[00:00:00] Speaker 03:
365 Wisconsin Alumni Research for SAFA.

[00:01:15] Speaker 01:
Thank you, Your Honor.

[00:01:18] Speaker 01:
May it please the Court.

[00:01:19] Speaker 01:
My name is Bill Ley, and together with my partners, Lauren Fletcher and Andrew Ganford, I represent Apple.

[00:01:26] Speaker 01:
From the briefing, the Court knows that we suggest that there are numerous errors below that resulted in a verdict not supported by the law or the evidence.

[00:01:37] Speaker 01:
Given the limits of time, I would like to focus on three issues today, but would be happy to address any issues the Court would like me to address.

[00:01:44] Speaker 01:
The three issues are these.

[00:01:46] Speaker 01:
First, the district court's mainly erroneous construction of the claim term prediction and its error in granting summary judgment on a disputed factual record.

[00:01:58] Speaker 01:
Second, the undisputed fact that the accused Apple products use hashed load tags corresponding to a group of load instructions and therefore do not make a prediction based upon a particular load instruction.

[00:02:14] Speaker 01:
And third, the district court's legally erroneous instruction on the issue of vicarious liability and the lack of any evidence that the pizza-sized wafers leaving the Samsung facility in Texas were capable of executing program instructions.

[00:02:29] Speaker 01:
Those are the three.

[00:02:30] Speaker 01:
And let me turn to the first.

[00:02:32] Speaker 02:
With respect to the second one, if we were to conclude that there was waiver with respect to the request for an instruction on the term particular,

[00:02:45] Speaker 02:
Where does that leave you?

[00:02:47] Speaker 01:
Your Honor, actually, there are two independent bases to address the fact that this undisputed fact that we use the hash load tags that are associated with a group of load tags should result in judgment for Apple.

[00:03:00] Speaker 01:
And there are two separate issues.

[00:03:02] Speaker 01:
One is not dependent upon the claim construction issue at all.

[00:03:06] Speaker 01:
It's not disputed.

[00:03:07] Speaker 01:
All of the experts agree that we hash the load tags.

[00:03:11] Speaker 01:
And as a consequence, every load tag is associated with a group of instructions.

[00:03:17] Speaker 01:
On the plain meaning of particular, that is not associating with a particular instruction.

[00:03:23] Speaker 01:
And all the court needs to do is consider the undisputed fact that we have load tags, the undisputed resulting fact that every load tag is associated with a group.

[00:03:34] Speaker 01:
And as Ronner knows, we can only have 4,026 load tags entirely.

[00:03:40] Speaker 01:
It's 2 to the 12th.

[00:03:41] Speaker 01:
There are millions and millions of load instructions.

[00:03:43] Speaker 01:
As a consequence, it's not dependent on it.

[00:03:46] Speaker 02:
But is the debate really over association or particular?

[00:03:51] Speaker 01:
It's particular, Your Honor.

[00:03:52] Speaker 01:
And it's not dependent upon the claim construction dispute, although that's an independent reason that there should be a new trial.

[00:03:59] Speaker 03:
Well, why is it not dependent on the claim construction?

[00:04:01] Speaker 03:
Isn't it dependent on both sides agree that it's plain and ordinary meaning?

[00:04:07] Speaker 03:
So if we're going to assess the jury verdict,

[00:04:10] Speaker 03:
against the claim construction, what do we do with that?

[00:04:14] Speaker 03:
I mean, we have to define what plain and ordinary meaning is.

[00:04:18] Speaker 01:
Your Honor, actually, that's correct.

[00:04:19] Speaker 01:
And I've tried to distinguish the two sets of circumstances, determining what the plain and ordinary meaning is, and going to Judge Romali's question about what happened when the court decided to construe the claim.

[00:04:31] Speaker 01:
From the outset, Apple took the position that the plain and ordinary meaning of the word particular was a single load instruction.

[00:04:39] Speaker 01:
That is, we took the position throughout the case that the focus should be on... By which you mean one and only.

[00:04:47] Speaker 00:
One and only.

[00:04:47] Speaker 00:
The problem is, a number of these formulations that get thrown around are not necessarily inconsistent with a load instruction, but possibly others as well.

[00:04:59] Speaker 00:
The question is, which side of that line does particular fall on?

[00:05:03] Speaker 01:
Your Honor, let me break it into two distinct arguments, if I could.

[00:05:07] Speaker 01:
Let's take the plain and ordinary meeting argument to go to Chief Judge Prost's question.

[00:05:12] Speaker 01:
We took the position, just as you say, that the plain and ordinary meeting is a single load instruction consistently.

[00:05:20] Speaker 01:
And we said the question is, for the hash load tags, are they associated with a single load instruction?

[00:05:27] Speaker 01:
And the answer is no.

[00:05:30] Speaker 01:
The only evidence you have about what the plain and ordinary

[00:05:34] Speaker 01:
ordinary meaning of particular is, is that A 144 to A 145.

[00:05:38] Speaker 01:
There, the district court quotes two dictionary definitions that are the plain and ordinary meaning.

[00:05:45] Speaker 01:
They both describe it as a single instruction.

[00:05:50] Speaker 02:
He then goes on to say- But you could have, if you're trying to predict something, right?

[00:05:56] Speaker 02:
Like you're trying to predict, like in the sad old days, whether or not the Cleveland Cavaliers were going to win.

[00:06:03] Speaker 02:
In a particular game, you would be saying it's not just associated with a particular event of LeBron James scoring so many points, because it could also be associated with the particular event of the other team not scoring as many points, or with someone else on the team scoring a particular number of points.

[00:06:23] Speaker 02:
In other words, just because you're talking about particular meaning one thing,

[00:06:30] Speaker 02:
doesn't mean it can't be associated with more than one.

[00:06:33] Speaker 01:
Actually, Your Honor, if you read the claim and the plaintiff made a claim, it does.

[00:06:38] Speaker 01:
It would be harder next year to predict whether the Cleveland Cavaliers are going to win without Ron James or not.

[00:06:43] Speaker 01:
That's a more difficult prediction.

[00:06:45] Speaker 01:
It's probably pretty easy to predict now.

[00:06:49] Speaker 01:
Game by game.

[00:06:50] Speaker 01:
But to go to Judge Bryson's question and to use your example,

[00:06:55] Speaker 01:
If you look at the claim, and this is plain meaning now.

[00:06:58] Speaker 01:
This is setting aside the claim construction dispute, the waiver issue that led us into this portion of the argument.

[00:07:04] Speaker 01:
The claim itself refers to the particular load instruction.

[00:07:10] Speaker 01:
The particular load instruction has an antecedent basis.

[00:07:14] Speaker 01:
The antecedent basis in the claim is a load instruction.

[00:07:18] Speaker 01:
And then after identifying a load instruction, you come down in the claim and it refers to the particular load instruction.

[00:07:25] Speaker 01:
As a district court said, the word particular has to have a meaning.

[00:07:29] Speaker 01:
If you eliminated the word particular and just said associated with the load instruction, it would have exactly the meaning that you've identified, exactly the meaning you've identified.

[00:07:40] Speaker 01:
There would be no additional meaning at all.

[00:07:42] Speaker 00:
Well, why isn't it possible?

[00:07:44] Speaker 00:
You've mentioned, and I think it's appropriate, the antecedent reference to the load instruction.

[00:07:51] Speaker 00:
Why isn't it?

[00:07:52] Speaker 00:
perfectly plausible reading of the language of the limitation A to say that they're using the word particular, in effect, as an equivalent of the word said.

[00:08:05] Speaker 00:
If the word said were there, I don't think your argument would hold water, would it?

[00:08:09] Speaker 00:
I mean, because it could be said instruction or it could be other instructions.

[00:08:14] Speaker 00:
It wouldn't be exclusionary.

[00:08:18] Speaker 01:
I'd say two things.

[00:08:19] Speaker 01:
One is it's not said.

[00:08:21] Speaker 00:
Well, I understand that.

[00:08:22] Speaker 00:
But the question is, can it be understood to be the equivalent of said?

[00:08:27] Speaker 00:
In other words, that when you say the particular instruction, the particular instruction to which we have referred above.

[00:08:35] Speaker 01:
And Your Honor, the answer is no for three reasons.

[00:08:38] Speaker 01:
The first is the word particular rather than the word said is used 20 times in the specification.

[00:08:44] Speaker 01:
And if I took your honor to the bottom of column 10 and the top of column 11, and you trace through what's happening in figures two, three, and four, you will see it as, without particular, the one instruction.

[00:08:56] Speaker 01:
So that's point number one.

[00:09:00] Speaker 01:
Point number two, the word particular is used consistently in these 20 different references to point to one and only one instruction.

[00:09:09] Speaker 02:
All throughout the course of the trial, you talked about a single instruction.

[00:09:15] Speaker 02:
And even when you at the last minute said you wanted, you said, you've probably already ruled on this, but we're going to ask you to give this instruction.

[00:09:24] Speaker 02:
And you said you wanted to say it was a single.

[00:09:28] Speaker 02:
You wanted to use a single load instruction.

[00:09:31] Speaker 02:
But single is very different from one and only one, which is now what you're arguing on appeal.

[00:09:37] Speaker 01:
Actually, Your Honor, what we're arguing is a single instruction.

[00:09:40] Speaker 01:
And if I could answer that and address the waiver issue, Your Honor,

[00:09:45] Speaker 01:
originally led to the question, and then actually address a couple of the other issues that I have during the course of our time.

[00:09:52] Speaker 01:
Here's the reason.

[00:09:53] Speaker 01:
Respectfully, there was a different set of circumstances than I think Your Honor just described.

[00:10:00] Speaker 01:
But my third point, going to Judge Bryson, is actually built upon what you just said.

[00:10:04] Speaker 01:
The judge determined what the plain, ordinary meaning was.

[00:10:08] Speaker 01:
He said it was the single instruction.

[00:10:11] Speaker 00:
Well, but then, on Jamal, he came back and gave a very

[00:10:15] Speaker 00:
my eyes, at least different interpretation of what the plain and ordinary meaning was, which is unhelpful to you.

[00:10:22] Speaker 01:
And Your Honor, there is a bizarre and unusual set of circumstances that led to that portion of the Jamwell opinion.

[00:10:31] Speaker 01:
But to go to Judge O'Malley's question, which is where it began, is we didn't ask for the construction.

[00:10:36] Speaker 01:
And that's the place why I was respectfully disagreeing.

[00:10:40] Speaker 02:
Yes, sir.

[00:10:40] Speaker 02:
In writing right here.

[00:10:41] Speaker 01:
No, no.

[00:10:42] Speaker 01:
What happened is this, Your Honor, on the third day of trial,

[00:10:46] Speaker 00:
This is Worf's motion to exclude your testimony to the contract.

[00:10:50] Speaker 01:
And then we said in response, we have said from the outset that this is plain meaning.

[00:10:55] Speaker 01:
It is still plain meaning.

[00:10:56] Speaker 01:
There's no construction required.

[00:10:58] Speaker 01:
But if there's a construction required, it is the single construction.

[00:11:03] Speaker 02:
But you never asked for it.

[00:11:04] Speaker 02:
And then at the end of the charge conference, you submitted one document that said, you've probably already ruled on this.

[00:11:11] Speaker 02:
We know we didn't ask for a construction.

[00:11:14] Speaker 02:
And now we want it to be a single, not one and only, but a single.

[00:11:17] Speaker 01:
No, Your Honor.

[00:11:18] Speaker 01:
Actually, what happened is when we responded to worse motion, we said, we don't think construction's required.

[00:11:26] Speaker 01:
But if it is, it is a single instruction.

[00:11:29] Speaker 01:
And if you decide to instruct, if you decide to construe, you should instruct the jury.

[00:11:35] Speaker 01:
On October the 8th, which was the third day of trial, the judge issued his opinion and construed.

[00:11:40] Speaker 01:
We then said the very next morning,

[00:11:43] Speaker 01:
Having adopted a construction, even though we suggested you didn't need to, the jury should be charged.

[00:11:48] Speaker 01:
This was still two days of trial left.

[00:11:51] Speaker 01:
It was before the charging conference.

[00:11:53] Speaker 01:
It was within 24 hours of his ruling.

[00:11:56] Speaker 01:
And that's why I said I was disagreeing with the facts, because it's a different set of circumstances.

[00:12:01] Speaker 03:
Can I just ask you a little about that?

[00:12:03] Speaker 03:
I mean, you've mentioned that this was a really set of unusual circumstances.

[00:12:06] Speaker 03:
Yes.

[00:12:06] Speaker 03:
And I can't disagree with you.

[00:12:10] Speaker 03:
What was going on here?

[00:12:11] Speaker 03:
I mean, even originally before this motion and your cross-examination of Dr. Conti, I guess, there seemed to be a pushback or attention and the court's aversion to your getting in too much detail as to what an ordinary plain and ordinary meaning is.

[00:12:27] Speaker 03:
And indeed, even in his, when he decided the granted your mo or denied the motion and he gave you the construction that you now say you wanted, which is single, his last sentence is,

[00:12:39] Speaker 03:
But don't emphasize the importance of that.

