[00:00:00] Speaker 03:
632 Phizen Electronics versus Silicon Motion Technology.

[00:00:39] Speaker 03:
Mr. Hoffman, whenever you're ready.

[00:00:44] Speaker 01:
May it please the Court, Your Honor, David Hoffman for the Appellant Fison.

[00:00:49] Speaker 01:
With the Court's permission, I'm going to address the issues in a different order than was in the briefing, because I think that it may be most helpful to address the NDP issue.

[00:00:58] Speaker 02:
Would you tell me where in the record you can test the PTAB's construction of predetermined data?

[00:01:06] Speaker 01:
Yes, Your Honor.

[00:01:08] Speaker 01:
If you look

[00:01:10] Speaker 01:
at A582, your honor.

[00:01:27] Speaker 01:
And on A582, we argued, this is the patentless response, that the 267 makes it clear that the new block indicator is not replacement data.

[00:01:40] Speaker 01:
That is an argument not based on the facts of what the prior art is showing, but an argument based on what the 267 patent discloses.

[00:01:48] Speaker 02:
Well, but that's not the same thing.

[00:01:50] Speaker 02:
You're not answering my question.

[00:01:54] Speaker 01:
I believe I am, Your Honor.

[00:01:57] Speaker 01:
For our point of view for claim construction, the issue is whether or not there is a dispute with the board over what the proper scope of the term predetermined data is.

[00:02:07] Speaker 01:
Yes, I agree that we both use the words

[00:02:11] Speaker 01:
replacement data.

[00:02:13] Speaker 01:
However, we believe that in the board's opinion, they did not use the replacement part of that.

[00:02:20] Speaker 01:
They only used data.

[00:02:21] Speaker 01:
And in doing so, it shows you in this section that we're very clear.

[00:02:26] Speaker 02:
You're claiming, you're arguing, I think, that they misapplied their construction.

[00:02:30] Speaker 01:
I don't think so, Your Honor.

[00:02:31] Speaker 01:
I think there's a difference because... There is a difference.

[00:02:34] Speaker 01:
There's definitely a difference, Your Honor.

[00:02:36] Speaker 01:
We believe that if it was a misapplication, we wouldn't be discussing whether or not the 267 patent itself and its disclosure of what predetermined data was included or did not include a particular aspect.

[00:02:48] Speaker 01:
And again, the indicator is not something that's only in the prior art.

[00:02:52] Speaker 01:
The indicator in this case is in the patent itself.

[00:02:56] Speaker 01:
And so when we talk about the patent itself and we describe what the patent itself allows for a particular claim term, and it's noted here for predetermined data,

[00:03:05] Speaker 01:
What we are saying there is that the scope of the claim does not include a single bit indicator.

[00:03:11] Speaker 01:
That's outside the scope of the claim.

[00:03:12] Speaker 02:
And that is a classic claim construction argument, not an argument that- So you didn't argue that predetermined data has to be the same size as the original data as a matter of claim construction?

[00:03:23] Speaker 01:
We did not argue that predetermined data need to be the same size as a matter of claim construction.

[00:03:27] Speaker 01:
No, Your Honor.

[00:03:29] Speaker 01:
Barbara, we did argue in claim construction

[00:03:31] Speaker 01:
is that the term predetermined data must be construed in line with the specification and the claims of a 267 patent.

[00:03:38] Speaker 01:
That's what you see on this page.

[00:03:40] Speaker 01:
And the specification and the claims of the 267 patent, and the claims make it clear that predetermined data is different and separate from an indicator, because an indicator is a claimed element in the 267 patent.

[00:03:54] Speaker 01:
Additionally, in the patent, on the first page of the 267 patent,

[00:04:02] Speaker 01:
Column 2, in the summary of the invention, and this language is important because it's not describing an embodiment.

[00:04:08] Speaker 01:
It actually says the invention of the 267 patent is directed to a data-accessing method, and that method prevents the host from receiving garbled codes.

