[00:00:00] Speaker 04:
Please be seated, ladies and gentlemen.

[00:00:09] Speaker 04:
We have five cases in the calendar this morning.

[00:00:15] Speaker 04:
Three patent cases from district court and one from the patent office.

[00:00:21] Speaker 04:
And two government employee cases, one of which is being submitted on the brief and will not be argued.

[00:00:28] Speaker 04:
The first case is Upte Incorporated versus Via Technologies, 2013, 1670, 2014, 1839.

[00:00:39] Speaker 04:
Mr. Carpenter, good morning.

[00:00:44] Speaker 04:
Please proceed.

[00:00:52] Speaker 00:
May it please the court.

[00:00:55] Speaker 00:
I'm Brian Carpenter, and along with my co-counsel, Tim Craddock, represent VIA Technologies, the appellants and cross-appellees.

[00:01:04] Speaker 00:
VIA requests that this court reverse and render judgment of non-infringement for two reasons.

[00:01:10] Speaker 00:
First, the determining function is not met because the accused VIA chipsets cannot determine whether data is cached in a modified state in the L1 cache.

[00:01:21] Speaker 00:
Second, Opti failed to lay a legally sufficient record upon which any reasonable juror could find that structures in the accused via chipsets were equivalent to the corresponding structures for the two means plus function, means plus function limitations.

[00:01:38] Speaker 00:
Alternatively, if in the event that this court finds that it cannot enter judgment, we at least request a new trial due to the erroneous jury instruction on the 112-6 equivalence issue.

[00:01:51] Speaker 00:
We also request the judgment that sequentially transferring means is indefinite.

[00:01:58] Speaker 00:
Since indefiniteness is clear, there's no reason to reach enablement.

[00:02:01] Speaker 00:
But as briefed, claim 26 is also invalid for lack of enablement, as explained in briefs.

[00:02:09] Speaker 00:
Regarding the determining function, there is no dispute about the one crucial fact that was admitted by Dr. Smith and confirmed by V as expert.

[00:02:21] Speaker 00:
And that is, the accused via chipsets cannot tell whether modified data is cached in the L1 cache or the L2 cache.

[00:02:31] Speaker 00:
As stated by Dr. Smith, Opti's expert,

[00:02:34] Speaker 00:
The accused via chipset simply do not know which one or both have modified data.

[00:02:42] Speaker 02:
What is the point here if I understand correctly at the time the cache memory could be a read back memory or a read through memory and the patent is dealing with cache memory which is read back memory.

[00:03:01] Speaker 02:
Is my understanding about that correct?

[00:03:03] Speaker 00:
I believe, in one small aspect, you're mistaken, Your Honor.

[00:03:06] Speaker 00:
And that is that it dealt with right back cache, not right through cache.

[00:03:13] Speaker 00:
The invention dealt with right back cache.

[00:03:16] Speaker 02:
So I said read back.

[00:03:18] Speaker 02:
I should have said right back.

[00:03:19] Speaker 02:
So the reason it's not concerned with the L2 cache at the time was that that was part of a secondary memory, right?

[00:03:32] Speaker 00:
That is correct.

[00:03:33] Speaker 00:
Figure one of the patents shows it as part of the secondary memory.

[00:03:36] Speaker 02:
So if I understand the accused infringing product correctly, both the L1 and the L2 are right back cash memories, right?

[00:03:54] Speaker 00:
To put it yes, but to put it a little differently, in the modern devices,

[00:03:59] Speaker 00:
Right-back is done with respect to all caches no matter how many there are.

[00:04:03] Speaker 00:
Right.

[00:04:03] Speaker 02:
So in terms of looking at it as a Doctrine of Equivalence issue, wouldn't you want this to apply to all right-back caches and not just the L1?

[00:04:18] Speaker 00:
Well, first of all, Doctrine of Equivalence was not argued on this function.

[00:04:22] Speaker 00:
There was no alleged equivalent function to the one with the claim.

[00:04:27] Speaker 00:
I'm not sure I understood your question.

