn't have much of an effect on the overall size. [00:26:38] Speaker 03: Or you could increase the interval to keep the 64-row structure and have one row that provides that K. But then there wouldn't be a saving. [00:26:48] Speaker 03: There would be a saving because you no longer have the data structure that used to store K. So is the idea that [00:26:54] Speaker 04: He certainly testified that there would be a motivation to combine the two into a single table. [00:27:00] Speaker 04: Is the idea that he testified that if you combine the two, then you would use the intervals to reduce the storage space? [00:27:08] Speaker 04: Is that the theory? [00:27:12] Speaker 03: He testified that there are a few alternatives. [00:27:14] Speaker 03: One thing he was examined on in his deposition was the idea [00:27:18] Speaker 04: that there could be an adjustment to the intervals, so that if you had... Does he say that there would be motivation to adjust the intervals? [00:27:25] Speaker 03: He did not say there was a motivation to adjust the intervals. [00:27:27] Speaker 03: That was an implementation question that he was being examined on. [00:27:30] Speaker 03: His motivation was, I don't need multiple structures. [00:27:34] Speaker 03: I can use only one, and I can use the existing data, or I can add this row. [00:27:39] Speaker 05: So he just thought, as someone who is looking at the computer structure, that's nicer to have. [00:27:47] Speaker 05: simpler data structures, but as somebody who's putting it together. [00:27:51] Speaker 05: But he did not testify to any advantage from that. [00:27:55] Speaker 03: He testified, as a person with 30 years in the industry, chief scientist at Digital Cameras at Kodak, that there would be an advantage. [00:28:02] Speaker 03: You would not have to have two data structures, you could have one. [00:28:05] Speaker 03: The board understood that to be the conservation of storage argument, what they, in the institution decision. [00:28:11] Speaker 00: Did it excite any evidence to support that proposition? [00:28:14] Speaker 03: They cited his testimony. [00:28:15] Speaker 00: What evidence did the other side introduce to say you don't actually save anything by eliminating a second memory location? [00:28:29] Speaker 03: There was no evidence to that point. [00:28:31] Speaker 03: The arguments that were made by Imperium below were really that this data was never used in the first place, and those aren't arguments that are made here. [00:28:39] Speaker 03: The Ram-Rom argument, the evidence that they purportedly cite to in the reply actually comes from the next appeal to different records. [00:28:45] Speaker 03: So there was nothing as to that argument either below or even in the opening brief, our contention is surely that's waived. [00:28:54] Speaker 03: Thank you. [00:29:04] Speaker 04: Thank you, Your Honors. [00:29:12] Speaker 00: Would you mind answering the question I just asked your friend on the other side? [00:29:16] Speaker 00: You're making an important distinction about, OK, so you get one memory location instead of two. [00:29:22] Speaker 00: And you're saying, so what? [00:29:23] Speaker 00: It doesn't save you anything. [00:29:24] Speaker 00: What evidence? [00:29:25] Speaker 00: And the board clearly understood their expert to kind of mean something like we're actually conserving resources. [00:29:34] Speaker 00: And that's a good thing. [00:29:35] Speaker 00: And whether the board could do that or not might depend on whether you gave them specific reason to think otherwise. [00:29:42] Speaker 00: And by reason, I don't mean lawyer reason. [00:29:46] Speaker 02: Yes, Your Honor, so there's two answers to that question. [00:29:49] Speaker 02: Two parts to answer, excuse me. [00:29:51] Speaker 02: The first is the evidence that all that we need comes from Shimada itself. [00:29:56] Speaker 02: Because as Samsung concedes, Shimada already teaches that its memory already includes the lookup table, the pre-flash duration, and the pre-flash power value. [00:30:10] Speaker 02: So these elements already exist in memory. [00:30:14] Speaker 04: That's not addressing the question that Judge Tarana asked you about, and that is whether you had a witness who said that combining these two different storage locations would not conserve storage space. [00:30:30] Speaker 02: Well, in the record before us, Your Honor, there is nothing that I can point to from our expert, Dr. Wright. [00:30:35] Speaker 02: Certainly Dr. Wright explained how these motivations were unfounded, but it focused on Samsung's theory about elimination of data structures. [00:30:43] Speaker 02: But if we actually look at what the evidence teaches, and again, the second part of my answer is going to be that it's always Samsung's burden to prove obviousness. [00:30:52] Speaker 02: And so all we have to do is show why there's no evidence to support that conclusion. [00:30:56] Speaker 02: But the evidence that we need to determine that Samsung's motivation is baseless is all we have to do is look at Shimada itself. [00:31:04] Speaker 02: Because Shimada already teaches that these three elements are in its memory. [00:31:08] Speaker 02: So moving all those pieces of information into one file does not conserve any storage space. [00:31:14] Speaker 00: It certainly would, wouldn't it, as a logical matter if the bookup table continued to be 64 by shrinking the gaps? [00:31:27] Speaker 02: That brings us to an important point, Your Honor. [00:31:29] Speaker 02: That argument was not put forth in Samsung's petition. [00:31:32] Speaker 02: At A108 is where Samsung puts forth its theory about modification. [00:31:37] Speaker 02: In there, it undisputedly says that their motivation will result in a lookup table that is longer, that it adds the pre-flash data to the pre-existing lookup, adds the pre-flash data to the pre-existing lookup table. [00:31:51] Speaker 02: So Shimada is actually getting a longer lookup table through this process. [00:31:56] Speaker 04: Now, moving these pieces of data into a different- I may be getting a longer lookup table, but the question is whether it takes less storage space than having the two separate. [00:32:07] Speaker 02: And again, your honor, this speaks to why the board should not go off on its own speculations, because there is no evidence that relates the ideas of Mr. Porolsky's discussion of data structures with the idea of conserving storage space by rearranging information in files. [00:32:22] Speaker 02: That is not in the record, and again, it's Samsung's burden to show that. [00:32:26] Speaker 04: Now, the... I think, unless there are other questions, we're out of time. [00:32:31] Speaker 04: Thank you, your honor. [00:32:32] Speaker 04: Thank you, Mr. Adel.