[00:12:42] Speaker 03:
So I'm just asking you for a little help here in terms of there was a constant pushback.

[00:12:47] Speaker 01:
Your Honor, I wish I could give you an explanation.

[00:12:50] Speaker 01:
I don't have one, even though I was there.

[00:12:52] Speaker 01:
But what happened was this.

[00:12:54] Speaker 01:
We go to Judge O'Malley's question.

[00:12:55] Speaker 01:
We have litigated this as a factual issue from the outset, plain and ordinary meaning we don't have a particular instruction we hatch.

[00:13:04] Speaker 01:
And the facts, as I said at the outset, are undisputed.

[00:13:08] Speaker 01:
That had been our expert view from the outset.

[00:13:11] Speaker 00:
Then, on the second day of trial... But contrary to the Wharf's expert's position.

[00:13:18] Speaker 01:
Who said, before trial, that the word particular had no meaning.

[00:13:23] Speaker 01:
It added nothing to the claim.

[00:13:24] Speaker 00:
Well, but that's... Well, okay.

[00:13:27] Speaker 01:
And then we get to the portion of the trial that Judge Prost is referring to, and the issue of whether

[00:13:34] Speaker 01:
is claim construction coming in under the guise of an expert's deposition becomes an issue.

[00:13:40] Speaker 01:
And there's a series of objections that I made.

[00:13:43] Speaker 01:
There is some that were sustained, some that were overruled.

[00:13:47] Speaker 01:
But then Whorf makes the motion.

[00:13:50] Speaker 01:
And the judge finds that there is a fundamental claim construction issue.

[00:13:54] Speaker 01:
He finds that he should construe it.

[00:13:57] Speaker 01:
He does construe it to say a single instruction.

[00:14:02] Speaker 01:
And then he says,

[00:14:03] Speaker 01:
But I'm not going to tell the jury.

[00:14:05] Speaker 01:
And by the way, you can't focus on it with your expert who's going to testify.

[00:14:09] Speaker 02:
Well, he didn't really say he was construing it.

[00:14:11] Speaker 02:
He didn't really say, I'm not telling the jury.

[00:14:13] Speaker 02:
What he said is, you all both said plain and ordinary meaning.

[00:14:17] Speaker 02:
And he said that your expert's testimony was consistent with the concept of plain and ordinary meaning, so I won't exclude it.

[00:14:24] Speaker 02:
And that's not exactly the same thing as saying,

[00:14:27] Speaker 02:
There's a claim construction dispute.

[00:14:29] Speaker 02:
I'm going to construe it.

[00:14:30] Speaker 02:
I'm not going to tell the jury.

[00:14:31] Speaker 02:
I mean, you're trying to set up this O2 micro thing, and that's not exactly how it happened.

[00:14:36] Speaker 02:
He was ruling in your favor by saying, go ahead.

[00:14:39] Speaker 02:
Make this argument.

[00:14:40] Speaker 01:
Give this testimony.

[00:14:41] Speaker 01:
But I don't think it's quite just go ahead.

[00:14:43] Speaker 01:
When someone says there's a fundamental claim construction dispute, I'm going to construe it, whichever construction you decide.

[00:14:51] Speaker 01:
And if you read his opinion, he addresses their argument

[00:14:55] Speaker 01:
surprising language somehow expands the scope of the claim.

[00:14:59] Speaker 01:
And then you come back and say, having construed it, I'm not going to instruct the jury.

[00:15:04] Speaker 01:
And you should not emphasize it, but I think there are two distinct issues here.

[00:15:10] Speaker 01:
One is there is a plain and ordinary meaning.

[00:15:12] Speaker 01:
It may be

[00:15:13] Speaker 03:
Well, let me ask you that.

[00:15:14] Speaker 03:
Let me ask you hypothetically, if none of this had happened during trial, if the parties had stuck to plain and ordinary meaning, they hadn't filed their motion to exclude your witness testimony and it had just gone on and there was no discussion of giving it an additional construction, where would we be?

[00:15:30] Speaker 03:
We'd have a jury verdict.

[00:15:31] Speaker 03:
You and Mr. Chu would be up here arguing.

[00:15:35] Speaker 03:
claim construction on plain and ordinary meaning?

[00:15:38] Speaker 03:
I mean, how would this go down?

[00:15:39] Speaker 03:
Sufficiency.

[00:15:40] Speaker 03:
It would be limited to sufficiency.

[00:15:42] Speaker 01:
That's exactly right.

[00:15:43] Speaker 01:
And I actually think that's a more articulate way of saying what I was trying to say at the outset, that there are two different issues here.

[00:15:49] Speaker 01:
There is a simple sufficiency argument, which is if we have hash load tags, and there are literally groups of millions or hundreds of thousands of instructions associated with a load tag, can that be particular given the plain and ordinary meaning?

[00:16:05] Speaker 01:
That's a separate issue.

[00:16:06] Speaker 01:
It's a sufficiency issue.

[00:16:07] Speaker 01:
We've argued it that way.

[00:16:08] Speaker 02:
This is a good segue then.

[00:16:09] Speaker 02:
So we have Dr. August's testimony and your own documentation where he testified that the chances in which a prediction will be associated with more than one instruction is 0.1%.

[00:16:24] Speaker 01:
Actually, Your Honor, I'm glad you brought that up because one of the things I'd like you to do is, it's like the Cleveland Cavaliers.

[00:16:32] Speaker 01:
If you look at Worf's brief in pages 17 to 18 and page 27, which is the point at which they're making that characterization of his testimony, and I think I got it correctly, but it's 17 to 18 and it's 27.

[00:16:52] Speaker 01:
And I'd like to answer this question very specifically.

[00:16:57] Speaker 01:
You will see that there's a quote provided to you.

[00:17:00] Speaker 01:
The quote is,

[00:17:01] Speaker 01:
A and B were the words only when inserted into the quote.

[00:17:06] Speaker 01:
And it's the only when that's inserted into the quotes of those pages that would allow someone to reach the conclusion you just reached about Dr. August's testimony.

[00:17:17] Speaker 01:
But if you go to the testimony that's, in fact, cited at appendix 2239 to 2240 and 2237 to 2238, what Dr. August said is,

[00:17:30] Speaker 01:
No.

[00:17:32] Speaker 01:
They hash 100% of the time.

[00:17:35] Speaker 01:
They are associated with a group 100% of the time.

[00:17:37] Speaker 03:
Where is the 100%?

[00:17:39] Speaker 03:
I'm looking.

[00:17:40] Speaker 03:
You've got some.

[00:17:41] Speaker 03:
Right.

[00:17:41] Speaker 03:
I thought he says 99.98.

[00:17:42] Speaker 01:
That is aliasing, Your Honor.

[00:17:45] Speaker 01:
And that's why hashing and aliasing, as Dr. Conti admitted at A2516, are two different things.

[00:17:54] Speaker 01:
We hash 100% of the time.

[00:17:57] Speaker 01:
Our millions of load instructions are associated with load tags.

[00:18:00] Speaker 01:
all the time.

[00:18:02] Speaker 01:
And this only when, which is key to communicating the argument that your honor just articulated, isn't in the testimony.

[00:18:10] Speaker 01:
In fact, if you read Dr. August's testimony, what he says is, yes, there's aliasing 1% of the time or so, but there's hashing 100% of the time.

[00:18:20] Speaker 01:
And 100% of the time, the load instructions are associated as a group with a load tag.

[00:18:28] Speaker 02:
So how do you define aliasing, and then how do you read

[00:18:31] Speaker 02:
this testimony in conjunction with that definition?

[00:18:34] Speaker 01:
Your Honor, the best I can do is if you actually look at A2516 to 2517, you'll see that Dr. Conte admits that hashing is different from aliasing.

[00:18:44] Speaker 01:
Aliasing is actually irrelevant to the particular limitation we're talking about today.

[00:18:49] Speaker 01:
Aliasing is when two different load instructions actually misspeculate with two different store instructions.

[00:18:59] Speaker 01:
And they're updating the same prediction at the same time because they happen to have the same load tag.

[00:19:05] Speaker 01:
And actually, and this is why this only when is very important, because what Dr. August said is, no, Apple hashes 100% of the time.

[00:19:17] Speaker 01:
When it hashes, all of the load instructions are associated with different load tags.

[00:19:24] Speaker 01:
The fact that aliasing occurs a small percentage of the time has nothing to do.

[00:19:29] Speaker 01:
with whether there are load tags that are hash load tags that apply to instructions.

[00:19:38] Speaker 02:
What about the documentation that says that your LSD predictor can be thought of as uniquely identifying load instructions?

[00:19:47] Speaker 02:
And that seems to imply from the technical documentation that that doesn't mean that it always occurs.

[00:19:55] Speaker 01:
Your Honor, first it's identifying the LSD predictor, which is identifying data dependencies rather than misspeculations.

[00:20:05] Speaker 01:
What happens in the Apple system, which is actually more conservative and more inclusive, I guess is the best way to put it, if a particular load instruction misspeculates and gets a prediction, if a particular load instruction has a data dependency in the Apple system and it gets a prediction,

[00:20:25] Speaker 01:
and a load tag.

[00:20:27] Speaker 01:
When that load tag comes up again, every single instruction that has that load tag will be prevented from speculating.

[00:20:38] Speaker 01:
And so while the 752 patent goes particular instruction by particular instruction, and Judge Bryson, if you looked at column four, they talk about the purpose of the invention, which is to identify a discrete number of load instruction and store instruction pairs that are likely to create

[00:20:55] Speaker 01:
problems when they misspeculate.

[00:20:57] Speaker 01:
The Apple system actually is more conservative because if you have a particular load instruction that has a data dependency, it gets a load tag, and then that load tag comes back around, every instruction that has that load tag is prevented from speculating.

[00:21:18] Speaker 01:
And it will prevent speculation in a circumstance where the 752 patent would allow it.

[00:21:25] Speaker 01:
Why?

[00:21:25] Speaker 01:
It's more conservative.

[00:21:28] Speaker 01:
The system, other portions of the system are more robust.

[00:21:31] Speaker 01:
It allows you to do that.

[00:21:33] Speaker 01:
It's a more conservative system.

[00:21:34] Speaker 01:
That's why, in particular, it's important to the claim.

[00:21:39] Speaker 01:
It's important in the specification.

[00:21:41] Speaker 01:
Now, I'm way over the point.

[00:21:43] Speaker 03:
Yeah, I do want you to talk to us a little about vicarious liability, and not on the legal question, but just on the evidence.

[00:21:51] Speaker 01:
And, Your Honor, if I could just address that and then

[00:21:53] Speaker 01:
just a minute or maybe 30 seconds or so on the question of steely and summary judgment.

[00:21:59] Speaker 01:
And I can do it in whichever order you prefer.

[00:22:02] Speaker 03:
I wanted to go to vicarious liability first, because they're relying.

[00:22:05] Speaker 03:
I mean, the question for us is, what was the evidence that the jury could have relied on in concluding that there was at least the demonstrated capability of infringing in Texas?

[00:22:17] Speaker 01:
And Your Honor, I think the key here is to look at

[00:22:20] Speaker 01:
appendix page 3924 to 3925.

[00:22:22] Speaker 01:
And what happened here is this.

[00:22:28] Speaker 01:
Early on... Hold on.

[00:22:30] Speaker 01:
Let me just reach the 3920... I'm sorry.

[00:22:36] Speaker 01:
What was the... 3924 to 3925, Your Honor.

[00:22:38] Speaker 01:
Okay.

[00:22:38] Speaker 01:
Thank you.

[00:22:41] Speaker 01:
And if you look at 3924, you will see that Dr. Conkey was asked a very broad question about

[00:22:48] Speaker 01:
what these wafers were capable of doing.

[00:22:51] Speaker 01:
And there was an objection.

[00:22:53] Speaker 01:
And you'll see, were they capable of executing instructions?

[00:22:56] Speaker 01:
Were they capable of functioning?

[00:22:58] Speaker 01:
And there is an objection.

[00:23:01] Speaker 01:
The objection is sustained because Dr. Conti has not done any testing himself.

[00:23:07] Speaker 01:
He hasn't identified what the test port can test for.

[00:23:11] Speaker 01:
It isn't in his expert report.

[00:23:13] Speaker 01:
Then we go to the next page, and he's asked,

[00:23:16] Speaker 01:
a question about the test ports.

[00:23:19] Speaker 01:
He answers it generally, but he never says again what the test port can test for, whether it was ever used to test for anything here in the United States or anywhere, and whether that test could tell you whether the wafers were capable of performing program instructions.

[00:23:42] Speaker 01:
If he had gone to the next question, which would have been

[00:23:45] Speaker 01:
Could they execute program instructions?

[00:23:47] Speaker 01:
Were they capable?

[00:23:48] Speaker 01:
We would have been right back to the question we objected to on 3924, and he wouldn't have been allowed to answer it.

[00:23:54] Speaker 01:
So the record, as it stands, and this is why we hope, we suggest, that reading 3924 and 3925 together are key, because what Worth has done in its brief is it's excerpted the very bottom of page 3925 as if that answer existed in the abstract and said, see, this is enough.

[00:24:16] Speaker 02:
But wait, he says that you test every chip.

