[00:00:00] Speaker 00:
Tech 2016, 1956 and 1957.

[00:00:41] Speaker 00:
Mr. Silvera?

[00:00:44] Speaker 04:
Thank you, Your Honor.

[00:00:44] Speaker 00:
Good morning.

[00:00:45] Speaker 04:
Good morning.

[00:00:46] Speaker 04:
May it please the Court, Matthew Silvera for Synopsys, Inc.

[00:00:50] Speaker 04:
Synopsys appeals two unpatentability determinations of the board, one on anticipation grounds and the other on obviousness grounds.

[00:00:58] Speaker 04:
Now, as Appellee's brief makes clear, affirming the board's determination on claim 32 would stretch the law of anticipation beyond all imagination.

[00:01:11] Speaker 04:
The obviousness determination on independent claim one is also fatally flawed.

[00:01:16] Speaker 04:
And both claims require reversal, which means the claims depending from claim one also require reversal.

[00:01:22] Speaker 04:
So I'm going to start with claim 32, which recites a method for fitting an integrated circuit design within a predefined area.

[00:01:30] Speaker 04:
And there are two limitations.

[00:01:31] Speaker 04:
You have your limitation of determining optimal placement of hard blocks within the design.

[00:01:37] Speaker 04:
And hard blocks essentially have a physically defined shape.

[00:01:41] Speaker 04:
Then the second step involves selecting the location for soft blocks and hierarchical blocks to fit around the hard blocks.

[00:01:49] Speaker 04:
Now all we had an issue here is anticipation.

[00:01:52] Speaker 04:
AtopTech actually asked to institute on obviousness as well, but the board denied it.

[00:01:58] Speaker 04:
So that's not an appealable decision because the board didn't institute on obviousness in the first place.

[00:02:03] Speaker 04:
It means all we're left with is anticipation on claim 32.

[00:02:06] Speaker 04:
As to anticipation, what the parties argued before the board was express disclosure or inherent disclosure.

[00:02:14] Speaker 04:
And you see this in the expert report of Dr. Gyasi.

[00:02:17] Speaker 04:
You see this in the briefing.

[00:02:18] Speaker 04:
You see this in the oral hearing.

[00:02:20] Speaker 04:
Consistently, it's express or inherent disclosure.

[00:02:23] Speaker 04:
And the board found in footnote two of its opinion at appendix 27, it found there's no inherent disclosure here.

[00:02:30] Speaker 04:
So the determining optimal placement limitation is not necessarily disclosed, obviously, within the Sioux reference, which is the sole anticipatory reference.

[00:02:42] Speaker 04:
So that leaves us with express disclosure.

[00:02:45] Speaker 04:
And if you look at the Sioux reference, there's simply no way to find substantial evidence of

[00:02:50] Speaker 04:
determining optimal placement of hard blocks within SUE.

[00:02:54] Speaker 04:
SUE isn't even focused on hard blocks.

[00:02:56] Speaker 04:
It's all about a performance-driven soft macro clustering and placement method.

[00:03:02] Speaker 04:
And soft macros are essentially the equivalent of soft blocks within the 967 pattern.

[00:03:06] Speaker 02:
And just to be clear, based on the prior art of record here, it seems to me there are multiple ways to skin a cat.

[00:03:13] Speaker 02:
You could improve performance by reducing efficiency, reducing heat consumption, reducing power.

[00:03:18] Speaker 02:
And placement of macros can result in improved efficiency, and in particular,

[00:03:25] Speaker 02:
reduced heat which is part of what the problem here is about where you locate them and you can do that by carefully addressing where to locate hard macros or alternatively carefully addressing where to locate soft macros or possibly both.

[00:03:41] Speaker 02:
Is my understanding of the technology correct?

[00:03:43] Speaker 04:
That's right and you know soft macro placement is the focus of Sue and that's very clear.

[00:03:49] Speaker 04:
The only disclosure of any hard macro placement

[00:03:52] Speaker 04:
in Sue is simply to say, well, first you place your hard macros, and then let's get into the real meat of this reference, and that's all about the soft macro placement.

[00:04:01] Speaker 02:
And it's all about creating an algorithm that carefully articulates, regardless of where the hard macros are placed, almost,

[00:04:09] Speaker 02:
They're each placed into various regions.

[00:04:12] Speaker 02:
But then what this is all about is maximizing performance by then locating the soft macros according to a precise algorithm around the various hard macros in a way that will reduce heat, among other things.

[00:04:26] Speaker 04:
Exactly.

[00:04:26] Speaker 04:
That's absolutely correct, Your Honor.

[00:04:28] Speaker 04:
And so what we have here is there's just no disclosure of hard macro or hard block placement, certainly not determining optimal hard block placement in the Sioux reference.

[00:04:37] Speaker 02:
And so the one thing the reference does say is according to some commercial floor plan.

