[00:01:02] Speaker 00:
We're ready when you are.

[00:01:06] Speaker 02:
Good morning.

[00:01:07] Speaker 02:
May it please the court.

[00:01:08] Speaker 02:
The PTAP instituted and conducted the 676 IPR despite the petition grounds being facially incomplete as to the required algorithmic structure of the computer implemented means plus function claims, and facially incomplete as to any explanation of structural equivalent to the prior art.

[00:01:27] Speaker 01:
In doing so, the board- Wouldn't you argue before the board

[00:01:30] Speaker 01:
that this had to be a computer with software?

[00:01:35] Speaker 02:
What we argued is that it was a computer-implemented invention.

[00:01:38] Speaker 02:
And what the board took from that is that there is an exception to this court's jurisprudence on algorithmic structure that's limited to software implementations.

[00:01:49] Speaker 01:
So you didn't argue before the board or didn't recognize before the board that there could be a hardware implementation, though you now agree that that's the case, right?

[00:01:59] Speaker 02:
That's correct.

[00:01:59] Speaker 02:
We've always agreed that it's possible, as it is in almost any case, to either have software executing code in a transient fashion, or you could hardwire or burn it into the semiconductor.

[00:02:13] Speaker 02:
It's still an algorithm.

[00:02:14] Speaker 01:
But I don't see that before the board, and correct me if I'm wrong about this, that you argue that if it's hardware, it's got to implement the algorithm and that the prior art didn't show that.

[00:02:28] Speaker 02:
Well, Your Honor, that's what we argued in our patent owner brief.

[00:02:32] Speaker 02:
And in the supplemental briefing, we pointed to the cross-examination of Dr. Russ.

[00:02:38] Speaker 02:
Dr. Russ stated that the algorithm is expressed in hardware.

[00:02:43] Speaker 02:
So it's just circuitry that is doing the algorithm.

[00:02:47] Speaker 01:
Where did you argue before the board that the prior art of hardware implementation didn't have an algorithm and therefore

[00:02:59] Speaker 01:
wasn't a problem.

[00:03:04] Speaker 02:
I'm not sure I understand the question, sorry.

[00:03:06] Speaker 01:
Well, what I'm concerned about is it seems to me, based on what I understand, and I may be wrong about this, that you didn't argue before the board, one, that there could be a hardware implementation, and that two, if there was, that it had to incorporate the algorithm.

[00:03:32] Speaker 02:
Let me give you some citations on that point.

[00:03:47] Speaker 02:
What we argued, and this is JA 279, is that the reproducing means is computer implemented and must be construed to include algorithmic structure.

[00:03:58] Speaker 02:
The point of the... So that's not a hard year.

[00:04:02] Speaker 02:
Well, any computer implementation has to have an algorithm to operate.

[00:04:08] Speaker 02:
The hardware point came in when the petitioner made its reply and was trying to excuse the fact that it didn't describe an algorithm.

[00:04:16] Speaker 02:
It said, well, there's this exception.

[00:04:18] Speaker 02:
And what we're saying is this is a hardware processor, and therefore it doesn't need an algorithm.

[00:04:22] Speaker 02:
None of that was in the petition.

[00:04:24] Speaker 01:
Where did you argue that if there was a hardware implementation, it had to have the algorithm and the hardware?

[00:04:36] Speaker 02:
That may have come up in the supplemental briefing.

[00:04:42] Speaker 02:
So this is, for example, JA 480, where we talk about, finally, even if hardware processors could somehow be algorithm independent, which they cannot, petitioners did not base their trial ground on a finite state machine or hardware processor.

[00:05:02] Speaker 02:
So this is what came in in the petitioner reply.

[00:05:07] Speaker 02:
thus even accepting Petitioner's hardware processor theory, which again was not in its petition.

[00:05:13] Speaker 02:
The trial ground is based on Yoshio's controller, which employs an algorithm.