[00:24:20] Speaker 02:
And so what you're saying is he never went to the point of saying that you have to test it before you export it?

[00:24:27] Speaker 01:
No, Your Honor.

[00:24:28] Speaker 01:
He actually, the problem was this, and this is what the judge identified.

[00:24:33] Speaker 01:
He never said, he said it's tested.

[00:24:34] Speaker 01:
He didn't say what it's tested for, what the test is.

[00:24:38] Speaker 01:
whether it's capable of performing program instructions.

[00:24:41] Speaker 01:
And in fact, if you look at 3924.

[00:24:42] Speaker 02:
He does say all the functionality is present, right?

[00:24:46] Speaker 01:
That's precisely what he was precluded from and why I suggest that if he were asked the next question, the same objection would have been.

[00:24:55] Speaker 03:
But what he says was it's quite close, I understand.

[00:25:01] Speaker 03:
Every two years, the chips, you use it to test the chips to make sure they work, to make sure they're capable.

[00:25:08] Speaker 01:
And Your Honor, if you go back just one page, look at the question at line nine.

[00:25:15] Speaker 01:
Did you reach a conclusion in your report as to whether, after printing in Austin, the functionality, circuitry, and design are established?

[00:25:22] Speaker 01:
Yes.

[00:25:23] Speaker 01:
What's the basis for that?

[00:25:25] Speaker 01:
And then that's what's excluded.

[00:25:27] Speaker 01:
And so I think reading that one line, and again, I would be arguing to you in any event that that one line doesn't tell you what's being tested for.

[00:25:37] Speaker 01:
to test whether the power will go on.

[00:25:40] Speaker 01:
It doesn't suggest that there's any program instructions that can be executed.

[00:25:44] Speaker 01:
That, I would be arguing to you, is insufficient as a matter of law in any event.

[00:25:48] Speaker 01:
But when you read these two pages together, and you realize what has occurred, which is the judge has stricken and excluded testimony about functionality, circuitry, and design, and whether they're established, because he hasn't disclosed it, because he hasn't done the test, because he hasn't described the test,

[00:26:07] Speaker 01:
That's insufficient.

[00:26:09] Speaker 01:
So let me make one final point on that, Your Honor, and then just two quick points on the summary judgment issue.

[00:26:16] Speaker 01:
On the legal issue, I know you wanted me to address the factual issue first.

[00:26:21] Speaker 01:
But the legal issue is important, because the Myers case, the Supreme Court case, says that if you're going to impose vicarious liability, it needs to be something more than direction and control.

[00:26:33] Speaker 01:
They actually literally use those words right out of the restatement.

[00:26:38] Speaker 01:
Akamai is direction or control.

[00:26:40] Speaker 01:
And what they say is required in addition is an agency relationship, a principal agent relationship where one person holds the other out as their agent and the other acts as the agent of the other.

[00:26:52] Speaker 00:
So do you think Akamai's characterization of method claims as being governed by, among other things, direction control is wrong and consistent with Myers?

[00:27:01] Speaker 01:
No, for this reason, Your Honor.

[00:27:03] Speaker 01:
Akamai Ashley comes right out and says that divided infringement is different than vicarious liability.

[00:27:09] Speaker 01:
Vicarious liability is saying that someone else is directly infringed, but I want to impose liability for that upon you for some reason.

[00:27:19] Speaker 00:
So if you accept that Akamai is correct and we can have divided infringement, then if I'm being charged with infringement based on making

[00:27:31] Speaker 00:
a device and I make, let's say, the case for a phone and then I contract with you to make the rest of the phone, then I'm liable for infringement for the entire phone.

[00:27:48] Speaker 00:
But if I contract with you to make the case and the rest of the components of the phone, then I'm not liable for direct infringement?

[00:27:56] Speaker 00:
Am I missing the correct distinction?

[00:27:59] Speaker 01:
Yes.

[00:27:59] Speaker 01:
And Your Honor, actually, Akamai says there's no distinction.

[00:28:02] Speaker 01:
Akamai at 1022 says that vicarious liability is a misnomer for the joint and divided infringement issue.

[00:28:11] Speaker 00:
Well, I understand that.

[00:28:12] Speaker 00:
But my question is, are you saying that as long as I play some role in the infringement by making a portion of the device that Akamai covers me,

[00:28:26] Speaker 00:
But if I contract out to you to make the entire device, Akamai and the law generally does not cover me.

[00:28:34] Speaker 00:
I'm not an infringer.

[00:28:36] Speaker 01:
No, but I think this is the precise explanation.

[00:28:40] Speaker 01:
I think as you've just posed the hypothetical, it's, as I understand it, an apparatus claim, not a method claim.

[00:28:46] Speaker 01:
Correct.

[00:28:47] Speaker 01:
Akamai is dealing with method claims.

[00:28:49] Speaker 00:
Well, I understand.

[00:28:49] Speaker 00:
But I thought you were saying that, well, go ahead.

[00:28:53] Speaker 01:
And Your Honor, here's the second part of the answer to the question.

[00:28:55] Speaker 02:
Akamai, the final Akamai.

[00:28:58] Speaker 02:
I don't know, some people call it five, some people call it six.

[00:29:00] Speaker 02:
But why?

[00:29:00] Speaker 02:
I mean, that is when the court turned to look at the definition of 271A.

[00:29:05] Speaker 01:
And that's correct, Your Honor.

[00:29:08] Speaker 02:
So it wasn't just saying, let's look at, you know, induced infringement of method claims.

[00:29:14] Speaker 02:
The Supreme Court told us to go back and rethink our definition of infringement, generally.

[00:29:21] Speaker 01:
Your Honor, I agree.

[00:29:23] Speaker 01:
But if you look at what the opinion says about the difference between vicarious liability and divided into an infringement, it makes the distinction that I think I'm trying to make between method claims and apparatus claims.

[00:29:37] Speaker 01:
That's the answer, in part, to your question.

[00:29:39] Speaker 01:
But Judge Bryson, there's a second answer to your question, which is 271B, 271A.

[00:29:44] Speaker 00:
Well, I would be liable for inducement, no doubt.

[00:29:47] Speaker 00:
But I'm focusing on 271A as Judge O'Malley.

[00:29:52] Speaker 01:
In a supply contract context, if entering into a supply contract and giving someone specifications is enough to become vicariously liable, no one's ever going to bring a 271B or 271C claim again.

[00:30:04] Speaker 01:
And in fact, what happened in this case is Whorf abandoned the indirect infringement claims to avoid the issues of intent that would be implicated by 271B and 271C.

[00:30:16] Speaker 00:
But do you agree with the premise of my question that if I made the case

[00:30:22] Speaker 00:
and contracted with you to make all the components other than the case, that then I would be directly liable under 271A, but not if I didn't make the case.

[00:30:36] Speaker 01:
No.

[00:30:36] Speaker 00:
You don't agree with that?

[00:30:37] Speaker 01:
I don't agree with that.

[00:30:37] Speaker 00:
I wouldn't be liable in either event?

[00:30:39] Speaker 01:
No, because the direct infringer is me as I make the case.

[00:30:43] Speaker 01:
You've contracted to have it.

[00:30:45] Speaker 00:
No, I made the case.

[00:30:46] Speaker 01:
You made the components.

[00:30:47] Speaker 01:
All right.

[00:30:48] Speaker 01:
All right.

[00:30:48] Speaker 01:
I may have it reversed.

[00:30:50] Speaker 01:
All right, that's fine.

[00:30:51] Speaker 01:
You're the direct infringer.

[00:30:52] Speaker 01:
That makes it easier.

[00:30:53] Speaker 01:
I'm actually asking you to directly infringe.

[00:30:56] Speaker 01:
The question is, am I vicariously or indirectly liable?

[00:31:00] Speaker 01:
The answer is there are two different boxes to put this into.

[00:31:04] Speaker 01:
One, it's a question of indirect infringement.

[00:31:06] Speaker 01:
That's 271B, 271C.

[00:31:10] Speaker 01:
A second box, which they're now trying to put these pizza-sized wafers into, is 271A.

[00:31:17] Speaker 01:
You are the direct infringer.

[00:31:19] Speaker 01:
When can I be directly liable?

[00:31:20] Speaker 01:
And that's where the Meyer Supreme Court case, quoting the restatement explicitly says, if you're going to make someone vicariously liable for someone else's act, then it has to be a principal-agent relationship.

[00:31:33] Speaker 02:
And this contract explicitly says... That case predates a lot of changes to the restatement where

[00:31:43] Speaker 02:
And we said in Akamai that the point is attribution.

[00:31:47] Speaker 02:
And that's what we should be looking at.

[00:31:48] Speaker 02:
And that vicarious liability is a misnomer because the concept is attribution.

[00:31:53] Speaker 02:
And how can one's acts be attributed to another?

[00:31:58] Speaker 01:
And I think, Your Honor, how can one's act be attributed to another so that that person is performing all the steps of a method claim?

[00:32:07] Speaker 01:
And I would suggest that if that was extended to

[00:32:11] Speaker 01:
your example with an apparatus claim, and to vicarious liability, I'm not sure who would continue to pursue a 271B or 271C claim.

[00:32:20] Speaker 01:
But more importantly, it's in- That might just be the reality.

[00:32:24] Speaker 01:
I think it'd be inconsistent with Myers and the Supreme Court precedent.

[00:32:28] Speaker 01:
And that is what we said to the district court judge.

[00:32:32] Speaker 03:
You get your extra minutes.

[00:32:34] Speaker 01:
Yes.

[00:32:35] Speaker 01:
The last minute is this.

[00:32:36] Speaker 01:
Quite apart from everything we discussed, there was a pre-trial summary judgment rule.

[00:32:40] Speaker 01:
It was based upon a claim construction of prediction that narrowed the term prediction.

[00:32:46] Speaker 01:
That is, as you know, we suggest legal error.

[00:32:51] Speaker 01:
It is basically taking the plain meaning, a plain meaning which we all agree the term has, and then implicitly, as the district court said, narrowing it.

[00:33:02] Speaker 01:
But even if the claim construction were correct, we had Dr. Caldwell, the chief architect of the Pentium Prowl,

[00:33:09] Speaker 01:
come in and give a declaration that described in detail how the Steely reference satisfied the limitations of the claim, no matter what the definition of prediction was.

[00:33:19] Speaker 01:
And the judge simply resolved that issue himself at page 831.

[00:33:23] Speaker 02:
He basically said... But didn't he say, even assuming, didn't he go then say, even assuming we accept all of Apple's expert testimony?

[00:33:32] Speaker 01:
No.

[00:33:33] Speaker 01:
What he said is that Apple's expert appears to have, it appears that Steely has an error

[00:33:39] Speaker 01:
in the manner in which the process is executed, even though Dr. Caldwell said there's no error, this is how it would work, even though the named inventor, Simon Steely, said this is how it worked.

[00:33:52] Speaker 02:
Even though... But he says, even assuming one might infer that such overwriting occurs and constitutes updating.

[00:33:58] Speaker 02:
And then he went on to conclude that Steely still doesn't.

[00:34:03] Speaker 02:
describe a tag replacement system that would constitute prediction.

[00:34:08] Speaker 01:
And that's error for two reasons.

[00:34:09] Speaker 01:
The first thing is, if that type of updating and overriding is specifically described in 752 pen as updating, it's described as temporal locality and deletion with a subsequent instruction coming through.

[00:34:24] Speaker 01:
And so as a matter of claim construction, it would be wrong.

[00:34:28] Speaker 01:
But even if you set aside the claim construction issue, that was a question for the jury to resolve.

[00:34:33] Speaker 01:
What you have on the record before you is this.

[00:34:36] Speaker 01:
You have Dr. Caldwell's testimony, Dr. Caldwell's description of what Steely does, the inventor's description of what Steely does.

[00:34:45] Speaker 01:
It says, yes, it does precisely what Dr. Caldwell says.

[00:34:49] Speaker 01:
And it overwrites tags as the process moves on.

[00:34:52] Speaker 01:
And you have the judge resolving that factual dispute because he thinks there's an error in Steely.

[00:34:59] Speaker 02:
But this only relates to anticipation.

[00:35:01] Speaker 02:
Did you then argue,

[00:35:03] Speaker 02:
an obviousness case relating to Steely?

[00:35:05] Speaker 01:
Well, Your Honor, we argued an obviousness case that had different aspects of Steely, but because the district court had decided that Steely didn't disclose a required prediction, we were precluded from arguing both anticipation and obviousness based upon Steely disclosing the prediction.

[00:35:25] Speaker 01:
So the answer is this summary judgment ruling precluded us from anticipation of Steely

[00:35:31] Speaker 01:
recruited us from an obviousness theory that we had articulated that was based upon steely disclosing a prediction.

[00:35:41] Speaker 01:
And all of this question about the prior was not only critically important to the Invalidity case, it was a big part of the damage in this case as well.

[00:35:50] Speaker 00:
Can I return you to particularity?

[00:35:54] Speaker 00:
Sure.

[00:35:54] Speaker 00:
As painful as it might be to replow this ground.

[00:36:01] Speaker 00:
One of the things that's troubling me, and we touched on it briefly before, but I think I never really got a resolution.