[00:04:41] Speaker 02:
Sure.

[00:04:42] Speaker 02:
You know, and so what made it really confusing to me is that clearly to me does not expressly disclose how the floor planning module will distribute the hard blocks.

[00:04:55] Speaker 02:
It seemed to me that I was genuinely confused by the board's determination that there was express.

[00:05:02] Speaker 02:
No, there was express, but no inherent.

[00:05:05] Speaker 02:
I kind of thought based on the expert testimony that if you had anything here, you had inherent.

[00:05:11] Speaker 02:
Because what the expert said is commercial floor planning, people would know, would choose some means of doing so.

[00:05:20] Speaker 02:
That seems to me that it would fit, if anywhere,

[00:05:23] Speaker 02:
in the inherent camp?

[00:05:24] Speaker 02:
Is that your understanding?

[00:05:25] Speaker 04:
I think that's if anywhere and I don't think it even would be inherent in this case but I do think that's if anywhere and in fact if you look at the expert's testimony at his deposition the expert said well really this you know certainly this would be obvious at least and I think that's ultimately what we're dealing with here.

[00:05:42] Speaker 04:
is that what they're really arguing is obviousness.

[00:05:44] Speaker 04:
And that's why in Pelli's brief, they fall back on this notion of implicit disclosure, which they don't define.

[00:05:50] Speaker 04:
They focus on a skilled artist and what a skilled artist would do.

[00:05:53] Speaker 04:
And that's the sort of gap filling that this court has repeatedly rejected, most recently in the NIDAC case.

[00:05:57] Speaker 02:
You might want to slow down a little.

[00:05:59] Speaker 02:
I listen fast, but you talk even faster.

[00:06:02] Speaker 04:
That's rare.

[00:06:04] Speaker 04:
So I think what we have there is we really have a misplaced analysis.

[00:06:08] Speaker 04:
At best, we would have some sort of obviousness determination that could be made.

[00:06:12] Speaker 04:
But once again, the board did not institute an obviousness.

[00:06:15] Speaker 00:
What about claim one?

[00:06:17] Speaker 04:
As for claim one, the focus there is on the flattening limitation.

[00:06:24] Speaker 04:
And the flattening focuses on eliminating superfluous levels of hierarchy

[00:06:29] Speaker 04:
above said atomic blocks.

[00:06:31] Speaker 04:
Now to understand said atomic blocks, you have to look up to the previous claim limitation, which says that you select atomic blocks at a level of the hierarchy above the bottom of the branches.

[00:06:44] Speaker 04:
And that is really the problem with the board's analysis on claim one.

[00:06:51] Speaker 04:
The board just simply did not address whether fields or SU

[00:06:56] Speaker 04:
actually disclosed an elimination of a level of hierarchy above the atomic blocks with the atomic blocks being above the bottom level.

[00:07:07] Speaker 04:
You know, the obviousness analysis is quite strange in the first place because essentially the board took two different references.

[00:07:16] Speaker 04:
It said, well, these are basically in the same field, and then it just sort of mixed and matched.

[00:07:20] Speaker 02:
Okay, so imagine this is a sandwich, right?

[00:07:23] Speaker 02:
You can flatten your sandwich, you know, panini style, or you can flatten it by removing ingredients.

[00:07:28] Speaker 02:
So if I take all the tomatoes out of the sandwich, it's gonna flatten the sandwich, right?

[00:07:31] Speaker 02:
The sandwich is gonna appear flatter.

[00:07:33] Speaker 02:
It's not gonna stand up as tall, because all the tomato, the whole tomato layer is gone, right?

[00:07:37] Speaker 04:
That would be a form of flattening.

[00:07:38] Speaker 02:
Yeah, it feels like here the board said it's flattening because we added more tomatoes to the layer.

[00:07:44] Speaker 02:
I don't get it.

[00:07:45] Speaker 04:
I don't get it either.

[00:07:47] Speaker 04:
And I think it's beyond I don't get it, and therefore they need to explain.

[00:07:51] Speaker 04:
I think they're just wrong when you look at the reference, which is why I think reversal is appropriate on claim one in addition to claim 32.

[00:07:57] Speaker 02:
They didn't eliminate the layer.

[00:07:58] Speaker 02:
They actually broke it out and made it more.

[00:08:01] Speaker 02:
Right.

[00:08:01] Speaker 02:
They added more.

[00:08:04] Speaker 02:
Yes.

[00:08:04] Speaker 02:
They took R and made it R0, R1, R2.

[00:08:08] Speaker 02:
Expanded the layer.

[00:08:09] Speaker 02:
Yep.

[00:08:10] Speaker 02:
I just want to make sure I'm not missing the technology.

[00:08:12] Speaker 04:
Not at all.

[00:08:13] Speaker 04:
I mean, it's clear from the title of the reference, right?