[00:04:30] Speaker 02:
Well, the identification in this claim of the first cash memory is designed to distinguish it from other cash memories which are part of the secondary memory, correct?

[00:04:51] Speaker 02:
So we have a problem of trying to figure out what this claim means where you have cache memories where more than one of them is a write-back memory.

[00:05:03] Speaker 02:
Is that fair or unfair?

[00:05:05] Speaker 00:
I believe that's a little unfair, because at the time of this invention, there were CPUs that had an L1 on board, but it was a write-through cache.

[00:05:15] Speaker 00:
And then the L2 would be in the secondary memory, and it would be write-back.

[00:05:20] Speaker 00:
We had a piece of prior art.

[00:05:21] Speaker 00:
In fact, we had a very, very early circuit board that operated that way.

[00:05:26] Speaker 00:
That is why via push for a construction of first cash memory, excuse me, as push for a construction of said first cash memory as the first memory, the first cash memory

[00:05:43] Speaker 00:
that was a write back cache and Opti resisted that and pushed forward.

[00:05:47] Speaker 02:
But it didn't contemplate, at the time of the patent, it didn't contemplate that the other cache memories would also be write back cache memories.

[00:05:55] Speaker 00:
Oh, to the contrary.

[00:05:59] Speaker 00:
There's a whole paragraph in the patent that talks about other levels of cache.

[00:06:04] Speaker 00:
But the time from this patent, early 1994, Intel had announced the pinion processor.

[00:06:10] Speaker 02:
The 486 process.

[00:06:12] Speaker 02:
Help me, because I'm not understanding here.

[00:06:15] Speaker 02:
At the time of the patent, they focused on the L1 cache, because if I understand correctly, that was the only cache memory that was a write-back memory.

[00:06:27] Speaker 02:
Other cache memories were write-through memories.

[00:06:29] Speaker 00:
No, no.

[00:06:30] Speaker 00:
Excuse me, Your Honor.

[00:06:31] Speaker 00:
I don't believe that's correct.

[00:06:35] Speaker 00:
The decision or the difference in various designs was whether or not L1 would be right through or right back.

[00:06:42] Speaker 00:
As to all the others, right back was a well-known established technique.

[00:06:47] Speaker 00:
So I think it's the reverse of the way, Your Honor, is thinking of it.

[00:06:51] Speaker 00:
But I thought you said the L2 was a right through memory.

[00:06:56] Speaker 00:
No, Your Honor.

[00:06:57] Speaker 00:
If I said that, I was mistaken.

[00:06:59] Speaker 00:
The L2 was always right back.

[00:07:02] Speaker 00:
The piece of prior art, Via,

[00:07:04] Speaker 00:
would have used had it had came to construction it pushed for was a design where the L2 was right back, the L1 was right through.

[00:07:12] Speaker 00:
And because of the construction Opti pushed for, we had to rely only on the

[00:07:22] Speaker 02:
the designs where... But I guess what I'm trying to say, and I don't want to use up all your time with this, is that the claim 26 is concerned with right-back memory, which is where the problem arises.

[00:07:38] Speaker 02:
And why shouldn't, I guess what I'm saying, why shouldn't claim 26 be construed to cover all right-back memories, whether they're the first memory or the second memory?

[00:07:50] Speaker 00:
Well, because, Your Honor, it was not construed that way.

[00:07:53] Speaker 00:
And if it is to be construed that way, I would contend we should have another shot at anticipation of a new trial, because it was not construed that way.

[00:08:02] Speaker 00:
The way it was construed, and it's construction I'll be pushed for, was a particular cache.

[00:08:07] Speaker 00:
And if I may make one more point.

[00:08:11] Speaker 00:
As you've noted, figure one of the patents shows L2 as far as secondary memory, and there's

[00:08:20] Speaker 00:
Even in the claim itself, the preamble distinguished between secondary memory and a first cash memory.

[00:08:27] Speaker 00:
There are two different items in the preamble.

[00:08:31] Speaker 00:
So that's one more reason we believe that the L2 is completely separate from first cash memory.