[00:04:19] Speaker 01:
And if you look at the claim where this language appears, it is also a data-accessing method.

[00:04:24] Speaker 01:
And transmitting a single-block indicator doesn't prevent the host from receiving garbled codes at all.

[00:04:32] Speaker 01:
It is an indicator of an erased block status.

[00:04:36] Speaker 01:
It's not predetermined data.

[00:04:39] Speaker 01:
It doesn't stop the garbled code problem.

[00:04:41] Speaker 01:
So between the clear disclosure of the patent, which says the invention solves the garbled code problem, and the differentiation between the use of these words in the claims and in the spec, we believe that the board erred in construing the term predetermined data, replacement data, to effectively not

[00:04:59] Speaker 01:
require replacement to merely be any other data beyond what's stored.

[00:05:05] Speaker 01:
It's not replacement data.

[00:05:07] Speaker 03:
Can I just move on to the obviousness analysis, because maybe all of this will be clarified with that.

[00:05:12] Speaker 03:
My understanding is that there's no dispute left, at least, that Sharon discloses everything except for the transmitting step, predetermining data.

[00:05:23] Speaker 03:
So then we're left with Bennett.

[00:05:25] Speaker 03:
And the board at A10,

[00:05:28] Speaker 03:
cites the quote from Bennett.

[00:05:31] Speaker 03:
Bennett is directed, and at the end it says, such systems return the erased status and or erased data in response to the red operation from an erased block.

[00:05:47] Speaker 03:
So is the erased data in Bennett predetermined data?

[00:05:52] Speaker 01:
That's the board's decision and its basis for finding that the... Can you just answer my question?

[00:05:57] Speaker 01:
Is the erased data?

[00:05:58] Speaker 03:
Yeah.

[00:05:58] Speaker 01:
That wasn't a subject.

[00:06:00] Speaker 01:
I don't know, Your Honor, because that wasn't the subject of the arguments below.

[00:06:04] Speaker 01:
The board's basis and the instituted grounds here rely on transmitting the ES bit and relying on that as being the predetermined data, not the other sections that have to be quoted in there.

[00:06:19] Speaker 01:
And you can see that

[00:06:25] Speaker 01:
on page A13 at the top.

[00:06:28] Speaker 01:
Although the FISN is correct that the ES bit of Bennett can be an indicator, it can also serve as replacement data.

[00:06:36] Speaker 01:
The board's decision is that the ES bit is the replacement data that's being transmitted.

[00:06:42] Speaker 03:
But, can you still answer my question?

[00:06:45] Speaker 03:
I mean, but Bennett is cited as the prior art, and they cite the ES bit

[00:06:52] Speaker 03:
and or erased data in response to a red operation from the erased block.

[00:06:58] Speaker 03:
Why isn't that sufficient?

[00:07:00] Speaker 03:
I mean, you're just focusing on the bit part of Bennett.

[00:07:03] Speaker 03:
But if you focus on the erased data, why is that not pre-determined?

[00:07:06] Speaker 01:
On behalf of Your Honor, because they are returning the erased data.

[00:07:09] Speaker 01:
Again, this wasn't the issue.

[00:07:10] Speaker 01:
I mean, the board decided that's the garbled code.

[00:07:13] Speaker 01:
If you read that section of a block that is the erased section and you decode it under the guise of sharing, what you're going to get is garbled code.

[00:07:22] Speaker 01:
So you can't practice the invention by returning the garbled code.

[00:07:26] Speaker 01:
That's exactly the opposite.

[00:07:27] Speaker 01:
That's why the board did not rely on that section, because you're just practicing essentially exactly the prior art if you return the garbled erased data instead of, in this case, they say an indicator, because that's what they have to rely on, because it wouldn't be decoded in the way that the data would.

[00:07:53] Speaker 01:
I could move on to NTP, unless there are more questions about that combination, Your Honor.

[00:08:01] Speaker 03:
Is your reading of NTP really that whenever two references have overlapping elements, they can't be combined as a matter of law?