[00:08:15] Speaker 04:
It's creating hierarchy in HDL-based HDL density design.

[00:08:21] Speaker 04:
So in any event, yes, you have a creation of hierarchy there.

[00:08:24] Speaker 04:
And then in the sue reference, which is the other one that the board relied on, there you're dealing with preserving hierarchy.

[00:08:31] Speaker 04:
And what the court didn't address as to sue at all is this notion of whether there was any elimination above the atomic block level, with the atomic block level being above the bottom level.

[00:08:42] Speaker 04:
So, once again, we just simply have references that don't teach or suggest or disclose or anything the claim limitations at issue here, in particular, the flattening limitation.

[00:08:55] Speaker 04:
Can I take you back to claim 32 for a minute?

[00:08:57] Speaker 03:
Sure.

[00:08:58] Speaker 03:
I mean, one of the difficulties, at least for me in these kind of cases, is

[00:09:03] Speaker 03:
In this area of technology, people sometimes use different words to describe the same concepts.

[00:09:08] Speaker 03:
And I am in no position to determine whether a word in one pat means the same or the other.

[00:09:14] Speaker 03:
And so here we have the expert saying that Sue, and we're relying on this,

[00:09:22] Speaker 03:
has things that correspond to hard blocks, and corresponds to optimal placement, and pointing to the right words for that.

[00:09:30] Speaker 03:
And do you disagree with that testimony, or do you just think even though with that testimony, if you take it as true, it's just not substantial estimate?

[00:09:40] Speaker 04:
Yeah, the testimony from the expert, I mean, it just essentially says... But that's what experts are supposed to do, right?

[00:09:48] Speaker 03:
I mean, these are complicated patterns, and they explain what these terms mean.

[00:09:54] Speaker 03:
And I'm not the fact finder.

[00:09:55] Speaker 03:
If the board says, well, that sounds good to me, that that's evidence, then we look at whether that's substantial evidence or not.

[00:10:02] Speaker 03:
And so it's kind of a two-step analysis.

[00:10:04] Speaker 03:
I mean, is his testimony so inherently wrong

[00:10:08] Speaker 03:
that we would find it not credible, and therefore the board here are relying on it, or even accepting it as true, it's still not substantial evidence.

[00:10:16] Speaker 03:
Right.

[00:10:17] Speaker 04:
I think it both steps, quite frankly.

[00:10:19] Speaker 04:
I think here it's so inherently incredible, because he does not provide any explanation for why he thinks that this disclosure ensue would amount to determining optimal placement.

[00:10:29] Speaker 03:
Well, he says it's performance driven.

[00:10:32] Speaker 03:
I mean, frankly, this claim limitation optimal placement seems a little bit of a stretch to be a patentable limitation anyway.

[00:10:40] Speaker 03:
It says, you know, you have this thing, and it determines the best fit to make it work well.

[00:10:45] Speaker 03:
Why isn't performance due to making it work well, too?

[00:10:49] Speaker 04:
Right.

[00:10:49] Speaker 04:
Well, first of all, there was no claim construction of determining optimal placement.

[00:10:53] Speaker 04:
And frankly, that is the petitioner's problem, because the petitioner is the one that was supposed to seek all claim construction that was necessary for its arguments.

[00:11:00] Speaker 04:
In fact, though, if you look at the specification, columns 10 and 11, there's a optimization algorithm for hard block placement.

[00:11:09] Speaker 04:
So when you actually dig in, had there been claim construction, it's clear that there's a disclosure of optimal, and it's a specific disclosure.

[00:11:17] Speaker 04:
And it's not simply use a commercial floor planner, which is simply a tool.

[00:11:22] Speaker 02:
And I think that Judge Hughes' question was a two-part question.

[00:11:26] Speaker 02:
You've done a very effective job of explaining the importance of the optimal limitation from a technological standpoint, which wasn't brief before us, but you can obviously have quite a good mastery of that technology.

[00:11:37] Speaker 02:
But I think his question more precisely

[00:11:40] Speaker 02:
was about the expert's mention of since Sue talks about performance driven.

[00:11:44] Speaker 02:
And I thought where you were going to go, which was the argument in your brief, was about how all the references of performance driven were directed at what?

[00:11:52] Speaker 02:
Were they directed at hard block placement?

[00:11:54] Speaker 02:
No, at soft macro placement.

[00:11:54] Speaker 02:
They were only soft block, because there are lots of ways to skin a cat, right?

[00:11:58] Speaker 02:
Right.

[00:11:58] Speaker 02:
And you can increase performance in a variety of ways, and that this reference had nothing to do with performance enhancement or performance driven regarding hard block placement.

[00:12:07] Speaker 04:
That's right.

[00:12:08] Speaker 04:
It takes it right back to the title.