[00:36:11] Speaker 00:
The judge said, and it sounds to me, looking at the timing of all of this, that a lot of what happened happened in a real hurry, in the middle of trial, the sort of hurly-burly of things that ended up getting decided probably and written overnight between the motion

[00:36:29] Speaker 00:
your response, the judge's order, and then the text order that came out the next day.

[00:36:34] Speaker 00:
So there's some rough edges there, obviously.

[00:36:38] Speaker 00:
But what bothers me is that you've got an instruction, pardon me, you've got the construction by the court that is the one that you latched onto, which is the single load instruction.

[00:36:51] Speaker 00:
But then when we get to the J-Mall,

[00:36:56] Speaker 00:
I mean, I find single at least open to some ambiguity as to what it means, because it doesn't say single and no other.

[00:37:03] Speaker 00:
So when we get to the JML, sure enough, the judge says, well, the meaning of the term is, in effect, he says, and the evidence is sufficient to show that a reasonable jury could conclude that the prediction was associated with a particular load instruction, even if that same prediction could be associated with other load instructions.

[00:37:23] Speaker 00:
All of which leads me to believe that, at least if the judge is being consistent, that he believed that the single load instruction construction was consistent with Worf's view that this load instruction could be accompanied in the load tags by many others.

[00:37:44] Speaker 00:
Why is that not right?

[00:37:45] Speaker 01:
Your Honor, I would say these things.

[00:37:50] Speaker 01:
Where you began was there is some confusion if you look at the different things.

[00:37:54] Speaker 01:
I don't think anybody disagrees with that.

[00:37:56] Speaker 01:
And the answer is there surely was.

[00:37:59] Speaker 01:
Let me try to unpack the confusion as best I can and answer your question.

[00:38:04] Speaker 01:
If you look at A144 to A145, we know the following things, and I don't think there's any dispute.

[00:38:11] Speaker 01:
One is Wharf brought the motion to exclude our expert from testify.

[00:38:15] Speaker 01:
No one disputes that.

[00:38:16] Speaker 01:
The second is we said

[00:38:19] Speaker 01:
No, it's plain meaning.

[00:38:20] Speaker 01:
This is a factual question.

[00:38:22] Speaker 01:
But if you're going to construe it, it means a single instruction.

[00:38:26] Speaker 01:
And here are dictionary definitions.

[00:38:30] Speaker 01:
And here is what the specification says.

[00:38:33] Speaker 01:
Whorf came back and said, no, no, no.

[00:38:36] Speaker 01:
It can be a group, which it had to do because the undisputed facts are that we hatch.

[00:38:42] Speaker 01:
And we have a limited number of load tags.

[00:38:44] Speaker 01:
And it said it can be the group not because particular had a different meaning.

[00:38:49] Speaker 01:
but because of the word comprising.

[00:38:51] Speaker 01:
So at A1, it's a separate argument altogether.

[00:38:54] Speaker 01:
But you get to A144 and 145.

[00:38:55] Speaker 01:
And I think this was something that Judge Prost asked me about.

[00:39:01] Speaker 01:
You have him saying there's a fundamental claim construction dispute.

[00:39:05] Speaker 01:
You have him saying that I'm going to construe it.

[00:39:08] Speaker 01:
You have him acknowledging worse argument that we have waived to go to Judge O'Malley's original question that got started down the particular path.

[00:39:17] Speaker 01:
And he doesn't decline to rule because either of us is waived.

[00:39:21] Speaker 01:
He said, no, I'm going to rule.

[00:39:22] Speaker 01:
And what he does then is he says, here are the dictionary definitions, both of which, Your Honor, support a single unique instruction.

[00:39:32] Speaker 01:
And that is critically important because it's consistent with the disclosed embodiments, every single disclosed embodiment in the patent, every single one of them.

[00:39:43] Speaker 01:
And then he says, here's the surprising argument.

[00:39:46] Speaker 01:
The comprising argument is not legally correct.

[00:39:49] Speaker 01:
And I construed it to be a single instruction.

[00:39:54] Speaker 01:
Then we get the, but I'm not going to instruct the jury.

[00:39:57] Speaker 01:
And you should not emphasize the word particular.

[00:40:01] Speaker 01:
So we try the rest of the liability case only takes about a day and a half.

[00:40:06] Speaker 00:
But the very next day in the text order, he comes back and he says, you have waived.

[00:40:10] Speaker 01:
Yes.

[00:40:10] Speaker 01:
And that's within 24 hours after he rules.

[00:40:15] Speaker 01:
having decided we hadn't waived initially, then deciding.

[00:40:19] Speaker 01:
The final view on the matter is that you waived.

[00:40:22] Speaker 01:
And Your Honor, that would be an abusive discretion.

[00:40:24] Speaker 01:
The idea that you find on Wednesday no one's waived, you then offer a claim construction.

[00:40:32] Speaker 00:
Well, it's only an abusive discretion fee is wrong, right?

[00:40:35] Speaker 00:
It's not an abusive discretion for him to change his mind.

[00:40:38] Speaker 01:
It's not an abusive discretion for him to change his mind, but it's an abusive discretion

[00:40:42] Speaker 01:
precisely because of the circumstances you described.

[00:40:46] Speaker 01:
So you get to the JMOL, and I think there's fair argument he's changed his mind.

[00:40:53] Speaker 01:
But the jury never got either of those two claim constructions.

[00:40:58] Speaker 00:
Well, if it had gotten the one that he settled on in the J-Mall, you wouldn't be here, at least on that issue.

[00:41:05] Speaker 00:
I mean, you clearly would have lost on that construction.

[00:41:10] Speaker 01:
If he had given that construction, what we would be here on is arguing to you that that claim construction was wrong.

[00:41:18] Speaker 01:
And I think that's why I think at the outset, and I think Judge Prost may have articulated it more clearly than me, there is a simple sufficiency of the evidence against the plain meaning.

[00:41:28] Speaker 01:
That does require, as the chief said, you decide what the plain meaning is.

[00:41:34] Speaker 01:
You have to decide what the plain meaning against.

[00:41:36] Speaker 01:
what the judge said in A144 to 145.

[00:41:39] Speaker 01:
You have to decide it against what we said was the plain meaning from the outset.

[00:41:43] Speaker 01:
And you have to consider what he said in the DMOL ruling.

[00:41:47] Speaker 01:
If the plain meaning is, as we say, a single instruction, then there is insufficient evidence to support the verdict, no matter what the waiver issues are.

[00:41:57] Speaker 01:
And that's why a fundamental question here is, in a patent that was in a crowded field of prior art,

[00:42:07] Speaker 01:
The IBM patent and the DEC patent were extremely close, as the district court said when it granted summary judgment of no willfulness.

[00:42:15] Speaker 01:
And you have a term particular that's in the specification 20 times and in the claim, and in a portion of the claim that has an antecedent basis.

[00:42:25] Speaker 01:
And the judge construes it to be consistent with our dictionary definitions.

[00:42:29] Speaker 01:
It says a single instruction.

[00:42:33] Speaker 01:
Can the verdict be sustained?

[00:42:34] Speaker 01:
we would suggest that the verdict cannot be sustained and should be reversed.

[00:42:38] Speaker 01:
At a minimum, we suggest a new trials report.

[00:42:40] Speaker 03:
OK, we'll reserve a little time for rebuttal, but thank you.

[00:42:44] Speaker 03:
Thank you.

[00:42:49] Speaker 05:
May it please the court, on behalf of the Wisconsin Alumni Research Foundation, Morgan Chu, and I'm here with my colleagues, Chris Abernathy to my immediate left and Gary Frischling to the far left.

[00:43:03] Speaker 05:
Let me discuss the particular issue first.

[00:43:07] Speaker 05:
There was a clear waiver, a clear waiver which must be assessed by abuse of discretion.

[00:43:17] Speaker 05:
The Apple request for a jury instruction or a claim construction using a single, instead of the words of the claim, the particular

[00:43:32] Speaker 03:
Well, let me ask you.

[00:43:32] Speaker 03:
I understand your argument.

[00:43:36] Speaker 03:
We understand your argument.

[00:43:37] Speaker 03:
But what about the point that let's assume this issue had never even come up and both parties adhered to plain and ordinary meaning and we get a jury verdict.

[00:43:49] Speaker 03:
Don't we have to then construe either whether your view of the plain and ordinary meaning or Mr. Lee's view of the plain and ordinary meaning

[00:43:58] Speaker 03:
Which of those is correct in order to evaluate the sufficiency of the evidence?

[00:44:03] Speaker 05:
No, if it came up exactly that way, then it's clear that both sides had waived any further construction.

[00:44:10] Speaker 03:
But how would we evaluate the evidence without ourselves understanding what the parameters of the claim are?

[00:44:15] Speaker 05:
Okay, let me do it with the words the particular, with the actual claim language.

[00:44:19] Speaker 05:
So it's a prediction associated with the particular load instruction.

[00:44:24] Speaker 05:
I'm using load instructions that are consumer instruction.

[00:44:28] Speaker 05:
99.98% of the time, there is that association.

[00:44:36] Speaker 05:
And there's no meaningful dispute on that.

[00:44:39] Speaker 05:
Dr. August basically agreed.

[00:44:42] Speaker 05:
Mr. Williams, an Apple engineer, agreed.

[00:44:45] Speaker 05:
Here's the point about the single argument today.

[00:44:50] Speaker 05:
After Wharf had wrested its case in chief and after Apple had wrested its case in chief during the liability portion of trial, it was after both parties had wrested its case in chief that Apple asked for the first time the use of the words a single

[00:45:10] Speaker 05:
to interpret the particular.

[00:45:12] Speaker 05:
And I can give you the particular sites.

[00:45:14] Speaker 00:
Well, but before you do that, I'm interested in the answer to Judge Prost's specific question also.

[00:45:22] Speaker 00:
And what I'm interested in is suppose contrary, and you're arguing the merits now, which I understand, but suppose that contrary to what happened, there had simply been an instruction, plain and ordinary meaning,

[00:45:36] Speaker 00:
And the argument made to us on appeal was insufficient evidence.

[00:45:43] Speaker 00:
What do we do?

[00:45:43] Speaker 00:
Do we say there is a range of possible plain and ordinary meanings, and we're not going to decide what plain and ordinary meaning means, and therefore if something fell within this range, we would sustain the juror's verdict?

[00:45:57] Speaker 00:
Or what?

[00:45:57] Speaker 00:
Do we construe the actual meaning at that point?

[00:46:00] Speaker 05:
As applied to the facts of this case, whether it's the words the particular or a single

[00:46:07] Speaker 05:
there is this undisputed fact that there is that association, prediction associated with the load.

[00:46:14] Speaker 00:
Yes, yes, I understand.

[00:46:15] Speaker 00:
But I'm looking for a more, to the more abstract question of what is our modus operandi here?

[00:46:24] Speaker 00:
What should we do by way of considering a situation in which there has been a plain and ordinary meaning, instruction and construction, and it comes up

[00:46:37] Speaker 00:
as a challenge to the sufficiency of the evidence.

[00:46:41] Speaker 05:
Let me focus on a way which we disagree with, that Apple could win, and that is if the construction, and it's not just one and only, but it's one and only every time.

[00:46:57] Speaker 05:
So if there's one time out of 1,000

[00:47:02] Speaker 05:
where it is not associated with one and only instruction, it's Apple's position they would win.

[00:47:09] Speaker 05:
They would lose under everyday meaning, any meaning that one wants to give that's reasonable, with the words of the particular, or the single, or even one and only, because that occurs 99.98% of the time.

[00:47:26] Speaker 05:
What occurs?

[00:47:27] Speaker 03:
That they use one and only instruction?

[00:47:30] Speaker 03:
No.

[00:47:32] Speaker 03:
Well, now we're getting into aliasing.

[00:47:36] Speaker 05:
No.

[00:47:36] Speaker 05:
The factual record is that there is a prediction associated with, I'll say, a single load instruction and only a single load instruction 99.98% of the time.

[00:47:56] Speaker 05:
And it's because of a number of things.

[00:48:00] Speaker 05:
There are only 192 entries.

[00:48:02] Speaker 05:
in the table.

[00:48:03] Speaker 05:
So talking about millions of instructions makes no sense.

[00:48:08] Speaker 05:
Hashing is irrelevant to this.

[00:48:10] Speaker 05:
So factually, there is always an association except for a tiny, tiny percentage of the times.

[00:48:18] Speaker 05:
So for 99.98% of the time, there's infringement

[00:48:22] Speaker 03:
Can you give us your best record sites for what witness testified?

[00:48:27] Speaker 03:
Oh, sure.

[00:48:28] Speaker 03:
Sure.

[00:48:29] Speaker 05:
First of all, let me give you Apple's Dr. August testimony on this.

[00:48:36] Speaker 05:
It's at Appendix 2239.

[00:48:37] Speaker 05:
The chances in which a prediction will be associated with more than one load instruction is 0.0-0.1%.

[00:48:51] Speaker 02:
Okay, now that's the testimony that I referred Mr. Lee to and he said, well, that's irrelevant because that's talking about aliasing.