[00:08:42] Speaker 00:
Do you want to move on to your other points?

[00:08:45] Speaker 00:
Yes, Your Honor.

[00:08:47] Speaker 00:
In terms of the means elements,

[00:08:50] Speaker 00:
That trial opting grossly oversimplified the infringement analysis at the expense of developing a legally viable record that any reasonable juror could find infringement on.

[00:09:01] Speaker 00:
As demonstrated in VIA's briefing for each of the means plus function elements, Dr. Smith oversimplified the functions and identified an oversimplified way and result.

[00:09:11] Speaker 00:
He then took the oversimplified function way result and looked for each of these in the accused devices.

[00:09:18] Speaker 00:
There is no precedent, nor has OPTI cited and demonstrated that the way and the result that could be used like claim limitations that you look for in the accused devices.

[00:09:27] Speaker 00:
The way result analysis requires identification and discussion of structural differences.

[00:09:34] Speaker 00:
In its response to be a second stated issue, that be as intelligent judgment as a matter of law, OPTI continued to argue about jury instructions.

[00:09:41] Speaker 00:
It is true that VIA led its opening brief discussing jury instructions.

[00:09:45] Speaker 00:
But it did so because that facilitated an early discussion in the briefing about the law on insubstantial differences.

[00:09:52] Speaker 00:
It also highlighted an important difference between the party's understanding of the law.

[00:09:58] Speaker 00:
He has demonstrated by its review and discussion of this court's precedence that the test for section 112 equivalence is insubstantial differences.

[00:10:07] Speaker 00:
Under this correct statement of the law, the court's de novo review is made much easier.

[00:10:11] Speaker 00:
as Opti's expert did not present any testimony regarding differences or insubstantial differences.

[00:10:16] Speaker 00:
But as to VIA's primary position that judgment should be entered as a matter of law, the jury instructions are irrelevant.

[00:10:23] Speaker 00:
The issue is that infringement should have never been allowed to go to the jury.

[00:10:28] Speaker 00:
Under either statement of the law reflected in the two jury instructions, Opti failed to conduct a proper way result analysis.

[00:10:41] Speaker 00:
Opti claims that VIA is trying to draft a new requirement, identification differences in the test, and that that's unsupported.

[00:10:48] Speaker 00:
Well, to the contrary.

[00:10:50] Speaker 00:
VIA's extensive analysis of the opinions of this court is that a proper way and result analysis begins with an identification differences between two sets of structures.

[00:11:02] Speaker 00:
The cases relied upon by Opti demonstrate that analysis of differences is required.

[00:11:07] Speaker 00:
For example, Opti relied on odetics, as does VIA.

[00:11:10] Speaker 00:
The court noted that this court in Cheminata, after construing the claim, identifying the structures, proceeded to analyze the differences between skid plate, the corresponding structure, and the assertively equivalent structure in the accused device, a set of soft work of wheels.

[00:11:28] Speaker 00:
To analyze differences, one must first identify the differences.

[00:11:33] Speaker 00:
The cases of Cheminata, Alpex, Solomon, discussed at length in his briefing provide additional good examples

[00:11:40] Speaker 00:
of a proper way-result analysis.

[00:11:43] Speaker 00:
All of these cases demonstrate that structural differences must be identified.

[00:11:49] Speaker 00:
Opti accuses VIA of insisting on component-by-component analysis, as shown in VIA's discussion of odetics, Caterpillar, Toro, and Solomon.

[00:11:56] Speaker 00:
The phrase component-by-component refers to the impermissible practice of further subdividing the corresponding structures into additional parts, contrary to what Opti would have his court believe.

[00:12:07] Speaker 00:
VIA did not insist on one-to-one correspondence, nor did it subdivide the corresponding structures.

[00:12:16] Speaker 00:
Dr. Smith simply failed to acknowledge any differences, much less make a comparison as between corresponding and accused structures.

[00:12:25] Speaker 00:
Requiring OPTI to account for the corresponding structures and equivalence thereof is necessary to prevent purely functional claiming.