[00:12:09] Speaker 02:
And so the fact that the word's performance-driven with regard to speed might be in a reference doesn't mean it's also going to be bigger, heavier, stronger.

[00:12:18] Speaker 02:
It just means it's faster.

[00:12:19] Speaker 02:
So you have to look at what those adjectives are attached to.

[00:12:22] Speaker 02:
That's what you argued in your brief.

[00:12:24] Speaker 02:
want to make sure you don't miss that it was a two-part question.

[00:12:27] Speaker 04:
Yes, absolutely.

[00:12:28] Speaker 04:
And the second part of the question is perfectly answered that way.

[00:12:31] Speaker 04:
I mean, it is a matter of what performance-driven is modifying.

[00:12:35] Speaker 04:
It's solely modifying the soft macros.

[00:12:38] Speaker 03:
Even though the expert testifies that Sue also talks about hard macros, which can be equivalent to hard blocks.

[00:12:44] Speaker 03:
Right.

[00:12:45] Speaker 04:
But hard macros, he never says it's performance-driven.

[00:12:51] Speaker 00:
Thank you.

[00:12:52] Speaker 00:
Mr. Severo, Mr. Marsh.

[00:13:01] Speaker 01:
Good morning to the police and court.

[00:13:03] Speaker 01:
Philip Marsh on behalf of the TopTech Inc.

[00:13:06] Speaker 01:
This case is a substantial evidence case.

[00:13:08] Speaker 01:
And the court should affirm the board's decision because it's supported by the substantial evidence.

[00:13:13] Speaker 02:
Why don't you start with flattened?

[00:13:15] Speaker 02:
Because I think that's probably your hardest, quite frankly.

[00:13:20] Speaker 02:
What the reference actually shows is the expansion of a layer and not the elimination of a layer.

[00:13:26] Speaker 02:
So how does that amount to flattening?

[00:13:28] Speaker 01:
I understand your point, Your Honor, and respectfully, I disagree.

[00:13:33] Speaker 01:
What's shown in the reference is in Figure 3, there's X0 and R0, right?

[00:13:38] Speaker 01:
And those eventually are grouped with, the reference tells us that they're regrouped and

[00:13:44] Speaker 01:
and divide it.

[00:13:45] Speaker 02:
Which page are you on just so we can get to the same place?

[00:13:48] Speaker 01:
Do you want to look at the reference or do you want to look at the board's opinion?

[00:13:51] Speaker 02:
Whatever you would like me to look at.

[00:13:53] Speaker 01:
Why don't we take a look at the reference?

[00:13:55] Speaker 01:
OK.

[00:13:55] Speaker 01:
So looking at appendix.

[00:13:58] Speaker 02:
Are you in sewer fields?

[00:14:00] Speaker 01:
In fields right now.

[00:14:01] Speaker 01:
Appendix 1013.

[00:14:02] Speaker 01:
1013.

[00:14:03] Speaker 02:
Yep.

[00:14:06] Speaker 01:
OK.

[00:14:07] Speaker 01:
So on the left-hand side in the second level of the hierarchy, there's R0.

[00:14:10] Speaker 01:
On the right-hand side, there's X0.

[00:14:13] Speaker 01:
The board credited Dr. Giasi's testimony below, the expert, who said that the regrouping in the way that they did eliminated a level of hierarchy.

[00:14:27] Speaker 01:
And that's really the part of the flattening limitation that Synopsis is challenging here.

[00:14:31] Speaker 01:
And if you look at Figure 4, you'll see that you don't see R0 or X0.

[00:14:35] Speaker 01:
Now, their complaint is that no, instead what you've done is you've replaced it with several different blocks.

[00:14:41] Speaker 01:
And I understand that that's your point.

[00:14:43] Speaker 01:
But respectfully, each of the sections that's shown here, R1, R2, R3, and R4, which were part of R0, those were part of R0 to begin with.

[00:14:53] Speaker 01:
We didn't create new hierarchy.

[00:14:55] Speaker 01:
Those were within that block labeled R0 before.

[00:14:58] Speaker 01:
They weren't called out in the hierarchical description.

[00:15:00] Speaker 02:
Well, in fact, nothing about R0 is disclosed to have been eliminated, right?

[00:15:04] Speaker 02:
Nothing was eliminated.

[00:15:04] Speaker 02:
It's like you took a piece of meat and you cut it into four chunks, but it's still the same quantity of meat.

[00:15:10] Speaker 02:
There's nothing in this reference that suggests

[00:15:12] Speaker 02:
than any portion of the meat was removed.

[00:15:14] Speaker 02:
Correct?

[00:15:15] Speaker 01:
Again, respectfully I disagree.

[00:15:16] Speaker 01:
What portion of the record?

[00:15:18] Speaker 01:
So what the expert said and what the board agreed with is that the superfluous level of the hierarchy, the R0 grouping of R1 through R4.

[00:15:26] Speaker 02:
That's not true.