[00:49:01] Speaker 02:
What's your response to that?

[00:49:03] Speaker 05:
Aliasing just means when there's association with two load instructions.

[00:49:10] Speaker 05:
There was no dispute about what aliasing means.

[00:49:14] Speaker 05:
And then there's the clear testimony by Dr. Conti

[00:49:18] Speaker 05:
about the 99.98% of the time.

[00:49:21] Speaker 05:
It's discussed at length in our brief.

[00:49:24] Speaker 05:
There's also a reference to the Apple engineer, Mr. Williams, who testified to the effect that it was less than 0.1%.

[00:49:33] Speaker 05:
So that would be less than 0.001 times.

[00:49:39] Speaker 05:
So it's not largely in dispute that well over 99% of the time, there is an association with one and only one.

[00:49:50] Speaker 03:
Now, one other point.

[00:49:51] Speaker 03:
But the testimony here, and even in your closing argument where you were talking about in 99.1% of the time, you were talking about aliasing, right?

[00:50:02] Speaker 05:
Aliasing is when there's an association with more than one load instruction.

[00:50:10] Speaker 05:
So the answer is yes, but it's the same as saying that there is not an association with more than one

[00:50:19] Speaker 05:
load instruction 99.98% of the time.

[00:50:28] Speaker 05:
And everyone agreed at trial.

[00:50:30] Speaker 03:
Where is the linking between aliasing and the one load instruction?

[00:50:39] Speaker 03:
Because in 2240, and is this Dr. August?

[00:50:44] Speaker 03:
And so if the jury is to credit this testimony, that would mean that for 99.9% of the time, the load, the prediction and entry of the table is associated with one load instruction, correct?

[00:50:56] Speaker 03:
And the answer is no.

[00:50:59] Speaker 03:
If the jury were to credit this, it goes on.

[00:51:01] Speaker 03:
And the witness is saying no to those questions, not yes.

[00:51:05] Speaker 03:
So I'm assuming I'm missing something.

[00:51:08] Speaker 05:
So I want you to tell me what it is.

[00:51:10] Speaker 05:
So Dr. Conti testified

[00:51:12] Speaker 03:
Well, I thought you were referring us to Dr. August's testimony.

[00:51:16] Speaker 05:
I did.

[00:51:18] Speaker 05:
Dr. August is Apple's expert.

[00:51:21] Speaker 05:
So he did fight on this in the sense that he was trying to picture it differently.

[00:51:27] Speaker 05:
So let me give you the setting, and then I'll give you Dr. Conti's site.

[00:51:30] Speaker 05:
The way to look at this is, was there substantial evidence?

[00:51:34] Speaker 05:
There was a big clash, largely between Dr. Conti for Wharf and Dr. August for Apple.

[00:51:40] Speaker 05:
They relied on lots of documentation.

[00:51:43] Speaker 05:
Significant evidence was put before the jury.

[00:51:46] Speaker 05:
And so both sides had a factual clash on these issues.

[00:51:52] Speaker 05:
And the jury ruled in favor of Whorf.

[00:51:54] Speaker 05:
So it ends up being a substantial evidence question.

[00:51:57] Speaker 05:
Now let me turn to some sites of what Dr. Conte said.

[00:52:02] Speaker 05:
Dr. Conte had said, based on his own calculations, Apple- Why don't you give me the page number?

[00:52:08] Speaker 05:
I can follow you through.

[00:52:10] Speaker 05:
If I say A, that means appendix, A 2517 to 2515 to 2517.

[00:52:20] Speaker 05:
And in particular, at 2517, starting at line 17, he explains, Dr. Conti, it's greater than a 99.9% chance that

[00:52:39] Speaker 05:
that the piece of information you use is going to be unique.

[00:52:45] Speaker 05:
That's at 2517.

[00:52:47] Speaker 05:
Those are just examples.

[00:52:49] Speaker 03:
I know.

[00:52:50] Speaker 03:
But actually, I think Mr. Lee responds to this in the briefing that maybe I've got the wrong thing.

[00:52:54] Speaker 03:
But if you keep reading, that seems to be explained or qualified to mean something different than if you just look at it in isolation.

[00:53:03] Speaker 03:
And indeed, what Apple does, the Apple engineers are smart.

[00:53:06] Speaker 03:
They work real hard to come up with good nicknames.

[00:53:08] Speaker 03:
They wouldn't use a zip code.

[00:53:09] Speaker 03:
Why would they use a zip code?

[00:53:11] Speaker 03:
They used something like a street address.

[00:53:13] Speaker 03:
That was that uniquely identified for 99.98% of the time.

[00:53:19] Speaker 05:
The zip code was an analogy that Dr. August used.

[00:53:22] Speaker 05:
I think it makes a lot more sense to see how the system actually operated.

[00:53:29] Speaker 05:
And it operated, I'll round it off, far more than 99% of the time.

[00:53:36] Speaker 05:
The prediction was associated.

[00:53:38] Speaker 05:
as Dr. Conti said, with a unique tag or with one load instruction.

[00:53:44] Speaker 05:
And there are many, many other citations in the record about that.

[00:53:51] Speaker 00:
But as I understand, Apple's argument is that that may be so in the operation of the system that as a given load instruction comes in that

[00:54:02] Speaker 00:
It's only that load instruction that is the subject at that moment of the prediction.

[00:54:08] Speaker 00:
However, the prediction applies not just to that load instruction, but to every load instruction that is covered through hashing by the same 12-bit address or truncated address.

[00:54:23] Speaker 00:
So why is that not an appropriate way to look at the question of whether this

[00:54:30] Speaker 00:
association applies to more than just one in the typical case instruction.

[00:54:37] Speaker 05:
There are two reasons for that.

[00:54:40] Speaker 05:
First, in the patent itself, the 752 patent at column 6, beginning about line 61, there is a discussion about using a partial address, which basically is what hashing is doing.

[00:54:56] Speaker 05:
It's dropping off part of the address.

[00:55:00] Speaker 05:
similar to what we do with patent numbers, right?

[00:55:03] Speaker 05:
We drop the first four digits, we just use the last three.

[00:55:07] Speaker 05:
Okay.

[00:55:08] Speaker 05:
It says, referring now to figure one, an ILP processor suitable for use with the present invention includes a memory, having a portion holding a stored program at a plurality of physical addresses, here depicted as XX1 to XX6, where the values XX

[00:55:30] Speaker 05:
indicate some higher ordered address bits that may be ignored in this example.

[00:55:38] Speaker 05:
So they are discussing doing it the same way.

[00:55:42] Speaker 05:
This is a discussion of hashing.

[00:55:44] Speaker 05:
So it's like dropping the higher order bits for a patent number or the first four digits.

[00:55:52] Speaker 00:
Point number one.

[00:55:53] Speaker 00:
Is this really a discussion of hashing, or is it just truncating the addresses for purposes of

[00:56:00] Speaker 00:
the discussion, i.e.

[00:56:03] Speaker 00:
just not having some very, very long address for purposes of explanation of how the invention works.

[00:56:10] Speaker 05:
No, this is a discussion when you drop the higher order bits.

[00:56:15] Speaker 05:
That's exactly what's done in hashing to make the system more efficient.

[00:56:21] Speaker 05:
And the goal is not to alias or to have an association with two or more load instructions.

[00:56:27] Speaker 05:
The goal is

[00:56:29] Speaker 05:
for the largest possible percentage, never to alias or not to alias.

[00:56:36] Speaker 05:
That's what the goal is.

[00:56:38] Speaker 05:
Go ahead.

[00:56:39] Speaker 05:
I'm sorry.

[00:56:41] Speaker 05:
So if you look at the patent, that's answer number one.

[00:56:44] Speaker 05:
Answer number two is the fact that there is this clash.

[00:56:50] Speaker 05:
In other words, Dr. August made similar arguments to the jury saying, well, somehow there is not this

[00:56:57] Speaker 05:
Prediction Association because there's hashing.

[00:57:01] Speaker 05:
Dr. Conti disagreed.

[00:57:04] Speaker 05:
They had extensive testimony about it.

[00:57:07] Speaker 05:
There was argument about it to the jury in closing argument.

[00:57:11] Speaker 05:
And the jury came out in favor of wharf on this factual dispute.

[00:57:17] Speaker 00:
Let me ask you about something in the specification that Apple highlights.

[00:57:22] Speaker 00:
Yep.

[00:57:22] Speaker 00:
On column 11, at the end

[00:57:25] Speaker 00:
first few lines of column 11, there's a reference to the, well, the reviewing, the predictor circuit reviews a prediction table shown generally in figure five to see if the, and the critical language then is, particular instruction identified by its physical address is in the prediction table.

[00:57:50] Speaker 00:
Why is the reference to physical address

[00:57:54] Speaker 00:
not an indication that the particular instruction is the whole single instruction, the unique instruction with that physical address.

[00:58:05] Speaker 00:
It could be, but this doesn't define the claims.

[00:58:09] Speaker 00:
In other words... But the term particular instruction shows up in the claims, and so why shouldn't we assume that that term is being used in the same way?

[00:58:17] Speaker 05:
Because it's also used in the context of the column six, where it's plainly discussing a situation

[00:58:24] Speaker 05:
where the full particular address is not there.

[00:58:29] Speaker 05:
Second, if you look at the claim language, there's nothing that says in the claim language that you have to have the full particular instruction address.

[00:58:39] Speaker 05:
I'm using the word full, right, because it's just an association of a prediction with either the particular load instruction or

[00:58:54] Speaker 05:
as Apple would have it, a single loading structure.

[00:58:58] Speaker 00:
But the term, I think, in my understanding at least, you tell me if this is wrong.

[00:59:03] Speaker 00:
But the term physical address typically refers to the full address, not some truncated form.

[00:59:09] Speaker 00:
Is that correct?

[00:59:10] Speaker 05:
I don't know one way or the other.

[00:59:13] Speaker 05:
I just don't know one way or the other.

[00:59:15] Speaker 05:
Because hashing functions are used throughout programming frequently, not just in this context.

[00:59:24] Speaker 02:
Is Dr. Conti saying that there are actual load tags that have only one instruction?

[00:59:32] Speaker 05:
Yes.

[00:59:33] Speaker 02:
And other than him simply saying that, what is the other evidence that would indicate that?

[00:59:41] Speaker 05:
Oh, he went through the code, the detector circuit.

[00:59:47] Speaker 05:
The detector circuit is made by what they call an RTL code.

[00:59:54] Speaker 05:
It is a hardware rendition, because these are hardware claims.

[00:59:59] Speaker 05:
But it's similar to what I think the members of the panel will understand source code to be.

[01:00:04] Speaker 05:
It's the exact technical way in which the circuit operates.

[01:00:11] Speaker 05:
And he discussed that in great detail before the jury on this and a variety of other issues.

[01:00:17] Speaker 02:
Well, you say he discussed it in great detail.

[01:00:20] Speaker 02:
And I'm having a hard time finding the precise

[01:00:24] Speaker 02:
places where you think he most clearly says it.

[01:00:26] Speaker 02:
He says in one place, Apple takes the load information and it produces one tag.

[01:00:31] Speaker 02:
You take in that load, you produce that tag for that load.

[01:00:34] Speaker 02:
So are you saying that they're saying that every single load only has one tag?

[01:00:42] Speaker 05:
No.

[01:00:42] Speaker 05:
I am saying an infinitesimally small number, less than 1%, can have more than one tag.

[01:00:52] Speaker 05:
And all that would mean

[01:00:54] Speaker 05:
is that 99.98% of the time there's infringement.

[01:00:59] Speaker 00:
Well, but I think Judge O'Malley's question, and I had the same question, is a slightly different one, which is, are there programs in which every single load instruction has its own load tag?

[01:01:12] Speaker 00:
Now, you make that argument at page 17 of your brief in the first full paragraph, but you don't say that there's any affirmative

[01:01:22] Speaker 00:
evidence to that effect.

[01:01:23] Speaker 00:
What you say is Apple has not contended that every program has multiple loads associated with every load tag, which is reversing the burden of proof.

[01:01:32] Speaker 00:
So I wonder if there is evidence that there are programs that would infringe, according to you, because even under their construction, because every single load instruction has its own load tag.

[01:01:48] Speaker 05:
First of all,

[01:01:50] Speaker 05:
I think I need to just go back to the 99.98 if I'm understanding the question.

[01:01:55] Speaker 00:
Maybe I'm misunderstanding the question.

[01:01:56] Speaker 00:
Suppose you have a very tiny program that only has 300 load instructions associated with it.

[01:02:04] Speaker 00:
And every single one of those 300 load instructions has its own load tag.

[01:02:11] Speaker 00:
That program would infringe even under Apple's construction, correct?

[01:02:17] Speaker 00:
Yes.

[01:02:18] Speaker 00:
And I think Judge O'Malley's question in mind, too, is whether what you mean to be saying in this paragraph on page 17 of your brief is that you have evidence to that effect, that there are such programs and that they infringe.

[01:02:33] Speaker 05:
I don't think there was any evidence of this very small type of hypothetical program where there is a unique load tag under all circumstances for every load instruction.

[01:02:48] Speaker 05:
OK?