[00:12:32] Speaker 00:
In Valmont, this court explained that Congress in 1952

[00:12:36] Speaker 00:
Patent Act added 112.6 to, in effect, overrule the Supreme Court's decision in Halliburton 329 U.S.

[00:12:42] Speaker 00:
1 at 9.

[00:12:44] Speaker 00:
The Supreme Court had prohibited means plus function language, fearing it would be over-broad and ambiguous.

[00:12:50] Speaker 00:
Congress decided to permit broad means plus function language, but added a standard.

[00:12:55] Speaker 00:
That standard is the second clause of paragraph 6.

[00:12:59] Speaker 04:
Mr. Carpenter, you're into your rebuttal time.

[00:13:01] Speaker 04:
You can continue while I say that as you wish.

[00:13:04] Speaker 00:
I'll finish this one quick point, and then I'll reserve the rest.

[00:13:08] Speaker 00:
The point of the second clause of section 112-6 is to avoid purely functional claiming.

[00:13:13] Speaker 00:
This stated in aristocrat 521, F-30, 1328.

[00:13:17] Speaker 00:
Allowing Opti and Dr. Smith to ignore the structures in the court's construction would be to allow purely functional claiming.

[00:13:25] Speaker 00:
That is, a black box for performing a function.

[00:13:28] Speaker 00:
In Alpex, this court granted judgment where the expert concluded

[00:13:32] Speaker 00:
a purely functional and not structural analysis, be it a request that this court do the same in interjudgment of non-intentional.

[00:13:41] Speaker 04:
Mr. Grest.

[00:13:42] Speaker 04:
Thank you, Your Honor.

[00:13:54] Speaker 01:
My name is Michael Brody.

[00:13:56] Speaker 01:
I'm here today with Jeff Eaton on behalf of Opti, Inc.

[00:13:59] Speaker 01:
Mr. Michael Mazzoni of Opti is present today as well.

[00:14:03] Speaker 01:
I thought I'd try to be very quick with respect to the issues that Mr. Carpenter addressed.

[00:14:09] Speaker 02:
I have some concern about this first cache memory point.

[00:14:13] Speaker 02:
I'm not at all sure if the district court was right in saying that it doesn't make any difference that the infringing device retrieves the L1 out of two cache memories.

[00:14:28] Speaker 02:
And I was trying to understand what was going on at the time, the patent issue.

[00:14:35] Speaker 02:
And if I understand correctly from the patent, the L1 memory was a right back memory, but the L2 memory at that time was part of the secondary memory.

[00:14:47] Speaker 02:
So the thing would only operate on the L1, but now the

[00:14:53] Speaker 02:
accused infringing device has more than one right bash memory.

[00:15:01] Speaker 02:
So you would want the patent to operate on both of those memories.

[00:15:06] Speaker 02:
Does that make sense to you?

[00:15:09] Speaker 01:
That was my understanding.

[00:15:11] Speaker 01:
I mean, I think there's another way of dealing with that issue, which is simply that in the Viya chip, in the Viya systems,

[00:15:23] Speaker 01:
Let me back up.

[00:15:24] Speaker 01:
What the claim requires is determining whether the L1 has data cache in a modified state.

[00:15:31] Speaker 01:
And nobody disputes that the bias systems determine that because they have to.

[00:15:37] Speaker 01:
If they didn't, the system would crash or you'd get even worse.

[00:15:42] Speaker 01:
It wouldn't crash, but you'd get faulty

[00:15:45] Speaker 01:
data being used for the various applications.

[00:15:48] Speaker 01:
But it operates on both the L1 and the L2, right?

[00:15:53] Speaker 01:
Yes, sir.

[00:15:53] Speaker 01:
And that's the way it determines if L1 is in a modified state.

[00:15:59] Speaker 02:
Well, it doesn't tell you whether it's L1 or L2, right?

[00:16:03] Speaker 01:
No, but the inverse, I think it is, is true.

[00:16:09] Speaker 01:
So if the L1 is in a modified state, that is determined by

[00:16:15] Speaker 01:
the chipset.