[00:05:21] Speaker 02:
That was the trial ground.

[00:05:23] Speaker 01:
Well, but all it seems to me is that there's a clause here which they cannot, which is hardly a developed argument, right?

[00:05:33] Speaker 02:
Well, this supplemental briefing came after the oral argument.

[00:05:37] Speaker 02:
The oral argument pointed out that the petitioner's own evidence, if you look at JA 1286, which is there.

[00:05:47] Speaker 02:
Was this argued at the oral argument?

[00:05:50] Speaker 02:
This was argued at the oral argument.

[00:05:51] Speaker 01:
Can you show us where it was argued?

[00:05:54] Speaker 02:
At the oral argument?

[00:05:55] Speaker ?:
Yeah.

[00:06:15] Speaker 02:
So the oral argument on a transcript starts at JA 497 or there about.

[00:06:30] Speaker 02:
So at JA 502, we start talking about the algorithm.

[00:06:38] Speaker 02:
We're arguing about hardware controllers.

[00:06:40] Speaker 02:
The ground doesn't depend on a hardware controller, so we're pointing

[00:06:44] Speaker 02:
that out, the ground actually points to a CPU that uses software, and so further down at line 14, so even accepting what their argument, that there's a controller, it's general purpose, there's an algorithm in the specification that's one structure, but it's not limited to that structure.

[00:07:04] Speaker 02:
There's this other structure, this hardware controller, but if you're consuming a means plus function term, you have to cover all these structures.

[00:07:15] Speaker 02:
That's it.

[00:07:17] Speaker 02:
So further on, J503, there's this discussion of the hardware structure being at ASIC.

[00:07:26] Speaker 02:
This is what came in in their petition or reply.

[00:07:32] Speaker 02:
Down the bottom, line 21, we never argued that our specification is limited to a general purpose computer that uses an instruction set.

[00:07:45] Speaker 02:
Further on 504, there's a discussion of hardware processor takes the instruction set and burns it into circuitry.

[00:07:53] Speaker 02:
That's always a statutory equivalent.

[00:07:56] Speaker 02:
We've never argued a hardware controller is not within the scope of these claims.

[00:08:02] Speaker 02:
What we said is an algorithm is required because you can't have a hardware processor.

[00:08:06] Speaker 02:
Finite state machines don't work in the abstract.

[00:08:09] Speaker 02:
And then we start walking through their exhibits where they explain that.

[00:08:13] Speaker 02:
Finite state machines operate in accordance with state machine language, SML.

[00:08:18] Speaker 02:
That's how you get the algorithm into the semiconductor.

[00:08:22] Speaker 01:
There's no evidence in the record.

[00:08:23] Speaker 01:
Even if you did, in fact, raise this as oral argument, isn't that too late under the board's scheduling order in these cases, which says that you can't raise things for the first time at oral argument?

[00:08:37] Speaker 02:
Well, in this case, we were responding to- Well, I'm right about the scheduling order.

[00:08:41] Speaker 02:
That is correct, but in this case, this hardware processor point was not in the petition.

[00:08:48] Speaker 02:
The grounds that were instituted was this was a CPU, full stop.

[00:08:54] Speaker 02:
Nothing about a hardware processor that doesn't need an algorithm.

[00:08:58] Speaker 02:
That's what came in in the petitioner reply when they realized that they were on the wrong side of the law, which they changed in the district court construction, which has also been briefed.

[00:09:10] Speaker 02:
So this argument

[00:09:11] Speaker 02:
came in, we were not allowed to submit any additional evidence.

[00:09:18] Speaker 02:
We were just rebutting something that wasn't even in the trial ground that we saw for the first time.

[00:09:23] Speaker 02:
There is no exception in this court's jurisprudence where you can ignore an algorithm just because you take code translated from software into circuitry.

[00:09:33] Speaker 02:
You still have to have the algorithm, and that's exactly what the petitioner's evidence makes clear.