[01:02:50] Speaker 05:
So go ahead.

[01:02:52] Speaker 03:
If you had something with particular before we exceed time more than we want to, I wanted to make sure you reached the vicarious liability questions that were raised.

[01:03:03] Speaker 03:
Let me then go there.

[01:03:06] Speaker 05:
OK.

[01:03:09] Speaker 03:
So on vicarious liability, if you're really ready to turn to that, I guess there are two parts.

[01:03:14] Speaker 03:
And the first part is really an evidentiary map.

[01:03:16] Speaker 03:
where Mr. Lee pointed us to the testimony which was kind of rejected or not allowed with respect to the witness saying this is what was happening because he wasn't there for the testing.

[01:03:30] Speaker 03:
So do you agree with what Mr. Lee said and do you have more evidence or other evidence or other testimony in that regard?

[01:03:37] Speaker 05:
I respectfully disagree with what Mr. Lee said.

[01:03:40] Speaker 03:
So that is the testimony.

[01:03:42] Speaker 05:
Well, I'll go to the exact testimony.

[01:03:44] Speaker 05:
The fact that a particular question was objected to and was sustained doesn't tell us what actually came in on the record.

[01:03:51] Speaker 05:
OK.

[01:03:52] Speaker 05:
Here at A, 3924 to 3925.

[01:04:02] Speaker 05:
Question, what's your basis for concluding that all the circuitry and functionality is present in the chip after printing?

[01:04:12] Speaker 05:
Objection.

[01:04:13] Speaker 05:
followed by colloquy, and then the witness is told by the court, you can proceed.

[01:04:21] Speaker 05:
Dr. Conti then says, quote, so there is a test for it.

[01:04:24] Speaker 05:
It's called JTAG.

[01:04:26] Speaker 05:
JTAG is an IEEE standard.

[01:04:29] Speaker 05:
It's IEEE standard 1149.

[01:04:32] Speaker 05:
My vice president of standards recertifies this every two years, and you use it to test the chips, to make sure they work, to make sure they're capable.

[01:04:43] Speaker 05:
You do that on every chip.

[01:04:45] Speaker 05:
A couple things to note.

[01:04:48] Speaker 05:
He was president of the IEEE.

[01:04:50] Speaker 05:
That's why he's referring to my vice president, so he's personally familiar with these standards.

[01:04:55] Speaker 05:
There is no dispute that the wafers manufactured in Boston, Texas, in fact, have these test ports.

[01:05:05] Speaker 05:
And Dr. Conte has other testimony.

[01:05:08] Speaker 05:
that relates to this.

[01:05:10] Speaker 02:
Let's go to... Okay, so he concedes he didn't do any testing himself, and so the judge says you can't talk about personal testing, but then he says these all have test ports, which means they're capable of being tested, and he says that the IEEE standard tells you how to test and all chips have to be tested.

[01:05:37] Speaker 02:
Is there any basis for him to say that that testing actually occurred in the United States?

[01:05:44] Speaker 05:
Since he wasn't on the premises to see the test, but the important point is the claim construction focused on whether the infringing circuits were capable of.

[01:05:59] Speaker 05:
That's the claim construction, and it was without objection by Apple.

[01:06:05] Speaker 05:
So naturally, Dr. Conti's thrust of his testimony on this subject was capable of.

[01:06:13] Speaker 05:
And the main apple argument was that you need to have bumping and fusing of various kinds.

[01:06:24] Speaker 05:
If I go to the end of the page I was just reading from, 3424, there's a question onto 3925.

[01:06:35] Speaker 05:
And does that require the blowing of the fuses?

[01:06:38] Speaker 05:
Answer, no.

[01:06:40] Speaker 05:
Now let me go to other Dr. Conte testimony.

[01:06:44] Speaker 05:
This is now at 2959.

[01:06:47] Speaker 05:
By the time you're away for, all the claimed circuits are here.

[01:06:52] Speaker 05:
They're already built.

[01:06:54] Speaker 05:
It's like you build a house.

[01:06:55] Speaker 00:
I don't think we have 2959.

[01:06:56] Speaker 00:
Is it 2959?

[01:06:59] Speaker 00:
I'm sorry.

[01:07:00] Speaker 00:
I was looking at 39.

[01:07:01] Speaker 05:
Oh, I may have misspoke.

[01:07:03] Speaker 05:
29.59.

[01:07:03] Speaker 00:
29.59.

[01:07:05] Speaker 00:
OK.

[01:07:05] Speaker 00:
No, you didn't.

[01:07:06] Speaker 00:
I misheard.

[01:07:08] Speaker 00:
OK.

[01:07:08] Speaker 00:
Go ahead.

[01:07:09] Speaker 05:
OK.

[01:07:09] Speaker 05:
By the time you have a wafer, all the claim circuits are here.

[01:07:13] Speaker 05:
39.24.

[01:07:14] Speaker 05:
39.29.

[01:07:25] Speaker 05:
Question.

[01:07:26] Speaker 05:
Where is the functionality created and existing for the Austin chips?

[01:07:32] Speaker 05:
Answer, it's created in Austin.

[01:07:34] Speaker 05:
Question, and does that have anything to do with when fuses are blown or when bumping occurs?

[01:07:40] Speaker 05:
Answer, no, no.

[01:07:42] Speaker 05:
Question, what does it have to do with?

[01:07:44] Speaker 05:
Answer, it has to do with the circuits that are laid down in Austin.

[01:07:50] Speaker 05:
Going to 2995, and this is on cross-examination.

[01:07:55] Speaker 05:
Question.

[01:07:56] Speaker 05:
And before bumping happens... 29.95.

[01:07:59] Speaker 05:
Excuse me.

[01:08:03] Speaker 05:
29.95.

[01:08:05] Speaker 05:
Question.

[01:08:06] Speaker 05:
And before bumping happens, you can't apply power to the circuits right?

[01:08:11] Speaker 05:
Answer, actually you can.

[01:08:14] Speaker 05:
But you can do that on a test fixture.

[01:08:17] Speaker 05:
29.96.

[01:08:19] Speaker 05:
Question.

[01:08:20] Speaker 05:
Still on cross.

[01:08:21] Speaker 05:
And it's correct that bumps are what allow the

[01:08:24] Speaker 05:
Power to the circuit's right answer.

[01:08:27] Speaker 05:
As I said, you can do testing without having to bump first.

[01:08:35] Speaker 05:
That's just a sampling of the testimony.

[01:08:38] Speaker 05:
He is demonstrating in those few small examples the capability.

[01:08:45] Speaker 02:
This was... Can you just explain to me... But wait, what about the next... I'm sorry, what about the next two questions?

[01:08:50] Speaker 02:
He says, you have no evidence that any such testing has been performed in the United States, right?

[01:08:55] Speaker 02:
And he says, I don't know one way or the other.

[01:08:57] Speaker 05:
Yes.

[01:08:58] Speaker 05:
Yes.

[01:08:59] Speaker 05:
So, well, he answered correctly there.

[01:09:04] Speaker 05:
But actually, proving testing wasn't a necessity.

[01:09:10] Speaker 05:
The necessity was proving capability.

[01:09:13] Speaker 05:
And the A7 had about three billion transistors.

[01:09:19] Speaker 05:
Every single one of them was laid down in Austin.

[01:09:23] Speaker 05:
They could be electrically tested.

[01:09:26] Speaker 05:
That's the purpose of the JTAG port.

[01:09:30] Speaker 05:
And he did provide background on what normally happens.

[01:09:37] Speaker 05:
Doesn't go directly to capability because the capability goes to the RTL coded circuits that are laid down

[01:09:49] Speaker 05:
in Austin.

[01:09:52] Speaker 05:
All right with that?

[01:09:54] Speaker 05:
I'm actually not.

[01:09:57] Speaker 03:
Why are you saying that there's a lot of emphasis on that there were these test ports in Austin?

[01:10:05] Speaker 03:
Are you saying that because we're dealing with capability, the capability is required irrespective of whether or not anything was ever tested in Austin or anything was ever done in Austin?

[01:10:16] Speaker 05:
I would say it's

[01:10:17] Speaker 05:
differently, and let me try and answer your question directly.

[01:10:21] Speaker 05:
On the capability question, the question is, what was done in Austin, did it make the three billion transistor circuits capable?

[01:10:31] Speaker 05:
The answer is yes.

[01:10:33] Speaker 05:
That's what Dr. Conti testified to about all the circuits being made and manufactured.

[01:10:40] Speaker 05:
No other circuits for the chips were added later, period.

[01:10:45] Speaker 05:
So then Apple made the argument, well, you cannot electrically test them.

[01:10:51] Speaker 05:
And that's an answer to your question, Your Honor.

[01:10:55] Speaker 05:
That's where Dr. Conte says, what are you talking about?

[01:10:59] Speaker 05:
There are these test ports.

[01:11:01] Speaker 05:
You can electrically test them.

[01:11:03] Speaker 05:
And generally, you do electrically test them.

[01:11:05] Speaker 05:
But he also says, I wasn't a witness to the actual electrical testing.

[01:11:09] Speaker 05:
But that is off the point.

[01:11:12] Speaker 05:
We're on the question of whether the circuits

[01:11:15] Speaker 05:
are capable, who is more responding to an Apple argument.

[01:11:20] Speaker 05:
I hope I clarified that.

[01:11:22] Speaker 02:
So he's saying that the circuits are capable, that they're capable without bumping?

[01:11:29] Speaker 05:
Yes.

[01:11:31] Speaker 05:
Yes, or fusing.

[01:11:32] Speaker 05:
Correct.

[01:11:33] Speaker 05:
To bump.

[01:11:34] Speaker 00:
So maybe I... Capable if a testing device is used with them, a specific testing device that's... No.

[01:11:41] Speaker 00:
Actually, they're capable...

[01:11:44] Speaker 00:
OK, so.

[01:11:45] Speaker 00:
Well, but they don't work unless you have the particular testing device to see if they are capable, right?

[01:11:58] Speaker 00:
I mean, they won't work without the testing device?

[01:12:00] Speaker 00:
No.

[01:12:01] Speaker 05:
The question is, we're talking about just a portion of these chips, the LSD, the load store detector.

[01:12:11] Speaker 05:
And those circuits, as they are laid down,

[01:12:14] Speaker 05:
Are they capable of doing the load store detection?

[01:12:19] Speaker 05:
That's the question.

[01:12:20] Speaker 05:
They do not have to be electrically tested in order to do that.

[01:12:27] Speaker 05:
Because the client construction is just capable.

[01:12:30] Speaker 05:
Again, it was undisputed, no objection to it.

[01:12:34] Speaker 05:
So the main clash is whether they're capable of.

[01:12:39] Speaker 05:
One witness says yes, the other says no.

[01:12:44] Speaker 05:
So let me give you an example.

[01:12:45] Speaker 00:
Let me ask it this way, the question that's bothering me.

[01:12:49] Speaker 00:
If you take the wafers without having any testing equipment attached to them, presumably they can't be made to work in the form that they are.

[01:13:05] Speaker 00:
They're just not going to work unless you hook them up to a testing machine.

[01:13:09] Speaker 00:
So in that sense, before you hook them up to the testing machine, they're not capable.

[01:13:15] Speaker 00:
They're capable only if you put the rest of the components together with them that allows them to perform the way they are intended to perform, right?

[01:13:26] Speaker 05:
This court has actually addressed it in a couple cases.

[01:13:29] Speaker 05:
And hopefully you'll find this to be a direct answer.

[01:13:33] Speaker 05:
In silicon graphics versus ATI, cited in the briefs, it involved a rasterization circuit.

[01:13:43] Speaker 05:
The district court said that circuit does not infringe because you need to add, I think it was an operating system for the rasterization circuit to rasterize.

[01:13:58] Speaker 05:
This court reversed the district court and said, no, it was capable of rasterizing.

[01:14:05] Speaker 05:
If I were to simplify it, let's say one has an allegedly infringing microwave oven.

[01:14:14] Speaker 05:
It's got an electrical power cord.

[01:14:16] Speaker 05:
Is it capable of warming up food?

[01:14:19] Speaker 05:
Yes, it is.

[01:14:21] Speaker 05:
I could say, yes, it is.

[01:14:22] Speaker 05:
In the box, it comes.

[01:14:24] Speaker 05:
It's got the electrical cord.

[01:14:25] Speaker 05:
But in fact, if you want to look at it from another perspective, you in fact have to plug it into the outlet.

[01:14:31] Speaker 05:
So the silicon graphics case stands for the proposition that even if something else needed to be done, it is still capable of the same holding

[01:14:43] Speaker 05:
effectively is in this court's fingen versus secure computing.

[01:14:50] Speaker 02:
So there's the silicon graphics fingen line of cases, and then there's a line, obviously, to where you get to the Nizomi case.

[01:14:59] Speaker 02:
So how do you distinguish your circumstances from the circumstances in Nizomi?

[01:15:05] Speaker 05:
If you're referring to the Supreme Court case on the fear of housing, or you

[01:15:11] Speaker 05:
referring to something?

[01:15:13] Speaker 02:
Referring to our case, though.

[01:15:14] Speaker 05:
OK.