[00:16:16] Speaker 02:
Yeah, but I'm not sure that the district court was correct in saying it doesn't make any difference that it's also operating on the L2 as well as the L1.

[00:16:25] Speaker 01:
Well, I appreciate that.

[00:16:29] Speaker 01:
And I think our position would be, first of all, the distinction that your honor was drawing when Mr. Carpenter was up as an appropriate one.

[00:16:40] Speaker 01:
The difference between operating on the one or the other is in itself an insubstantial difference.

[00:16:47] Speaker 01:
And third, what?

[00:16:47] Speaker 02:
Was the doctrine of quillin's argued with respect to this aspect of the claim?

[00:16:51] Speaker 01:
No, it wasn't.

[00:16:53] Speaker 01:
But third, what the claim calls for is determining whether the L1 is cached in a modified state.

[00:17:03] Speaker 01:
And that is what the chipset does.

[00:17:06] Speaker 01:
The way the VIA chipset is designed,

[00:17:11] Speaker 01:
that can be accomplished by checking the L2, by basically checking both of them, or the L2 cache.

[00:17:17] Speaker 01:
It doesn't matter.

[00:17:19] Speaker 01:
But the fact that the Viachip set does more than what the claim is required doesn't mean that it fails to do what the claim is required.

[00:17:28] Speaker 01:
And that, I think, is the position that was basically argued below.

[00:17:33] Speaker 01:
and uh... and i think that's the simplest answer to this whole question but we're doing exactly what the paper could they're doing exactly well that that that one can only hope for another i guess i'm not understanding why the simple answer is uh... unsatisfactory so the claim is calling for a particular determination to be made

[00:18:03] Speaker 01:
the nature of that requirement.

[00:18:09] Speaker 02:
It wouldn't make any sense that in modern times where all the cash memories are write back memories that wouldn't operate on all the write back memories.

[00:18:21] Speaker 02:
But I guess I'm struggling how to fit that into the claim language.

[00:18:26] Speaker 01:
Because what the claim language requires is a determination as to the status of the L1

[00:18:34] Speaker 01:
memory.

[00:18:36] Speaker 01:
And the reason it requires that is because if there's any modification, then you have all the problems that the patent is addressed to.

[00:18:44] Speaker 01:
There's no dispute, as I understand it, that if an L1 memory

[00:18:50] Speaker 01:
is in a modified cash, is in a modified state.

[00:18:53] Speaker 03:
Are you saying, basically, that the accused should check both caches?

[00:18:59] Speaker 03:
We can't tell which one it's checking, but it has to check the L1 in the course of its operation.

[00:19:04] Speaker 03:
And by doing that, it infringes your patent.

[00:19:08] Speaker 01:
I think that's right.

[00:19:08] Speaker 01:
To be honest, Your Honor, I'm having a little bit of a brain cramp.

[00:19:11] Speaker 01:
And I can't remember whether it checks both or whether the L2 has a copy of the L1.

[00:19:16] Speaker 01:
But whatever it's doing, it can't

[00:19:20] Speaker 01:
not determine what the status of the L1 is.

[00:19:23] Speaker 01:
That is just a prerequisite to the machine continuing to operate.

[00:19:28] Speaker 01:
And that's what the claim language requires.

[00:19:33] Speaker 03:
The point of the determining language, at least based upon the agreed claim construction, is just determining whether there's anything in the cache in a modified state.

[00:19:41] Speaker 01:
Exactly.

[00:19:42] Speaker 01:
And if there is anything in the cache in a modified state, that is discovered in the Viya chipset.

[00:19:49] Speaker 01:
Because it can't not be discovered.

[00:19:51] Speaker 01:
If it were in a modified state and that were overlooked somehow, then the system wouldn't work the way it's designed to work.

[00:20:01] Speaker 02:
So I guess there's an argument here, and I think you make it, that if they wanted to make this point, they should have made it as a claim construction argument, which wasn't made as a claim construction.

[00:20:11] Speaker 01:
I think that's right.

[00:20:12] Speaker 01:
The claim construction dispute.