[00:08:09] Speaker 03:
No, Your Honor.

[00:08:12] Speaker 03:
What is your reading of NTP?

[00:08:13] Speaker 01:
Our reading of NTP is in a situation where you have two references that are being combined.

[00:08:20] Speaker 01:
The obviousness combination for mapping to the element requires that you add to a reference something that is completely superfluous.

[00:08:28] Speaker 01:
That is where NTP kicks in.

[00:08:30] Speaker 01:
So in this particular example, Sharon doesn't disclose an element.

[00:08:34] Speaker 01:
They merely take in the Bennett ES bit and replaced the ES bit in Sharon with the indicator in Sharon with the indicator from Bennett.

[00:08:45] Speaker 01:
Classic obviousness analysis under KSR.

[00:08:48] Speaker 02:
What they did here is because the claims also require... Where does it say that they replaced the indicator in Sharon with the bit from Bennett?

[00:08:59] Speaker 01:
They didn't, Your Honor.

[00:09:00] Speaker 01:
That's, I think, the problem.

[00:09:02] Speaker 01:
If they had done a classic... What they could have said.

[00:09:05] Speaker 01:
Well, they couldn't say that here.

[00:09:06] Speaker 01:
I think that's why they get themselves into an NTP problem.

[00:09:09] Speaker 02:
So your statement is they replaced the indicator in Sharon with the bit from Bennett.

[00:09:15] Speaker 02:
That's not correct.

[00:09:16] Speaker 01:
No, I was setting up a hypothetical, Your Honor.

[00:09:18] Speaker 01:
If they had done that as part of an analysis, if that had been the only thing they had done, then the overlap in the references wouldn't have made a difference.

[00:09:26] Speaker 01:
But because here, what they did was they added the ES bit of Bennett, but because the claims also require another indicator, and they couldn't use the ES bit, which is itself an indicator, they couldn't use that for the indicator also in the claims, because they had changed its function.

[00:09:42] Speaker 02:
They didn't.

[00:09:43] Speaker 02:
They used the flag from...

[00:09:45] Speaker 02:
They used the five for sharing.

[00:09:46] Speaker 02:
And they said that the bit from Bennett was predetermined data.

[00:09:52] Speaker 01:
They did, Your Honor.

[00:09:53] Speaker 01:
But importantly, they never said in their holding that the ES bit changed.

[00:09:57] Speaker 01:
It is still the ES bit.

[00:09:58] Speaker 01:
It is still an indicator.

[00:09:59] Speaker 01:
It's just been termed to be, for the purposes of mapping, replacement data.

[00:10:05] Speaker 01:
So what we get there is a situation where there is no function anymore remaining in having the

[00:10:12] Speaker 01:
the Sharon indicator.

[00:10:13] Speaker 01:
Everything the Sharon indicator does, and the board doesn't dispute this, can be done by the ESBIT.

[00:10:18] Speaker 01:
So from a technological standpoint, it is, as NTP explains, it's the height of hindsight for when you're mapping elements to a claim to take something which in a real system would have absolutely no function at all, because it had been completely replaced, and to add that in merely for the purposes of mapping the claims.

[00:10:35] Speaker 01:
And that's the situation in this case, where it's not merely an overlap, it's that they have to rely on

[00:10:41] Speaker 01:
a completely meaningless functionality in the system in order to get to the claims.

[00:10:46] Speaker 01:
And a person's matter of law, when doing a combination, a person wouldn't do that, because it's the definition of hindsight, when the piece has no function in a real system anymore, completely superfluous.

[00:11:03] Speaker 01:
And I have my light.

[00:11:06] Speaker 01:
Other questions?

[00:11:06] Speaker 03:
Why don't you hear from the other side?

[00:11:17] Speaker 00:
May it please the court?

[00:11:22] Speaker 00:
I don't believe that there is a claim construction dispute that can be raised here on the record below.