[00:09:39] Speaker 02:
They say, well, we don't have to

[00:09:41] Speaker 02:
describe an algorithm because there's this block diagram, and therefore that's a hardware processor.

[00:09:47] Speaker 02:
Well, we wouldn't have any of these 112 decisions on algorithms if you could just point to a picture with a block diagram in it.

[00:09:53] Speaker 02:
That's most of those cases.

[00:09:56] Speaker 02:
You've only shown a block diagram.

[00:09:58] Speaker 02:
There's no description of the algorithm.

[00:10:00] Speaker 02:
Well, if you could just throw up your hands and say it's a hardware processor, we'd be done.

[00:10:04] Speaker 02:
We wouldn't have to get into all of this.

[00:10:06] Speaker 02:
So this came in late.

[00:10:08] Speaker 02:
It was not the trial ground.

[00:10:10] Speaker 02:
We can't argue against something that we don't see until the petitioner reply.

[00:10:17] Speaker 02:
And then separate and apart from that, what we've said is, well, even accepting that this hardware processor theory makes technical sense, the Yoshio reference does not describe a hardware processor.

[00:10:29] Speaker 02:
It's a CPU that operates with an instruction set.

[00:10:33] Speaker 02:
And if your argument is,

[00:10:34] Speaker 02:
Well, this is a means plus function claim term.

[00:10:37] Speaker 02:
It has structures A and B, one of which has an algorithm, one of which is a hardware processor.

[00:10:42] Speaker 02:
If your prior art is the version that has the algorithm, then you need to compare the algorithms in any event, and you need to do the equivalency analysis of, is that algorithm a structural equivalent to the algorithm that's in this patent?

[00:10:57] Speaker 02:
And that wasn't done either.

[00:10:59] Speaker 02:
So separate and apart from the fact that this hardware processor issue came in late,

[00:11:04] Speaker 02:
The prior art would fail under that theory.

[00:11:10] Speaker 02:
So this is a relatively narrow, straightforward issue.

[00:11:14] Speaker 02:
I don't want to belabor the point.

[00:11:16] Speaker 02:
But there's no sidestepping the court's algorithmic structure jurisprudence based on a hardware processor theory.

[00:11:23] Speaker 02:
Whether you're executing instructions temporarily as code or you're burning the code as a circuit, you still need the algorithm to meet the function of that claim.

[00:11:33] Speaker 02:
There's no evidence in this record that some off-the-shelf hardware controller could be picked up and plugged in that would do this algorithm.

[00:11:41] Speaker 02:
And that was the petitioner's burden.

[00:11:43] Speaker 02:
It was not in trial grounds.

[00:11:45] Speaker 02:
It was not briefed in the trial because this hardware processor theory came in later to explain away the mistake in the law.

[00:11:54] Speaker 02:
So the grounds as well are deficient.

[00:11:57] Speaker 02:
There's no structural equivalence to the algorithm in the specification.

[00:12:01] Speaker 02:
There's no dispute in the record.

[00:12:03] Speaker 02:
that the patent describes an algorithm and links it to this means plus function claim.

[00:12:08] Speaker 02:
It was simply sidestepped under this hardware processor theory, which I submit is incorrect as a matter of law.

[00:12:16] Speaker 02:
Why don't we hear from the other side?

[00:12:18] Speaker 02:
Good morning.

[00:12:27] Speaker 03:
Good morning.

[00:12:28] Speaker 03:
May it please the court?

[00:12:30] Speaker 03:
Your Honor, just touching on Judge Dyke's point, I think you're right.

[00:12:33] Speaker 03:
Sony did argue in its briefing to the board that the controller was computer implemented.

[00:12:40] Speaker 03:
They said it must be computer implemented.

[00:12:42] Speaker 03:
And that simply is not supported by the intrinsic evidence or the extrinsic evidence.

[00:12:48] Speaker 03:
If you look at the 676 specification, consistently throughout the spec, it just refers to a controller.