[00:20:20] Speaker 01:
was Opti had proposed that L1 was the first level of cache memory, commonly known as L1.

[00:20:27] Speaker 01:
And VIA proposed that it was the first level of cache memory that uses a write back protocol.

[00:20:35] Speaker 01:
The rejection of theirs in favor of ours, I think, was simply being true to the patent.

[00:20:41] Speaker 01:
But again, I think that if they had wanted to claim, argue, that

[00:20:51] Speaker 01:
Infringement was only possible by a particular technique of determining what's in the L1 cash.

[00:21:00] Speaker 02:
I'm sorry?

[00:21:02] Speaker 01:
Yes.

[00:21:03] Speaker 01:
That could have been argued, but that wasn't.

[00:21:05] Speaker 01:
And I don't think the patent would have supported that.

[00:21:07] Speaker 01:
I mean, I think the patent is about cash consistency or cash coherency.

[00:21:13] Speaker 01:
And that was an issue for L1 caches.

[00:21:17] Speaker 01:
And it's an issue that's addressed via chipset.

[00:21:20] Speaker 01:
And that's why we think they imprinted that limitation.

[00:21:26] Speaker 01:
On the equivalency points, I think our basic position is in the briefs.

[00:21:36] Speaker 01:
I'll be very short on it.

[00:21:40] Speaker 01:
Even if you accept everything that Mr. Carpenter said, the reality is that our expert, the structures that were identified

[00:21:49] Speaker 01:
for each of the means were for the sequentially transferring means, the system controller and peripheral controller, and for the determining means, the same two structures, and in particular, the logic that was shown in them for generating two signals that accomplished the snooping operation.

[00:22:10] Speaker 01:
Dr. Smith discussed each of those structures, explained how the patent characterized their function

[00:22:16] Speaker 01:
their method of operation and the result they achieved and compared that directly to via chipsets relying extensively on their documentation and on testing.

[00:22:26] Speaker 01:
It was a classic means plus functional analysis, classic function weight result analysis.

[00:22:34] Speaker 01:
We'd respectfully submit that the jury's determination on that was fine.

[00:22:39] Speaker 02:
They say that your expert didn't address

[00:22:43] Speaker 02:
structural equivalence.

[00:22:45] Speaker 01:
That's simply not true.

[00:22:46] Speaker 02:
Can you show us where you did address the structural equivalence?

[00:23:06] Speaker 01:
It's basically the

[00:23:07] Speaker 01:
record passages that are discussed at pages 34 through 40 of our brief.

[00:23:13] Speaker 01:
Which page?

[00:23:23] Speaker 01:
So with respect to the means for determining, it's at

[00:23:34] Speaker 01:
a 1259 line 24 through 1262 I'm sorry through 1268 line 8 1259 through 1268 so

[00:24:03] Speaker 01:
He discussed the function.

[00:24:07] Speaker 01:
This was all done with respect to... Where does he discuss the structural equivalence?

[00:24:15] Speaker 01:
The structural equivalence is in that passage, Your Honor.

[00:24:20] Speaker 01:
That's where he's talking.

[00:24:21] Speaker 02:
What page?

[00:24:23] Speaker 01:
What line?

[00:24:25] Speaker 01:
I don't have the transcripts in front of me.

[00:24:27] Speaker 01:
I can't read you the line.

[00:24:32] Speaker 01:
Structural equivalents.

[00:24:37] Speaker 03:
What is the structural equivalent?

[00:24:40] Speaker 01:
So the structures in the patent are the chipsets.

[00:24:46] Speaker 01:
The structures in the Viya product are their north bridge and south bridge.

[00:24:50] Speaker 03:
Wait, I don't care about the Viya products.

[00:24:53] Speaker 03:
So you say in the patent the structures are the chipsets.

[00:24:55] Speaker 03:
How are those chipsets defined in the patent?

[00:24:58] Speaker 01:
The chipsets are referred to as the system controller.

[00:25:03] Speaker 01:
And the south bridge is the IPC, which is the Integrated Peripherals Control.

[00:25:07] Speaker 01:
Those referred to a particular chipset that was in the market.