[00:15:27] Speaker 02:
Let's turn to your brief at page 21 because you know I actually don't do this very often but I think in this case I need to point it out to you.

[00:15:38] Speaker 02:
At page 21 in your brief

[00:15:41] Speaker 02:
You have a sentence, it's about two-thirds of the way down, where you say, the board disagreed, noting that in fields, modules R0 and X0 were removed, while in SU, node SM1 and 2 was eliminated.

[00:15:58] Speaker 02:
And then you cite Appendix 21.

[00:16:00] Speaker 02:
If you go to Appendix 21, the board didn't actually ever make that holding.

[00:16:06] Speaker 02:
Instead, they said, you argue that.

[00:16:09] Speaker 02:
Zero and x-ray removed and ensue s1 and s2 were eliminated then the board goes on and says In fields are zero next year removed, but they never make the additional holding that you are attributing to them So why don't you take a look and tell me what I'm missing right?

[00:16:27] Speaker 01:
So let's look at appendix 21 the board's decision what the board said is Where did they say sue?

[00:16:35] Speaker 02:
Nodes SM1 and 2 were eliminated.

[00:16:37] Speaker 02:
Where does the board say that?

[00:16:39] Speaker 01:
Right.

[00:16:39] Speaker 01:
So let me start.

[00:16:42] Speaker 01:
Because of the way that the board phrased it, I need to start with fields and then get to sue.

[00:16:46] Speaker 02:
Do you see on page 21 where you say, in reply, petitioner asserts in fields, modules R0 and X0 removed, and soon node SM1 and SM2 were eliminated.

[00:16:56] Speaker 02:
Do you see where the board says that?

[00:16:58] Speaker 02:
You assert that.

[00:16:59] Speaker 02:
Petitioner asserts.

[00:17:00] Speaker 02:
Do you see that language in the board opinion?

[00:17:02] Speaker 02:
Yes, I see that.

[00:17:03] Speaker 02:
And so then down at the bottom, the board does seem to say R0 and X0 have been divided and renamed.

[00:17:10] Speaker 02:
You see that?

[00:17:11] Speaker 01:
Correct.

[00:17:12] Speaker 02:
And then on the next place, nowhere do they say ever that SM1 and SM2 were eliminated and sued.

[00:17:19] Speaker 02:
They never say that anywhere.

[00:17:22] Speaker 02:
And you have represented to us in the red brief that the board made that holding and noted it expressly.

[00:17:28] Speaker 02:
I think that you've made a little bit of a mistake.

[00:17:30] Speaker 02:
I think what you really have done was taken their sentence that you asserted something and attributed it to them as an actual holding.

[00:17:37] Speaker 02:
And I'm not sure I can find that.

[00:17:41] Speaker 01:
Perhaps that wasn't worded as artfully as it could have been, Your Honor.

[00:17:44] Speaker 01:
I think what happened here is, as you point out, at the bottom of page 21, the board adopted Dr. Gyasi's testimony and said, as testified to by Dr. Gyasi in fields,

[00:17:56] Speaker 01:
basically R0 and X0 are eliminated from Figure 4, just like I said.

[00:18:01] Speaker 01:
And then as you continue on, the board talks about how, and we should have probably had Appendix 22 cited there, where it says in Figure 2B of SU the two hard macros, HM1 and HM2, are first shown at different levels and then are depicted after soft macro formation in the same layout plane where each hard macro is assigned

[00:18:23] Speaker 01:
into its corresponding region.

[00:18:25] Speaker 02:
But that has nothing to do with SM-1 or SM-2 being eliminated.

[00:18:28] Speaker 02:
HM-1 and HM-2 are hard macros.

[00:18:31] Speaker 02:
SM-1 and SM-2 are soft macros.

[00:18:33] Speaker 02:
I don't see how I could interpret that statement by the board the way that you represent it in the red brief, that it affects the elimination of a whole layer of soft macros.

[00:18:43] Speaker 02:
The board did not hold that.

[00:18:47] Speaker 01:
Perhaps that's a fair point, Your Honor.

[00:18:49] Speaker 01:
It is true that Dr. Gyasi testified.

[00:18:52] Speaker 01:
I'll let you move on.

[00:18:53] Speaker 01:
Thank you.

[00:18:53] Speaker 01:
It is true that Dr. Giasi testified that that node, SM1,2, that that was removed from figure 2B of Sue in the placement.

[00:19:04] Speaker 01:
But he also testified that the node directly above HM2, that that node became a superfluous level of hierarchy that was eliminated as well.

[00:19:15] Speaker 01:
That node's not labeled.

[00:19:17] Speaker 01:
And so it's kind of hard to address that note directly.

[00:19:19] Speaker 02:
But the board.

[00:19:20] Speaker 02:
It's just kind of crazy to me to say you eliminated something when all you did was parse it up into pieces.