[01:15:16] Speaker 02:
That talks about how you can define what the testing will tell you with respect to functionality.

[01:15:25] Speaker 05:
Well, again, there are two answers.

[01:15:28] Speaker 05:
There's no requirement to test when the claim construction is capable of.

[01:15:35] Speaker 05:
As one of this Court's decision

[01:15:40] Speaker 05:
referred to an automobile that had a propulsion system in it.

[01:15:46] Speaker 03:
Can we stick to silicon graphics?

[01:15:48] Speaker 03:
Sure.

[01:15:48] Speaker 03:
That dealt with just adding, activating.

[01:15:52] Speaker 03:
You're simply activating.

[01:15:53] Speaker 03:
Whereas I think Apple's position here, and I don't know if you would disagree, there's testimony to support it, that what they're talking about doing in Korea, which is fusing and bumping, that's not simply activating it.

[01:16:06] Speaker 03:
It's making it functional.

[01:16:08] Speaker 03:
And if that were

[01:16:10] Speaker 03:
true, then I think you would agree silicon graphics would be distinguishable, or not.

[01:16:17] Speaker 05:
Well, first, we, with respect, disagree with Apple's position, and I don't think it would make it distinguishable.

[01:16:26] Speaker 05:
In other words, the rasterization circuit had no possibility, with a test port or not, to operate, and the question was whether it was capable.

[01:16:39] Speaker 05:
Was it capable of rasterizing?

[01:16:41] Speaker 05:
This court said yes, even though, and I think it was an operating system, it was an addict.

[01:16:47] Speaker 05:
And FinGen was to the same effect.

[01:16:49] Speaker 05:
One had to get an activation code to activate it.

[01:16:54] Speaker 05:
And this court said, no, there's infringement.

[01:16:57] Speaker 02:
Now let me go to the- But there's a difference between just activating and understanding whether the functionality is all there, right?

[01:17:07] Speaker 02:
And isn't that why we needed more data with respect to what the testing would show and what happens when that JTAC port is accessed?

[01:17:20] Speaker 05:
Again, there's no need to access it to know that the circuits are capable of.

[01:17:29] Speaker 05:
If the claim construction said that there had to be some electrical test

[01:17:37] Speaker 05:
Yes, that would cause a different outcome.

[01:17:40] Speaker 05:
But that's not what the claim construction was.

[01:17:45] Speaker 05:
So if I can go to the bumping and fusing for a moment.

[01:17:51] Speaker 05:
So what are they involved?

[01:17:55] Speaker 05:
The chips are put in a package, and then they need electrical connections to the rest of the product.

[01:18:01] Speaker 05:
That's what bumping is.

[01:18:02] Speaker 05:
It's referring to solder balls where

[01:18:05] Speaker 05:
there's an electrical connection to the rest of the product.

[01:18:09] Speaker 05:
We have to keep in mind this is not the operation of the chip that in French it's capable of.

[01:18:17] Speaker 05:
The fusing is done for a number, a couple different reasons.

[01:18:24] Speaker 05:
And what it is, it's like a little hardware switch or switches.

[01:18:30] Speaker 05:
And Apple uses these as an example to

[01:18:35] Speaker 05:
have security for software purposes, because Apple doesn't want its iOS operating system to be used on processors that are made by someone else.

[01:18:48] Speaker 05:
So these switches are pulled that said, OK, now it can be used in its intended environment, which is the Apple iPhone.

[01:18:59] Speaker 05:
So it's very similar to the rasterization circuit

[01:19:04] Speaker 05:
and what's going on in FinGen.

[01:19:07] Speaker 05:
I have some dated August testimony about this.

[01:19:12] Speaker 05:
And this is at 3643 at line 10.

[01:19:18] Speaker 05:
Question, and with respect to the fuses you mentioned, what is the importance of blowing the fuses?

[01:19:24] Speaker 05:
Answer by Dr. August.

[01:19:26] Speaker 05:
Well, there are actually components in the microprocessor in the dye that aren't fully configured.

[01:19:32] Speaker 05:
There are some measurements that need to be taken to figure out, for example, how to interact with cache memory.

[01:19:39] Speaker 05:
And once those measurements are taken, the fuses can be customized and blown for that particular die.

[01:19:45] Speaker 05:
And at that point, we'll actually be able to communicate with memory.

[01:19:50] Speaker 05:
So that actually activates part of the circuitry or completes some of the circuitry before that's done.

[01:19:57] Speaker 05:
So what he's basically saying, similar to... Before that's done, there's no it's not function.

[01:20:03] Speaker 05:
Well, he does say that, but Dr. Conti's testimony is exactly the opposite.

[01:20:11] Speaker 05:
I'm pointing out that at least his testimony, Dr. August's testimony, is that this is activation very similar to what's happening in silicon graphics and in Fingent.

[01:20:28] Speaker 05:
The main point is all of the operational circuitry for which there was no dispute.

[01:20:34] Speaker 05:
was all manufactured, put on the wafers in Austin, and at no other place in the world were any of the circuits for the entire chip put on, much less the circuits for the LSD, which was the infringing component.

[01:20:54] Speaker 03:
So your position does not necessarily need any reliance on the J-port testing?

[01:21:00] Speaker 05:
Correct.

[01:21:03] Speaker 05:
Case was tried on the capable of jury instruction.

[01:21:08] Speaker 05:
That's point number one.

[01:21:10] Speaker 05:
And point number two on that, you had significant evidence being proffered by each side with different conclusions drawn by the two experts, opposing arguments by both sides.

[01:21:24] Speaker 05:
And there was substantial evidence to support the jury's verdict.

[01:21:30] Speaker 03:
I didn't think it was possible, but we're almost getting close to the time we took with Mr. Lee.

[01:21:35] Speaker 03:
If you want to ask a factual question, just to be sure.

[01:21:41] Speaker 00:
This is going to be a one-word answer.

[01:21:44] Speaker 03:
No, no, I want you to go on.

[01:21:47] Speaker 00:
Yes, yes, yes, but I did just very quickly.

[01:21:51] Speaker 00:
There were allusions in the brief that I wasn't sure of indicating that perhaps

[01:21:56] Speaker 00:
The chips from the Austin plant were not all of the chips that issue in the case, but I couldn't figure out whether that's true or not.

[01:22:03] Speaker 05:
That's correct.

[01:22:04] Speaker 00:
What percentage of the chips involved in the case are the Austin chips that are affected by this damages question?

[01:22:13] Speaker 00:
By this question.

[01:22:16] Speaker 05:
Roughly 40%.

[01:22:18] Speaker 00:
OK, that's fine.

[01:22:19] Speaker 05:
Thank you.

[01:22:19] Speaker 05:
And you don't need an explanation as to where the other chips came from?

[01:22:23] Speaker 05:
No, no.

[01:22:24] Speaker 05:
OK, then I'll go on.

[01:22:25] Speaker 05:
There was another piece to our discussion with Mr. Lee on vicarious liability.

[01:22:29] Speaker 03:
That was more of a legal question.

[01:22:31] Speaker 05:
Yes.

[01:22:32] Speaker 05:
So the question on Steeley and the summary judgment motion of anticipation.

[01:22:39] Speaker 02:
Well, first, I think Judge Cross was trying to get you to the point of, in Akamai, were we interpreting 271A for all purposes, or were we only

[01:22:50] Speaker 02:
limiting our interpretation of 271A for circumstances involving method claims?

[01:22:59] Speaker 05:
I think the first reading was that when 271A was discussed in the Yakamae opinion, it was a discussion of vicarious liability.

[01:23:10] Speaker 05:
And the court basically said control and direction is applicable.

[01:23:18] Speaker 05:
And control and direction can be found by prior cases in two ways.

[01:23:23] Speaker 05:
Number one, agency, or number two, by a contractual relationship.

[01:23:30] Speaker 05:
And then the actual decision in ACMI says there's another third way.

[01:23:35] Speaker 02:
We know what we said about how you can show attribution, since we said vicarious liability is not the right phrasing.

[01:23:43] Speaker 02:
We know what we said there.

[01:23:44] Speaker 02:
The question is, did we say it?

[01:23:46] Speaker 02:
all 270 for all infringement analyses or only for method claims?

[01:23:52] Speaker 05:
I believe, looking at the language, the fairest reading, because there was no limitation to it, was that it covered all claims.

[01:24:01] Speaker 05:
And there's another Federal Circuit case, centillion data systems versus Quest.

[01:24:08] Speaker 05:
And in the centillion system, Quest, the alleged infringer, argued

[01:24:15] Speaker 05:
Pre-existing law only applies to method claims, not to system or apparatus claims.

[01:24:22] Speaker 05:
And the Centillion Federal Circuit decision said, no, we reject Quest's argument here so that we can apply the law and apply it, as I said earlier, in connection with the Akamai

[01:24:42] Speaker 05:
decision.

[01:24:43] Speaker 05:
The order was sentillian was first.

[01:24:46] Speaker 05:
Akamai came later.

[01:24:48] Speaker 05:
So it had already been settled that the law for vicarious liability could be applied to system and method claims.

[01:25:01] Speaker 02:
So talk about Steeley.

[01:25:05] Speaker 05:
So I'll just focus on the claim construction aspect.

[01:25:10] Speaker 05:
unless there are going to be questions about other things.

[01:25:13] Speaker 05:
We start with the claim language itself.

[01:25:16] Speaker 05:
And there's a very detailed decision by Judge Connolly.

[01:25:21] Speaker 02:
This is the static versus dynamic?

[01:25:23] Speaker 05:
Yes.

[01:25:24] Speaker 05:
Predictions?

[01:25:24] Speaker 05:
Yes.

[01:25:25] Speaker 05:
OK.

[01:25:26] Speaker 05:
So first point, if you look at the claims themselves, the last limitation of claim one

[01:25:34] Speaker 05:
refers to a prediction threshold detector preventing data speculations for instructions.

[01:25:41] Speaker 00:
Where are you exactly?

[01:25:43] Speaker 05:
Claim one, column 14, line 50.

[01:25:47] Speaker 05:
So it's subparagraph V of claim one.

[01:25:50] Speaker 05:
I'll wait till everyone gets there.

[01:25:52] Speaker 05:
Yeah, OK.

[01:25:53] Speaker 05:
OK.

[01:25:54] Speaker 05:
A prediction threshold detector preventing data speculations for instructions having a prediction

[01:26:01] Speaker 05:
quote, within a predetermined range, end quote.

[01:26:06] Speaker 05:
And predetermined range would make no sense unless there is some updating or unless it's dynamic, because all Steely discusses is having a flag when there's been a misspeculation.

[01:26:22] Speaker 05:
So we have this predetermined range, language, and the claims themselves.

[01:26:28] Speaker 05:
Elsewhere in the intrinsic evidence, the specification, there's a discussion about an object of the invention.

[01:26:38] Speaker 05:
And this is in column 4 at about line 31, where it reads in part that an, quote, object of the invention to provide a predictor circuit that may identify

[01:26:58] Speaker 05:
data dependencies on an ongoing or dynamic basis.

[01:27:04] Speaker 05:
So it's saying that an object of the invention is to be dynamic, starting at about line 67.

[01:27:12] Speaker 02:
Maybe I'm wrong, but I understood Apple's argument to be that even if we accept that claim construction, even if static and dynamic are both encompassed, that Steely still doesn't, or at least

[01:27:29] Speaker 02:
that Steely still doesn't explain the tag comparison.

[01:27:35] Speaker 05:
Here's my understanding of the Apple argument that was based on Dr. Colwell's testimony.

[01:27:43] Speaker 02:
He was... That Steely has nothing to do with predicting, right?

[01:27:49] Speaker 02:
Isn't that what he said?

[01:27:51] Speaker 02:
That Steely's tag comparison does not constitute a prediction.

[01:27:56] Speaker 05:
That's not my understanding of their position.

[01:28:01] Speaker 05:
What I think Your Honor may be referring to, and it is addressed by Judge Connolly's claim construction order, which is in the record, particularly around Appendix 19, addresses a lot of the argument.

[01:28:20] Speaker 05:
And I should note one other thing.

[01:28:23] Speaker 05:
There was a Dr. Colwell declaration

[01:28:26] Speaker 05:
But a fair amount of what Apple argues in its wreaths have to do with Dr. Colwell's testimony at trial, which wasn't before the court in summary judgment.

[01:28:40] Speaker 05:
But even if everything was before the court for purposes of summary judgment, Dr. Colwell speculated as to how the Steeley system might operate.

[01:28:52] Speaker 05:
He could not and did not

[01:28:54] Speaker 05:
point to any disclosure in Steeley.

[01:28:58] Speaker 05:
He would draw a conclusion such as a trial.

[01:29:01] Speaker 05:
Well, I think that's the way it's supposed to operate.

[01:29:05] Speaker 05:
And he provided some testimony that some associations are overwritten.

[01:29:14] Speaker 05:
And what Judge Connolly said was, well, that's just discarding something.

[01:29:20] Speaker 05:
That's not updating the prediction.

[01:29:23] Speaker 05:
as opposed to what is in the intrinsic evidence in the spec of the patent, where you're incrementing a counter, decrementing a counter, or doing something else, which is always updating the prediction to make it better.