[00:25:14] Speaker 03:
It was a commercial chipset.

[00:25:16] Speaker 03:
That operated in a particular way.

[00:25:17] Speaker 01:
Yes, sir.

[00:25:18] Speaker 01:
And there are figures in the patent that lay out the logic circuitry that implements the functionality.

[00:25:26] Speaker 01:
There are figures in the patent that show the operation, the behavior of the chipset.

[00:25:32] Speaker 01:
in very specific terms.

[00:25:35] Speaker 01:
What Dr. Smith did was he compared that functionality and that structure, which in this case is essentially the logic that... Yeah, but the problem is you can't point us where.

[00:25:47] Speaker 02:
You're supposed to come to the oral argument to be able to answer the questions.

[00:25:56] Speaker 01:
My colleague points me to page 170... I'm sorry, A1260.

[00:26:01] Speaker 01:
So, do VIA products have a chipset or are they chipsets?

[00:26:07] Speaker 01:
Yes.

[00:26:07] Speaker 01:
And you just put up the data sheets for the CN700 Northbridge, the VT8237R plus Southbridge.

[00:26:15] Speaker 01:
Now I'm going to focus most of these questions on these two particular products and then later on I'll ask you to explain how the technology you've seen these products relate to other products that are also at issue.

[00:26:24] Speaker 01:
Do the VIA chipsets, do they operate in the same technological context as the claim requires?

[00:26:30] Speaker 01:
And yes.

[00:26:31] Speaker 01:
And then he goes through in the following pages and points to the particular logic blocks in each of those chipsets, the signals that they generate, and compares those to the correlate disclosures in the patent.

[00:26:53] Speaker 01:
And where does he say that structure is equivalent?

[00:27:03] Speaker 01:
I guess what I'd say, Your Honor, is that what I just read to you is a statement that the structure is equivalent.

[00:27:12] Speaker 01:
He's saying, these are the structures that I'm comparing, and here is how they are equivalent.

[00:27:17] Speaker 01:
I mean, it's lengthy testimony.

[00:27:20] Speaker 02:
He says it's the same technological complex.

[00:27:23] Speaker 02:
That's not the same thing as saying they're equivalent.

[00:27:32] Speaker 01:
Well, for example, on page 1262, so let's just focus on these two and then we'll get to the signals coming out of the north bridge.

[00:27:39] Speaker 01:
I'll ask you whether those are also, I'm sorry, that's the wrong passage.

[00:28:07] Speaker 04:
Mr. Burden, we have to move on.

[00:28:10] Speaker 01:
Yes, sir, I understand.

[00:28:11] Speaker 01:
I'm using up my rebuttal time, but I think it's not worth it.

[00:28:14] Speaker 04:
You volunteer rebuttal time.

[00:28:15] Speaker 04:
Well, you don't have rebuttal time except for the cross appeal.

[00:28:19] Speaker 01:
Right, I understand.

[00:28:20] Speaker 04:
And that's only if it's raised by Mr. Coppenter.

[00:28:24] Speaker 04:
And we'll save your final half minute if you like.

[00:28:30] Speaker 01:
I'd rather spend it by giving Judge Steick a reference if I can.

[00:28:50] Speaker 01:
I guess the best answer I can give you, sir, is that I thought what we had done in the brief, and the reason I'm having trouble, there's not like a neat punch line.

[00:29:01] Speaker 01:
The reason I'm having trouble is because he did go through specific structures and say, do they function the same way?

[00:29:12] Speaker 01:
And he gave that evidence.

[00:29:15] Speaker 01:
Do they perform the same function?

[00:29:17] Speaker 01:
Do they do it in the same way, and do they achieve the same result?

[00:29:20] Speaker 01:
And he did it with reference to particular structures in each instance.

[00:29:26] Speaker 01:
I can't give you a place where it's wrapped up in a bow, but I thought we had in the briefing given you the passages where each of those steps was performed, and that's why I'm having trouble

[00:29:40] Speaker 01:
just giving you a punchline like that.