[00:19:27] Speaker 02:
You know?

[00:19:28] Speaker 02:
You didn't eliminate.

[00:19:28] Speaker 02:
And you have to flatten a layer.

[00:19:30] Speaker 02:
You have to flatten a whole layer.

[00:19:32] Speaker 02:
You know?

[00:19:32] Speaker 02:
And you have the same number of layers in Sioux and in Fields.

[00:19:36] Speaker 02:
You just have them dispersed.

[00:19:39] Speaker 01:
Well, I think maybe it's important to understand what layers we're talking about here.

[00:19:43] Speaker 01:
So we're talking about hierarchical representation of a schematic.

[00:19:47] Speaker 01:
And so in that sense, the way that it's described is by grouping logical ideas together.

[00:19:52] Speaker 01:
And R1 through R4 were grouped together in R0.

[00:19:58] Speaker 01:
X1 and X2 were grouped together in X0.

[00:20:01] Speaker 01:
That superfluous level of hierarchy, R0 and X0, were removed, and they disappeared, the same ensue

[00:20:09] Speaker 01:
SM 1, 2.

[00:20:11] Speaker 03:
I don't understand how you can say it disappeared when they're still the same level all the way across.

[00:20:16] Speaker 03:
It's just more at the level, not less at the level.

[00:20:20] Speaker 01:
Well, Your Honor.

[00:20:21] Speaker 02:
I assure you, I have four children.

[00:20:23] Speaker 02:
If they stand in a group, there are four children.

[00:20:25] Speaker 02:
If I stand them up individually, I still have four children.

[00:20:29] Speaker 01:
That's true.

[00:20:29] Speaker 01:
But if you were drawing up a hierarchical representation and you didn't group them, let's say, as boys and girls,

[00:20:36] Speaker 01:
And you didn't group them that way in the hierarchical representation, but you had a flatter representation that just said, you know, Joey, Annie, and whoever they were.

[00:20:44] Speaker 01:
That's what's shown in Figure 4.

[00:20:46] Speaker 01:
That's a flatter representation.

[00:20:47] Speaker 01:
You've removed some hierarchical grouping that is R0 and X0.

[00:20:52] Speaker 01:
Those are categories that group together those R1 through R4 and X1 and X2.

[00:21:00] Speaker 01:
When you remove that, that's what the patent is talking about, removing

[00:21:03] Speaker 01:
the superfluous level of hierarchy above that flattened tier.

[00:21:07] Speaker 03:
But in the patent, it removes it.

[00:21:09] Speaker 03:
It doesn't replace it with anything.

[00:21:12] Speaker 03:
Your reference removes it, but then replaces it with things.

[00:21:17] Speaker 03:
So why isn't that level of hierarchy still there?

[00:21:20] Speaker 01:
Great question.

[00:21:21] Speaker 01:
So the patent does replace it eventually with what's called PRUs.

[00:21:25] Speaker 01:
But this is a claim that has a comprising transition language.

[00:21:31] Speaker 01:
and it's under the broadest reasonable interpretation.

[00:21:33] Speaker 01:
And so the fact that there might be another layer added or inserted, I think, doesn't mean that we don't remove the X0 and R0.

[00:21:42] Speaker 01:
Those groupings, that level of hierarchy, it appears in figure three, but doesn't appear in figure four.

[00:21:48] Speaker 01:
It's gone.

[00:21:49] Speaker 01:
It's gone from the logical hierarchy.

[00:21:51] Speaker 01:
That's what Dr. Gyasi testified to, and that's what the board agreed to.

[00:21:55] Speaker 01:
Now, it may be true that we could

[00:21:59] Speaker 01:
that reasonable minds could differ on this, that we could come to different reasonable conclusions.

[00:22:04] Speaker 01:
But under the substantial evidence standard, what we understand from this court is that as long as there's more than a scintilla of evidence that would support the board's decision, and here they cite Dr. Chiasi and they cite the references themselves, as long as there's that kind of evidence, we have to affirm the board, even if we disagree, even if there's another reasonable interpretation.

[00:22:26] Speaker 01:
So the board walked through on claim one.

[00:22:29] Speaker 03:
But the expert's testimony has to be a reasonable description of the prior art.

[00:22:40] Speaker 03:
I mean, he can say this discloses the things, but if it doesn't actually disclose removing a level of hierarchy.

[00:22:50] Speaker 03:
And again, I have the same problem.

[00:22:52] Speaker 03:
here that I had with your friend is this is very complicated technology and we're relying largely on experts and the board to tell us what these references say but it certainly doesn't seem like what he's saying about this reference we're talking about getting rid of RO matches up with pictures but I'm a little worried that I'm relying too much on my

[00:23:19] Speaker 03:
you know, kind of simplistic view of the picture.