[00:12:54] Speaker 03:
It doesn't disclose any type of computer or processor.

[00:12:58] Speaker 01:
I think at this point they agree with that.

[00:13:00] Speaker 01:
It seems to me what they're saying is that the petitioner didn't argue invalidity or unpatentability based on the hardware embodiment of this

[00:13:22] Speaker 01:
But they didn't have an opportunity to address that in the patent owner response as a result.

[00:13:29] Speaker 01:
I think that's what they're saying.

[00:13:30] Speaker 03:
Yeah.

[00:13:31] Speaker 03:
And in response to that argument, I would say they were put on notice of what the petitioner, how they were asserting Yoshio met the controller limitation.

[00:13:39] Speaker 03:
So they certainly had an opportunity to respond to that argument.

[00:13:43] Speaker 03:
And in their patent owner response, they expressly stated they were not going to put on rebuttal evidence because they thought the claim construction was overly broad and incorrect.

[00:13:52] Speaker 03:
They certainly had an opportunity at that point if they chose to.

[00:13:54] Speaker 03:
They could have put on some rebuttal evidence.

[00:13:58] Speaker 03:
They could have cross-examined Dr. Russ.

[00:14:00] Speaker 03:
They actually took the step of withdrawing their own expert testimony.

[00:14:04] Speaker 03:
So that was a decision they made, but they certainly were put on notice that Yochiho

[00:14:09] Speaker 03:
I was mapped to the claims and under petition.

[00:14:12] Speaker 01:
Was there an argument in the petition that Yoshio showed a hardware implementation of this?

[00:14:20] Speaker 03:
No, the Yoshio controller and I believe Eris's counsel stated this during the oral argument.

[00:14:26] Speaker 03:
The Yoshio controller is a computer implemented controller and it was mapped because it has a processor or CPU or something.

[00:14:34] Speaker 03:
So it is computer implemented.

[00:14:36] Speaker 03:
But under the claim construction, as long as you have a controller, and the claims require a controller, Yoshio discloses a controller, it performs, and the board explained, and Dr. Ross explained how it performs the recited functions, and that's sufficient.

[00:14:50] Speaker 03:
And if Sony wanted to make an argument that they thought it didn't meet the claims a little literally, and there should have been some type of equivalency analysis, they could have raised that argument during the proceedings, but they chose not to.

[00:15:02] Speaker 03:
And so in our brief, we find that they waived the right to do so before this court.

[00:15:08] Speaker 03:
So unless you have any other questions about some of the issues that were briefed, I'm happy to answer them.

[00:15:16] Speaker 03:
If you have questions about anything else.

[00:15:22] Speaker 01:
Well, what the board said in its decision was that computer implementation wasn't required, right?

[00:15:29] Speaker 01:
Correct.

[00:15:30] Speaker 01:
and that a hardware implementation would count.

[00:15:34] Speaker 01:
Did they find the hardware implementation in the prior art?

[00:15:39] Speaker 03:
No.

[00:15:40] Speaker 03:
So they found that the claims were not limited to a computer-implemented controller.

[00:15:46] Speaker 03:
It could be either hardware or computer-implemented, but they found based on the intrinsic evidence, based on the spec, how it's depicted in Figure 7 as a hardware element along with other hardware elements.

[00:15:57] Speaker 03:
And based on Dr. Russ's expert testimony, he pointed out that these claims were filed, had priority back to 1991.

[00:16:03] Speaker 03:
And at that time, it was well known that controllers could be implemented in hardware.

[00:16:08] Speaker 03:
Dr. Russ gave examples of like a dishwasher, a washing machine, as a hardware controller, would have a hardware controller and a traffic light.

[00:16:16] Speaker 03:
So it was well known.