[00:29:43] Speaker 01:
And I apologize if that means I'm unprepared.

[00:29:46] Speaker 01:
But I do believe that we addressed it very specifically in the briefs and that he gave exactly that testimony.

[00:29:53] Speaker 01:
So I appreciate your time and your questions.

[00:29:56] Speaker 04:
I think I'm through.

[00:30:05] Speaker 00:
Picking up on the point that was just left off on.

[00:30:10] Speaker 00:
We went via great links in its reply brief to set out and fairly present Mr. Spitz's testimony.

[00:30:17] Speaker 00:
I think you'll see there that there was no side-by-side comparison of differences.

[00:30:23] Speaker 00:
If you look at this court's opinions in Cheminata and the cross opinion, 424 at 3rd, 315, this court has spent columns discussing relatively simple structures, comparing the differences, talking about the differences.

[00:30:39] Speaker 00:
When you get to a more complex structure, like in the Solomon case that came out of the ITC, there's pages of discussions about the differences of these highly complex systems.

[00:30:51] Speaker 00:
In this case, that record doesn't exist.

[00:30:54] Speaker 00:
OPTI failed to present the record.

[00:30:56] Speaker 00:
Another reason VIA discussed a new trial was we needed to walk through Mr. Mac Alexander's testimony to show this court how it should have been done.

[00:31:08] Speaker 00:
Mr. Mac Alexander.

[00:31:09] Speaker 00:
His testimony is presented in via his opening brief, clearly presented a compare and contrast of differences, discussed differences, and the figure at A2012 shows his conclusions and where he used certain differences between the two sets of structures as mileposts or guideposts to determine in substantial differences and ultimately, and he also discussed way and result and ultimately reached the conclusion

[00:31:37] Speaker 00:
They're not insufficient differences.

[00:31:39] Speaker 00:
In the last 30 seconds, there's a lot we could say about L1, L2.

[00:31:44] Speaker 00:
But the reality is technology has far progressed beyond where the patent was.

[00:31:48] Speaker 00:
And the patent at the top of, I believe it's column 20, talks about snooping L2 as yet a separate step.

[00:31:55] Speaker 00:
In modern architectures, the duffers are so deep that there is no need to snoop each cache all the North Bridge cares about.

[00:32:05] Speaker 00:
Is there modified data somewhere?

[00:32:07] Speaker 00:
writes back everything, flushes the buffers, and still keeps up with the PCI transfer.

[00:32:14] Speaker 00:
Thank you, Your Honors.

[00:32:16] Speaker 02:
Can I have one other question?

[00:32:19] Speaker 02:
Did you argue on J-Mall that they didn't produce any evidence of fill-in structure?

[00:32:25] Speaker 00:
Yes, Your Honor.

[00:32:26] Speaker 00:
That is our position.

[00:32:27] Speaker 02:
No, no.

[00:32:28] Speaker 02:
It is now.

[00:32:28] Speaker 02:
But did you argue that?

[00:32:30] Speaker 00:
Oh, yes, Your Honor.

[00:32:31] Speaker 00:
We started with Dalbert trying to.

[00:32:33] Speaker 00:
We saw the expert report realized

[00:32:36] Speaker 00:
They were not doing a legally sufficient analysis.

[00:32:38] Speaker 00:
We moved again at the conclusion of the plaintiff's case in chief.

[00:32:42] Speaker 00:
We moved again at the end of the defendant's case in chief.

[00:32:45] Speaker 00:
And we moved post-trial.

[00:32:47] Speaker 00:
So at every step of the way, we were telling the court, he should not be allowed to use these oversimplified function and way and result as proxies for claim annotations.

[00:32:56] Speaker 00:
He must do a differences analysis.

[00:32:59] Speaker 00:
That was always our argument.

[00:33:01] Speaker 04:
Thank you, Mr. Coppenter.

[00:33:04] Speaker 04:
There was nothing on the cross-appealing, so we'll conclude the argument.

[00:33:08] Speaker 04:
Take the case and revise it.

[00:33:11] Speaker 01:
Thank you.