[00:23:22] Speaker 03:
There's still three levels, so why not flattening when he said something else?

[00:23:27] Speaker 03:
So that's precisely why I get back to the standard, Your Honor.

[00:23:30] Speaker 03:
Here, in this case... But you still have to give me some reason to believe that he's correct.

[00:23:35] Speaker 03:
And I still don't, just intuitively, and I apologize if I'm just flat out wrong about this, but intuitively, it does not seem to me when you take away R0,

[00:23:46] Speaker 03:
But at the same level in the drawing, replace it with R1, R2, R3, that you've taken away any level of hierarchy.

[00:23:54] Speaker 03:
You've just reformatted that hierarchy.

[00:23:59] Speaker 01:
So I might take issue with that a little bit, and reasonable minds, I think, could differ.

[00:24:03] Speaker 03:
But can you point to me somewhere in his testimony that explains why that's not a correct articulation of what's happening here?

[00:24:12] Speaker 01:
Yes, I can.

[00:24:14] Speaker 01:
In Appendix 1099, Dr. Gyasi testified looking at fields that R0 and X0 are eliminated from Figure 4.

[00:24:25] Speaker 01:
That was his understanding of the reference.

[00:24:27] Speaker 01:
OK, but that's not good enough, right?

[00:24:29] Speaker 01:
That's the problem.

[00:24:30] Speaker 01:
I'm just getting started.

[00:24:31] Speaker 01:
OK.

[00:24:32] Speaker 01:
So he then said that by selecting atomic blocks for regrouping, X0 and R0 are rendered a superfluous level of hierarchy above the atomic blocks

[00:24:42] Speaker 01:
that are eliminated and flattened in the process.

[00:24:46] Speaker 01:
That's his interpretation of this technical reference.

[00:24:49] Speaker 01:
He then said that based on that, one of ordinary skill in the art would have understood that in selecting atomic blocks and then regrouping those atomic blocks, the plurality of hierarchically arranged branches above the atomic blocks would be flattened and eliminated.

[00:25:04] Speaker 03:
Now, those are- But he doesn't address the addition of a new level, the R123 level.

[00:25:12] Speaker 03:
and why that doesn't destroy the flattening effect.

[00:25:18] Speaker 01:
He doesn't expressly say that.

[00:25:20] Speaker 01:
He didn't need to do that to meet the claims.

[00:25:22] Speaker 01:
I mean, we're talking again about a broadest reasonable interpretation here.

[00:25:26] Speaker 03:
OK, let me just lay it on the line, because that testimony is not getting me there.

[00:25:32] Speaker 03:
I don't see, if you have three hierarchies, you have an A level, a B level, and a C level.

[00:25:39] Speaker 03:
And you get rid, and there's one block in each.

[00:25:42] Speaker 03:
Or maybe there's more in the lower blocks.

[00:25:44] Speaker 03:
It really doesn't matter We're talking about the top level and you take away the a block and replace it with a 1 a 2 Don't you still have three levels of hierarchy?

[00:25:54] Speaker 03:
So I think the may have eliminated one level of hierarchy in a certain format You've replaced it at the same level with more blocks I think the problem here is that in r0 and x0 is

[00:26:08] Speaker 01:
R1 through R4 is there.

[00:26:10] Speaker 01:
Everything in R1 through R4 is within R0.

[00:26:13] Speaker 01:
But that's the problem with your argument.

[00:26:15] Speaker 03:
If everything in R0 includes R1 through R4, and you're just expanding it, you're not taking away any level.

[00:26:25] Speaker 03:
You're just breaking it up.

[00:26:27] Speaker 03:
It's still there.

[00:26:29] Speaker 03:
But nothing was added.

[00:26:30] Speaker 03:
I mean, my point, if we... Can I just turn you back to 32 before you completely run out of time?

[00:26:37] Speaker 03:
I find that one even more baffling, because I don't understand his testimony.

[00:26:40] Speaker 03:
I don't understand the Sioux reference, what he said just before.

[00:26:44] Speaker 03:
Your friend said about the performance driven in the reference only relating to the soft modules.

[00:26:50] Speaker 03:
If that's true, then that seems to undermine the expert's testimony.

[00:26:55] Speaker 03:
Can you address that?

[00:26:57] Speaker 03:
I'm sorry, if what's true?

[00:26:58] Speaker 03:
That Dr. Ghiasi's testimony about Sioux

[00:27:02] Speaker 03:
and pointing to performance-driven as meaning the optimal thing, that the performance-driven was all about the soft macros and not the hard macros, which corresponded to hard modules.

[00:27:13] Speaker 01:
So that's not what he said.

[00:27:14] Speaker 01:
And that's not what Sue's about exclusively.

[00:27:19] Speaker 01:
Sue is titled, it's a performance-driven soft macro clustering and placement, but it talks about

[00:27:25] Speaker 01:
using floor planning to place the hard macros as well.