[00:16:16] Speaker 03:
And so Dr. Russ thought, even though, or concluded, that one in skill in the art, looking at the intrinsic evidence, would find that the hardware controller

[00:16:25] Speaker 03:
the controller in the 676 spec was more likely a hardware controller, but that it could be implemented as a hardware controller or a computer-implemented controller.

[00:16:36] Speaker 01:
So, and they found the computer implementation in Yoshio?

[00:16:40] Speaker 03:
Correct.

[00:16:41] Speaker 03:
And they found that met the controller limitation because it was a controller, and it performed the recited functions, and the claims weren't limited to that specific algorithm.

[00:16:51] Speaker 03:
And that was sufficient.

[00:16:52] Speaker 00:
Well, just to be clear, then, the position of the office is that for computer-implemented procedures, it's irrelevant whether the algorithm is or is not disclosed or suggested in the prior art.

[00:17:08] Speaker 03:
No, I think if the spec and the claims were limited to a computer-implemented controller,

[00:17:14] Speaker 03:
Then under this court's precedent, you'd have to include the algorithm recited in the spec.

[00:17:19] Speaker 03:
But here, since there's no requirement that the claims were not limited to a computer-implemented controller that could be formed on a hardware controller, and it was more likely to be a hardware controller, you weren't required to perform that specific algorithm.

[00:17:33] Speaker 03:
Just performing the recited functions would be sufficient to meet that claim limitation, so long as you had a controller of some type in the prior.

[00:17:40] Speaker 00:
So for a means plus function, again,

[00:17:44] Speaker 00:
You're saying it's irrelevant whether the means contains an algorithm in order for the function to be read on the prior art.

[00:18:01] Speaker 03:
In this case, since it was a hardware controller, the algorithm was not part of the claim construction.

[00:18:07] Speaker 00:
I want to understand the difference between this case and whatever

[00:18:11] Speaker 00:
generalization, whatever general practice is being imposed on computer-implemented means, claims to computer-implemented means.

[00:18:22] Speaker 03:
Yeah, so if it was solely a computer-implemented means that was disclosed, like a general purpose computer or a processor, then the spec would have to disclose some type of algorithmic structure.

[00:18:32] Speaker 03:
Otherwise, if it didn't, your claims would be indefinite.

[00:18:36] Speaker 01:
So, Yoshio didn't disclose the algorithm.

[00:18:39] Speaker 03:
There is no finding.

[00:18:41] Speaker 03:
The court actually didn't address the algorithm, because it wasn't part of the claim construction.

[00:18:45] Speaker 03:
And so there's no fact finding on whether Yoshio disclosed that algorithm.

[00:18:51] Speaker 01:
So the theory is that Yoshio did disclose a controller and a synthesizer, and that was sufficient.

[00:19:02] Speaker 03:
Yes, because the controller and the synthesizer performed the recited function of reproducing the audio according to the designated channel.

[00:19:11] Speaker 03:
The board actually went through it.

[00:19:12] Speaker 03:
Dr. Russ explained how in Yoshio, spanning columns six and seven, it had a controller, 25, which would look at to determine the designated channel, which is basically which language should you reproduce the audio in.

[00:19:26] Speaker 03:
It would send that to the decoder.

[00:19:28] Speaker 03:
The decoder would then

[00:19:31] Speaker 03:
create a digital audio file, it would send it to the converter, which would then convert the digital file into an audio file.

[00:19:38] Speaker 03:
And so together, you had the controller plus those two other components.

[00:19:42] Speaker 03:
We met the synthesizer limitation.

[00:19:44] Speaker 03:
And together, they performed the recited function of reproducing the audio according to the designated channel.

[00:19:49] Speaker 03:
And that met the claim limitation.

[00:19:52] Speaker 03:
And again, Sony didn't provide any rebuttal evidence to that.

[00:19:55] Speaker 03:
They just kind of argued solely that the claim instruction was incorrect.

[00:20:01] Speaker 03:
Thank you.

[00:20:02] Speaker 02:
Thank you.