[01:29:39] Speaker 05:
So every embodiment in the 752 patent describes what I just described, a dynamic circumstance.

[01:29:49] Speaker 05:
Even where Apple says, wait a minute, part of the spec

[01:29:53] Speaker 05:
is not dynamic, and this is toward the top of number 14, starting at about line three, where this portion describes incrementing and then Apple quotes out of context in its brief a reference to weighting schemes or pattern recognition.

[01:30:17] Speaker 05:
But if one reads that full language about weighting schemes as

[01:30:22] Speaker 05:
Judge Connolly found that not only does not support Apple's position, but it's really a description of something that is dynamic.

[01:30:32] Speaker 05:
Thank you.

[01:30:33] Speaker 05:
We have your argument.

[01:30:34] Speaker 03:
Thank you very much.

[01:30:37] Speaker 03:
We'll restore four minutes, five minutes if you need it.

[01:30:41] Speaker 03:
That will keep things even.

[01:30:43] Speaker 01:
Yes.

[01:30:43] Speaker 01:
Thank you, Your Honor.

[01:30:44] Speaker 01:
Let me take it in reverse order and deal first with prediction, then with vicarious liability, and come back to

[01:30:50] Speaker 01:
question of particular, which is where we started.

[01:30:53] Speaker 01:
On this question of prediction, when the briefing is all done, no one disputes that there's a plain meaning of prediction.

[01:31:01] Speaker 01:
It includes both static and dynamic.

[01:31:04] Speaker 01:
No one has suggested to you anything otherwise.

[01:31:07] Speaker 01:
No one has suggested that WARF was its own lexicographer.

[01:31:11] Speaker 01:
No one has suggested their specification disavow, which Connolly said is there is implicitly

[01:31:17] Speaker 01:
in the specification a requirement that you limit the plain meaning to dynamic.

[01:31:23] Speaker 01:
And that's wrong for two reasons.

[01:31:26] Speaker 01:
One is the claim doesn't limit itself to dynamic.

[01:31:30] Speaker 01:
It just says prediction.

[01:31:31] Speaker 01:
The second is at column 14, when you consider the argument just made to you, Mr. Chu said that the system of the 752 pattern requires incrementing and decrementing.

[01:31:45] Speaker 01:
That's the dynamic aspect.

[01:31:47] Speaker 01:
Column 14 says, you don't have to increment.

[01:31:50] Speaker 01:
And it describes a waiting scheme.

[01:31:53] Speaker 01:
And a waiting scheme could be dynamic.

[01:31:54] Speaker 02:
Are you saying that Steely anticipates because there's any static?

[01:32:04] Speaker 01:
We say for two reasons.

[01:32:05] Speaker 02:
Any static functionality?

[01:32:06] Speaker 02:
I mean, the district court found that Steely does not allow for updates, right?

[01:32:11] Speaker 01:
The district court actually disputed Dr. Caldwell's testimony and Mr. Steely's testimony.

[01:32:16] Speaker 01:
that in fact allows for updates.

[01:32:18] Speaker 01:
One of the things the specification of the 752 patent says is there's temporal locality.

[01:32:25] Speaker 01:
That is, if there's an old misspeculation, we can replace it with a new one.

[01:32:29] Speaker 02:
Right.

[01:32:29] Speaker 02:
So it has to have a dynamic capability, right?

[01:32:32] Speaker 01:
That's exactly what COE does.

[01:32:33] Speaker 01:
And that's the disputed issue that the jury should have resolved rather than the judge resolving it.

[01:32:38] Speaker 01:
But at column 14, and I want to do this quickly to stay within my five minutes,

[01:32:44] Speaker 01:
A waiting scheme is this.

[01:32:45] Speaker 01:
There may be a misspeculation that occurs earlier in the program.

[01:32:49] Speaker 01:
It's not a big deal.

[01:32:51] Speaker 01:
So you give it a three.

[01:32:52] Speaker 01:
But there's a misspeculation that can occur later with a pair that is a big deal because it could affect the performance of the processor.

[01:33:00] Speaker 01:
So you give it a seven.

[01:33:02] Speaker 01:
You have a prediction threshold indicator of five because it says which ones are important and which ones aren't.

[01:33:09] Speaker 01:
That's static.

[01:33:10] Speaker 01:
That's a waiting scheme.

[01:33:12] Speaker 01:
And when the specification says you don't have to increment or decrement, you could have these weighting schemes, they could be dynamic too.

[01:33:19] Speaker 01:
But they also could be static.

[01:33:22] Speaker 01:
And that's why the argument that's being made to you is that the plain meaning could include dynamic and static.

[01:33:30] Speaker 01:
There's no lexicography.

[01:33:32] Speaker 01:
But we think the preferred embodiment should somehow limit the claims not just to dynamic, but to a particular type of dynamic.

[01:33:40] Speaker 01:
And that wouldn't make any sense as a matter of law.

[01:33:43] Speaker 01:
And what Dr. Caldwell explains, and this is the last thing I'll say in this issue, is no, if you read Steely, it does two things.

[01:33:50] Speaker 01:
One is it tags load and store instructions that have misspeculated.

[01:33:55] Speaker 01:
And when you get to a subsequent misspeculation with a different load store pair, it tags them.

[01:34:02] Speaker 01:
But when you get to a subsequent misspeculation with the same load, it will update.

[01:34:06] Speaker 01:
It's the temporal dislocation.

[01:34:08] Speaker 01:
that's discussed in the 752 patent.

[01:34:12] Speaker 01:
However you judge that argument, it should have gone to the jury.

[01:34:18] Speaker 01:
It was not for the judge to resolve.

[01:34:20] Speaker 01:
On the issue of vicarious liability, let me say, to answer a couple questions quickly, Judge O'Malley, that you raised.

[01:34:27] Speaker 01:
The question of whether Akamai is 271A for all purposes has been answered post-Akamai by this court.

[01:34:35] Speaker 01:
In a number of cases, there's the Lada case,

[01:34:37] Speaker 01:
at 838 Fed 2nd at 1339, 1340.

[01:34:40] Speaker 01:
It basically says, we're applying the proposition to joint infringement of method claims, but not the system claims.

[01:34:49] Speaker 01:
It's answering the question of whether it applies to all of 271A.

[01:34:53] Speaker 01:
Second point on vicarious liability is this.

[01:34:56] Speaker 01:
The answer to your question is, there is the Finjin silicon graphics line of cases.

[01:35:01] Speaker 01:
There is the Nizome cases.

[01:35:03] Speaker 01:
Here are the distinctions.

[01:35:05] Speaker 01:
The first is that the nizome cases require some structural modification before you can use them.

[01:35:13] Speaker 01:
That's what bumping and fusing is.

[01:35:14] Speaker 01:
It's a structural modification and you can't use them.

[01:35:18] Speaker 02:
The second distinction... And so that's the key question I'm trying to figure out is, are we just talking about turning on the functionality versus modifying the functionality?

[01:35:32] Speaker 01:
And, Your Honor, this will get to my second

[01:35:35] Speaker 01:
My last point on this, but let me just finish one thought on the distinction in the cases that Your Honor identified.

[01:35:43] Speaker 01:
In FinGEN, fantasy sports, silicon graphics, the user could turn them on without structural modification.

[01:35:51] Speaker 01:
They had to insert a passcode.

[01:35:53] Speaker 01:
They had to buy something off the internet.

[01:35:55] Speaker 01:
But the user could make it happen.

[01:35:58] Speaker 01:
The user can't make anything happen to these wafers that are pizza-sized, not fused, not bumped, not cut.

[01:36:04] Speaker 01:
That's why they're going to Samsung in Korea for the purpose of being bumped, fused, and cut.

[01:36:11] Speaker 01:
To go to the point I think your honor is raising, which is my last point on this, is the testimony that Mr. Chu cited to you can be addressed in this way.

[01:36:22] Speaker 01:
The requirement is just not capable.

[01:36:23] Speaker 01:
It's capable of executing program instructions.

[01:36:27] Speaker 01:
It's a specific capability.

[01:36:29] Speaker 01:
All of the testimony he cited to you before pages 3924 and 3925, which I discussed with you, all it says is you can test.

[01:36:39] Speaker 01:
It doesn't say what for.

[01:36:40] Speaker 01:
It doesn't say whether they're capable of performing program instructions.

[01:36:45] Speaker 01:
That is the key.

[01:36:46] Speaker 01:
I agree with Mr. Chu on that.

[01:36:47] Speaker 01:
When you got to that question.

[01:36:49] Speaker 02:
So your point is that the functionality is all present.

[01:36:53] Speaker 02:
It's just not, you just can't access it.

[01:36:56] Speaker 01:
No, Your Honor, our point is this, and I should have been more precise.

[01:37:00] Speaker 01:
The transistors and wires are there.

[01:37:03] Speaker 01:
That's what happens during the lithography process.

[01:37:05] Speaker 01:
That's not enough.

[01:37:07] Speaker 01:
Until you bump, fuse, cut, you can't use the chips.

[01:37:14] Speaker 01:
And the testimony that Mr. Chu cited to you, all of which was before the judge struck Mr. Conte's testimony, none of it says anywhere.

[01:37:26] Speaker 01:
anywhere that those wafers are capable of performing program instructions.

[01:37:31] Speaker 01:
It doesn't tell you how to load software onto those wafers.

[01:37:34] Speaker 01:
It doesn't tell you whether software was loaded onto those wafers.

[01:37:38] Speaker 01:
It doesn't tell you whether they were capable of performing at all.

[01:37:41] Speaker 01:
The place where the rubber met the road was at page 3924 to 3925.

[01:37:47] Speaker 01:
And what Mr. Que did in his argument is the same thing they did in their brief, and that is that

[01:37:55] Speaker 01:
you ignore the page and a half that comes before when the judge says no.

[01:38:00] Speaker 01:
If you didn't test it, if you weren't there, if you don't know what the test was for, I'm not going to let you say what it was capable of doing.

[01:38:08] Speaker 01:
That is insufficient as a matter of law.

[01:38:11] Speaker 01:
Last point is on this question of particular.

[01:38:14] Speaker 01:
And I think, Judge Bryson, as you basically framed the issue, which is let's set aside all of the issues that happened with worse motion,

[01:38:25] Speaker 01:
claim construction, the waiver, no waiver than waiver than the JMO opinion.

[01:38:30] Speaker 01:
Just set them aside.

[01:38:31] Speaker 01:
Now, I think that they add color that requires us to look carefully at this issue.

[01:38:36] Speaker 01:
The issue before you is, OK, if it had gone to the jury on plain and ordinary meaning, it did.

[01:38:43] Speaker 01:
What is a plain and ordinary meaning?

[01:38:47] Speaker 01:
And is there substantial evidence to support a finding of particular under the plain and ordinary meaning?

[01:38:53] Speaker 01:
And we suggest the answer is,

[01:38:55] Speaker 01:
No.

[01:38:56] Speaker 01:
And the answer is in these citations, Your Honor.

[01:39:01] Speaker 01:
At the bottom of Worf's brief at pages 17 and 18, this is where the issue really crystallizes.

[01:39:08] Speaker 01:
And what Worf says is, at the very bottom, even Apple's Dr. August confirms such scenarios exist.

[01:39:17] Speaker 01:
And he agreed.

[01:39:17] Speaker 01:
And then you'll see, quote, a prediction will be associated with more than one load instruction, close quote.

[01:39:26] Speaker 01:
only when, quote, aliasing occurs, close quote.

[01:39:29] Speaker 01:
And then you'll see the citations.

[01:39:31] Speaker 01:
This is the linchpin of their argument responding to the question of Judge Prost and Judge O'Malley on aliasing.

[01:39:37] Speaker 01:
This is the key.

[01:39:40] Speaker 01:
When you go to those pages 2239 to 2240, what you find is the portion that Judge O'Malley quoted to me.

[01:39:49] Speaker 01:
But on the very next page, 2240, Dr. August says, no.

[01:39:54] Speaker 01:
They're hashed all the time.

[01:39:56] Speaker 01:
They are associated with load instructions all the time.

[01:40:00] Speaker 01:
And that is consistent with what Dr. Conti said.

[01:40:04] Speaker 01:
At A1621, Dr. Conti said, Apple hashes 100% of the time.

[01:40:11] Speaker 01:
At A2516-17, Dr. Conti said, hashing and aliasing are different.

[01:40:20] Speaker 01:
Hashing and aliasing are different as an undisputed fact of the matter.

[01:40:23] Speaker 01:
Apple hashes 100% of the time as an undisputed matter.

[01:40:28] Speaker 01:
As a consequence, every load tag is associated with a group of loads.

[01:40:33] Speaker 01:
At A1605 to A1606, Dr. Conti conceded that there are only 4,000 or so load tags for the Apple system, but millions of load instructions.

[01:40:43] Speaker 01:
It has to be a group.

[01:40:47] Speaker 03:
I think we're going to bring this to an end.

[01:40:49] Speaker 01:
Thank you, Your Honor.

[01:40:50] Speaker 03:
Thank you.

[01:40:50] Speaker 03:
We thank both sides, and the case is submitted.