[00:27:28] Speaker 03:
And what Dr. Ghiasi tests... So where does it talk about placement of the hard macros being performance-driven as opposed to the soft macros?

[00:27:36] Speaker 01:
Well, so I think there we have to look at anticipation through the lens of what those skilled in the art would understand when they read the term.

[00:27:43] Speaker 01:
And so that's where Dr. Ghiasi comes in.

[00:27:46] Speaker 01:
That's what the case law tells us.

[00:27:47] Speaker 01:
Dr. Ghiasi comes in and says, OK, Sue teaches that each hard macro is assigned into its corresponding region according to a floor planning result.

[00:27:55] Speaker 01:
That's true.

[00:27:57] Speaker 01:
He says, Sue shows placement in both figures 3A and 2B, and Sue's hard macros are the patent's hard blocks.

[00:28:05] Speaker 01:
Then he says that floor planning is part of a circuit design, and that floor planning process does optimize.

[00:28:12] Speaker 01:
That's what his testimony is as an expert in the field.

[00:28:14] Speaker 02:
But wouldn't that be the inherent part?

[00:28:17] Speaker 02:
Or, I mean, don't, see, this is the bit of the problem with anticipation versus obviousness.

[00:28:22] Speaker 02:
I'm not saying we can't make an obviousness case here, but my problem is

[00:28:25] Speaker 02:
anticipation you have to understand that floor planning and that all commercial floor planning mechanisms because no particular one is specified in sue so it could be any off-the-shelf commercial floor planning mechanism okay so you have to understand that those would necessarily optimize hard macro placement

[00:28:48] Speaker 02:
That's where I feel like you've moved either to obviousness or at least to maybe inherency if your expert can establish that they would always, every time, do that.

[00:28:57] Speaker 02:
But you don't have an explicit disclosure.

[00:28:59] Speaker 02:
That's where I'm struggling.

[00:29:01] Speaker 01:
Well, to the point about every tool, Your Honor, it does call out specifically Cadence's ensemble program for floor planning.

[00:29:10] Speaker 01:
And I'll note, and this is kind of an important point, the problem that we run into here is it's a little difficult to know what optimal is.

[00:29:18] Speaker 01:
I mean, my colleague here told you that optimal was described in an algorithm in columns 10 and 11 of the patent.

[00:29:25] Speaker 01:
That's not quite true.

[00:29:26] Speaker 01:
The hard block placement doesn't really have that algorithm.

[00:29:30] Speaker 01:
That algorithm is about placing the soft blocks around the hard blocks.

[00:29:34] Speaker 01:
What it says about hard block placement is that good hard block placement is necessary for the placement optimization step, which is step 22 in figure 1.

[00:29:41] Speaker 01:
That happens later.

[00:29:43] Speaker 01:
Now, if you look in the patent and if you look to column

[00:29:47] Speaker 01:
Get this right.

[00:29:50] Speaker 01:
If you look to column 16, at lines 14 through 16, what you'll see is when they talk about that step 22, which again is described as the placement and routing optimization step in figure one, it says when you do that step, that you use software tools such as Silicon Ensemble or PKS.

[00:30:12] Speaker 01:
It's the same exact disclosure that we have

[00:30:15] Speaker 01:
in the reference.

[00:30:16] Speaker 01:
So to the extent that they have support for this determining optimal placement in the patent, which I agree is maybe a questionable limitation, we have the exact same disclosure.

[00:30:26] Speaker 00:
We disclose that exact same program.

[00:30:37] Speaker 00:
Take four minutes if you need it.

[00:30:39] Speaker 04:
Thank you, Your Honor.

[00:30:40] Speaker 04:
I'll be very brief.

[00:30:43] Speaker 04:
I think Mr. Marshall's main point was, you know, the standard review, substantial evidence, you know, can reasonable minds differ.

[00:30:49] Speaker 04:
And I think the issue here is that reasonable minds really can't differ.

[00:30:53] Speaker 04:
We have references that don't contain the limitations that are at issue here.

[00:30:59] Speaker 04:
For that reason,

[00:31:00] Speaker 04:
I think the court should reverse the board's determination.

[00:31:05] Speaker 04:
I'll just briefly touch on one other thing, and that's this notion of optimal placement in claim 32 and what's disclosed.

[00:31:14] Speaker 04:
Once again, it all comes back to what Sue is pointing at, and that's soft macro placement.

[00:31:19] Speaker 04:
And that's all that's performance driven.

[00:31:21] Speaker 04:
For that reason, we don't have the same disclosure in any way.

[00:31:25] Speaker 04:
Unless the court has any questions, I know we've already gone over as a whole here.

[00:31:30] Speaker 00:
Thank you.

[00:31:31] Speaker 00:
Council will take the case and revise it.

[00:31:33] Speaker 04:
Thank you.