[00:11:29] Speaker 00:
In its preliminary response, Fizon argued that predetermined data is replacement data stored before user operation.

[00:11:38] Speaker 00:
The board and its institution decision agreed that it was replacement data, but declined to say that it was stored before user operation.

[00:11:46] Speaker 00:
At footnote four of page 13 of Faison's brief, they're not challenging the fact that stored before user operation was not part of the claim construction.

[00:11:57] Speaker 00:
Faison down below never proposed a claim construction that said replacement data must be larger than a bit.

[00:12:02] Speaker 00:
They never proposed a construction.

[00:12:04] Speaker 00:
He admitted that.

[00:12:06] Speaker 00:
And they also never proposed a construction that said the data must be the same size.

[00:12:10] Speaker 00:
They also never proposed a construction that said the garbled code problem must be

[00:12:14] Speaker 00:
solved in the host.

[00:12:15] Speaker 00:
I agree they made those arguments with respect to application, but they did not make it with respect to claim construction.

[00:12:21] Speaker 00:
Thus, there was nothing to appeal with respect to claim construction.

[00:12:25] Speaker 00:
We know this from Faison's own words and its patent owner response.

[00:12:30] Speaker 00:
At 8564, they say, first, the combined system would not practice the predetermined data limitation as construed by the board.

[00:12:41] Speaker 00:
And at A578, which is what the board relied on when it said Faison agreed with them, Faison says, in the institution decision, the board properly recognized that the claimed predetermined data must be replacement data.

[00:12:56] Speaker 00:
However, in applying this construction to the system of Sharon modified to use an ES bit, the board made an error.

[00:13:03] Speaker 00:
So I think it's clear that this is subject to the substantial evidence standard and not de novo review.

[00:13:10] Speaker 00:
FISN cannot meet that standard with respect to the obviousness combination.

[00:13:15] Speaker 00:
Sharon discloses a flash memory device having a flash controller that executes flash management software.

[00:13:20] Speaker 00:
That's a finding of the board at A9.

[00:13:24] Speaker 00:
The board also found at A8-9 that Sharon's invention is directed to reducing error rates.

[00:13:30] Speaker 00:
Yet Sharon does explicitly say you can set a flag or an indicator to indicate a new block.

[00:13:36] Speaker 00:
However, Sharon doesn't tell you

[00:13:38] Speaker 00:
What you do when you read that indicator, they say, that's in the prior art.

[00:13:41] Speaker 00:
Look to the prior art logic.

[00:13:43] Speaker 00:
That's at A9.

[00:13:45] Speaker 00:
The prior art logic includes Bennett.

[00:13:47] Speaker 00:
Bennett and Sharon are both patents filed by a memory company, Sandisk, their related art.

[00:13:53] Speaker 00:
The board found you could properly combine these.

[00:13:55] Speaker 00:
And what it did is it said, how does Bennett deal with the new block?

[00:14:00] Speaker 00:
It sends back a status byte.

[00:14:02] Speaker 00:
For example, the ES bit and or the erased data.

[00:14:06] Speaker 03:
But it never focused on the erased data aspect.

[00:14:09] Speaker 00:
The board did not focus on the erase data.

[00:14:11] Speaker 03:
And you have one paragraph in your brief that calls out.

[00:14:14] Speaker 00:
We did make the argument down below that the erase data could be replacement data, the erase data plus the ESBIT, because it's an ANDOR, or the status byte itself.

[00:14:24] Speaker 00:
But the board did focus in on the ESBIT, correct, in their holding.

[00:14:31] Speaker 00:
And as far as the NTP issue goes, I don't think this is a situation where we need

[00:14:35] Speaker 00:
two networks and we're relying on one network from one and one network from another.

[00:14:38] Speaker 00:
That was the NTP case.

[00:14:40] Speaker 00:
Here, we have the flag cell in Sharon.

[00:14:43] Speaker 00:
Sharon doesn't tell you what to do when you read that flag cell.