[00:20:13] Speaker 02:
Just a couple of points on Yoshio.

[00:20:15] Speaker 02:
There's no dispute that it uses software.

[00:20:17] Speaker 02:
What we've said in the briefing is that it doesn't describe any algorithm.

[00:20:21] Speaker 02:
It just states it does the function.

[00:20:24] Speaker 02:
It's not our burden to rebut something that's not in the record.

[00:20:27] Speaker 02:
It's the petitioner's burden to explain

[00:20:30] Speaker 02:
If that is your ground and that uses an algorithm, then how is it equivalent to a different structure, if you accept the petitioner's argument, of a hardware processor that can use any algorithm?

[00:20:42] Speaker 02:
So what's being argued here, in essence, by the PTO is that as long as you have a discrete box and a figure, you own every way of implementing a function.

[00:20:53] Speaker 02:
That can't possibly be correct.

[00:20:56] Speaker 01:
I don't think that's what they're arguing.

[00:20:59] Speaker 01:
I think they're arguing that you didn't, before the oral argument, before the board, present this theory that in a hardware implementation, you still had to have an algorithm.

[00:21:13] Speaker 02:
Well, Your Honor, I would submit that it's not our duty to argue against ourselves.

[00:21:17] Speaker 02:
The ground was the claim requires a CPU, period.

[00:21:22] Speaker 02:
Nothing about a specific hardware processor

[00:21:25] Speaker 02:
Nothing about a specific hardware processor that's already preloaded the circuitry to perform the function.

[00:21:31] Speaker 02:
That was their burden.

[00:21:33] Speaker 02:
We can't imagine what they're going to argue when we point out that their claim construction was fatally flawed.

[00:21:41] Speaker 02:
It's clear in the specification by using words such as executing, processing, the flow chart, that there's an algorithm there.

[00:21:49] Speaker 02:
It's their duty to compare that structure to the prior art.

[00:21:53] Speaker 02:
And if you want to argue,

[00:21:54] Speaker 02:
that the specification has structures A and B, and you rely on B, and yet your ground points to A, there still needs to be an equivalency analysis there.

[00:22:07] Speaker 02:
Essentially, the argument here is a computer has software, period.

[00:22:12] Speaker 02:
A computer can't be a specific semiconductor that's already had the circuitry burned into it.

[00:22:19] Speaker 02:
But that flips this cord-storied prunes on its head.

[00:22:22] Speaker 02:
What makes that processor

[00:22:24] Speaker 02:
a special processor is the executing algorithm.

[00:22:28] Speaker 02:
So what difference does it make whether it's operating in software or you burn it into the code?

[00:22:34] Speaker 02:
And then finally, all of the petitioner's evidence on point makes clear that these processors are burned algorithms into the semiconductor.

[00:22:43] Speaker 02:
The algorithm doesn't go away just because it's a hardware processor, and that's in the petitioner's own evidence.

[00:22:52] Speaker 02:
So there's no way

[00:22:53] Speaker 02:
to sidestep the need for an algorithm.

[00:22:56] Speaker 02:
As I said, every case that's been before this court where there's been an indefiniteness issue or lack of written description support for lack of an algorithm is usually there's just a block diagram with a processor box you didn't describe an algorithm.

[00:23:12] Speaker 02:
Well, why doesn't everybody just argue, well, that's a hardware processor and I don't need it?

[00:23:16] Speaker 02:
Because it doesn't make any technical sense.

[00:23:19] Speaker 02:
get computers and a hardware processor is a computer to operate without an algorithm.

[00:23:26] Speaker 02:
So for that reason, these grounds should be reversed and these claims found not patentable.

[00:23:35] Speaker 02:
Thank you.

[00:23:37] Speaker 02:
We thank both sides.

[00:23:38] Speaker 02:
And the case is submitted.

[00:23:40] Speaker 02:
And I guess everybody can stay where they are.

[00:23:43] Speaker 02:
And the next case is