[00:14:47] Speaker 00:
Bennett tells you you can transmit back a status byte.

[00:14:50] Speaker 00:
It includes, for example, the ES bit.

[00:14:52] Speaker 00:
That's all the boards relying on is the ability to transmit back to the host.

[00:14:58] Speaker 00:
And for those reasons, we would ask that the board's decision be affirmed.

[00:15:02] Speaker 00:
Are there any questions?

[00:15:17] Speaker 01:
Thank you, Your Honors.

[00:15:18] Speaker 01:
I'll be fairly quick.

[00:15:19] Speaker 01:
We only have five minutes.

[00:15:22] Speaker 01:
We don't argue, we don't dispute that we also argued application.

[00:15:26] Speaker 01:
We argued both at the term scope, the claim had the wrong scope.

[00:15:30] Speaker 01:
We also argued the application.

[00:15:31] Speaker 01:
In pointing to portions of our brief, we argued application.

[00:15:34] Speaker 01:
Again, we're not appealing the application.

[00:15:35] Speaker 01:
We're only appealing the scope of the claims.

[00:15:38] Speaker 01:
If we would note to the erased data, that wasn't the basis of the board's decision.

[00:15:43] Speaker 01:
So under the general, it's not a sufficient ground to,

[00:15:47] Speaker 01:
to affirm their decision.

[00:15:49] Speaker 01:
As I noted, there was a specifically disputed topic.

[00:15:52] Speaker 01:
As I recall, they tried to raise that issue at the board, and it was subject to a request for reconsideration that was denied.

[00:15:58] Speaker 03:
So it was not... Well, the board didn't need to go there because it disagreed with your claim, your newly revised claims instruction, right?

[00:16:07] Speaker 01:
It's possible, Your Honor, though I believe the board wouldn't have gone there because the petitioners saw it, they didn't put it in their petition, and they saw it in their request for reconsideration.

[00:16:16] Speaker 01:
to reconsider the scope, and that was denied.

[00:16:18] Speaker 01:
So I believe that would it be outside the scope of what they could have argued.

[00:16:22] Speaker 01:
So if it was remanded back, I don't believe it would be within the scope of what the case should be considered.

[00:16:33] Speaker 01:
I haven't mentioned it yet, but I think it is significant to point out the significant teaching away.

[00:16:38] Speaker 01:
I do know that it's a fact issue here, and that is obviously reviewed with substantial evidence.

[00:16:45] Speaker 01:
Even the petitioner doesn't try to defend the board's rationale for why it found that there wasn't teaching away.

[00:16:51] Speaker 01:
The Sharon reference, many times, over eight times, says that a significant part, an inventive part of the patent is to not reduce the size of the data.

[00:17:00] Speaker 01:
The board, in dismissing that argument, said that the disclosures in Sharon were directed to compression, and they need to consider them.

[00:17:08] Speaker 01:
But if you look at the Sharon reference, there are numerous statements about the need to not reduce the size of the data.

[00:17:14] Speaker 01:
that have nothing to do with compression.

[00:17:16] Speaker 01:
And in trying to defend the board's judgment, like I said, the petitioner doesn't argue the same grounds as the board, but they try to argue is that separate distinction that discussion in Sharon about don't shrink the size of the data is somehow only connected to written blocks and not to new blocks.

[00:17:34] Speaker 01:
And I think that shows that on the merits of looking at Sharon, there isn't really any evidence to establish that the limitation and the constant teaching

[00:17:44] Speaker 01:
is restricted to just the compression, because it's used so many places.

[00:17:49] Speaker 01:
And it's such a clear statement of the purpose of the invention is to reduce the size, is not reduce the size of the data.

[00:18:01] Speaker 01:
Are there any more questions, Your Honor?

[00:18:02] Speaker 01:
If not, no.

[00:18:03] Speaker 01:
Thank you.

[00:18:04] Speaker 01:
We thank both counsel and the cases submitted.