[00:00:00] Speaker 00:
Our first case for argument is 23-2133, Samsung versus Netlist.

[00:00:03] Speaker 00:
Sorry.

[00:00:04] Speaker 00:
Mr. Hollis, please proceed.

[00:00:09] Speaker 03:
Good morning.

[00:00:10] Speaker 03:
May it please the court?

[00:00:11] Speaker 03:
First, I'm glad to see the collegial clerk tradition continues.

[00:00:16] Speaker 03:
With regard to this appeal, I would start with section 103, which focuses on the claim subject matter.

[00:00:24] Speaker 03:
And here, the board did not focus on the claim subject matter.

[00:00:28] Speaker 03:
It looked at unclaimed details when it required bodily incorporation with respect to the two references it issued, rather than following this court's law and looking at whether the teachings of the references could be combined in view of the level of ordinary skill in the art.

[00:00:46] Speaker 03:
A level of ordinary skill in the art that I would note included familiarity with memory design and also with built-in self-test, neither of which the board referenced in its decisions.

[00:00:58] Speaker 03:
Samsung's simple combination of functional teachings was based on a first motivation of adding self-testing to a particular memory module, where, as noted, the level of skill included knowledge of self-testing.

[00:01:13] Speaker 02:
Mr. Haas?

[00:01:14] Speaker 03:
Yes, gentlemen.

[00:01:15] Speaker 02:
You said simple combination, and that sort of struck a nerve inside of me, because when I looked at the proposed combination,

[00:01:22] Speaker 02:
that you proposed, it seemed more like quite an elaborate proposal with multi-layers of modifications that would be required to Ellsbury's memory module in order to stick in the different pieces of the other references self-testing system in order to match up with all the claimed elements as arranged in the claim.

[00:01:50] Speaker 02:
I wish you wouldn't say simple combination, because I find it to be a distraction.

[00:01:57] Speaker 02:
Could we get into a little bit more detail about why I shouldn't think it's so complex, the proposal of modifying the Ellsbury reference with using different pieces of jet alone in different places in Ellsbury, and then of course, yet doing further modifications

[00:02:19] Speaker 02:
to the switches and the pathways, et cetera.

[00:02:22] Speaker 03:
Respectfully, Your Honor, I think we can approach that by looking at what the board said versus what Samsung actually said in its petition, which I think helps illustrate what the actual combination was that was put in front of the board.

[00:02:35] Speaker 03:
The board in the appendix on page 40 stressed that its analysis was with regard to whether portions of Jettilo's self-test module can be disassembled and distributed.

[00:02:47] Speaker 03:
across Ellsbury's memory module.

[00:02:50] Speaker 03:
Respectfully, that is very much the type of bodily incorporation that this Court has seen previously.

[00:02:56] Speaker 03:
But that doesn't match up with what Samsung actually proposed.

[00:03:00] Speaker 03:
And for that, I would turn the Court to the petition that Samsung had, which, and specifically to page 123 in the appendix, where Samsung, and my guess is this is the figure you're thinking of, Your Honor, when you refer to it being complex,

[00:03:17] Speaker 03:
and that's on 124 where you'll see the different portions in the different figures.

[00:03:23] Speaker 03:
But it's important to look at the language that Samsung used in describing what was going on in 124.

[00:03:31] Speaker 03:
Samsung talked about, specifically, the combination of the functionality of the particular block diagram components

[00:03:41] Speaker 03:
being incorporated into Ellsbury.

[00:03:44] Speaker 03:
And then again, at the bottom of page 123, the functionality of another set of components being incorporated into another part of Ellsbury.

[00:03:53] Speaker 03:
It wasn't a question of disassembling the module, taking particular pieces of circuitry.

[00:03:59] Speaker 03:
It was, in accordance with this court's usual law, a question of the functionality, the teachings.

[00:04:05] Speaker 03:
And it's important in looking at the claim subject matter to realize that the issues that the board focused on aren't in the claim.

[00:04:13] Speaker 03:
If you actually look at the specific claim, which of course is on the inside of both the blue brief and the red brief, part of the claim that we're looking at is the very last clause, which is a where-in clause describing the second mode.

[00:04:29] Speaker 03:
In each part of that clause, it is a device configured to, or a module configured to, or logic elements configured to.

[00:04:38] Speaker 03:
So very broad terms, focusing only on functionality.

[00:04:44] Speaker 03:
So there was really no teaching of specific timing or synchronization or particular types of circuitry that are required by this claim.

[00:04:53] Speaker 03:
Instead, the claim just asked for functionality.

[00:04:56] Speaker 03:
And it relied on the person with the ordinary skill in the art

[00:04:59] Speaker 03:
to implement that functionality because, and it's undisputed, there's no teaching in the 5-2-3 patent of the details that the board was focused on.

[00:05:10] Speaker 03:
In other words, the patented issue relied on the person of ordinary skill in the art to implement the very functionality that Netlist now relies on in saying that the person of ordinary skill in the art wouldn't have found this obvious.

[00:05:22] Speaker 03:
If the person of ordinary skill in the art was able to implement the functionality with respect to these very broad functional claims,

[00:05:29] Speaker 03:
They certainly could do so with respect to the combination that was proposed by Samsung.

[00:05:35] Speaker 03:
And the reason I say it's simple, Your Honor, is because Jettilo is an example of built-in self-test, which is both standardized under the JEDEX standards, but also part of the skill of the person, ordinary skill in the art.

[00:05:50] Speaker 03:
That's something that sets this aside, or sets it apart, frankly.

[00:05:55] Speaker 03:
from what we see in other obvious instances.

[00:05:58] Speaker 00:
I'm struggling with that argument.

[00:06:00] Speaker 00:
I mean, your petition relied on Jettilo, right?

[00:06:04] Speaker 00:
And Jettilo supplied the self-test and you put it in

[00:06:09] Speaker 00:
inserted it into Ellsbury by basically dividing the functionality and the associated components that drove that functionality and locating them in two different parts of Ellsbury, right?

[00:06:22] Speaker 00:
I mean, that's what you've asked for and suggested renders these claims obvious, correct?

[00:06:29] Speaker 03:
I agree with you, Your Honor.

[00:06:31] Speaker 03:
In accordance with Jettilo, we said you should implement the testing

[00:06:35] Speaker 03:
in accordance with the normal operation of the memory module.

[00:06:38] Speaker 03:
And for that, I would point you that Jettilo specifically suggests that, and that is in column three at lines, looks like 47 through 51, where Jettilo says, you should self-test in the same way that normal operation occurs.

[00:06:54] Speaker 03:
So, in Ellsbury, you have data going in one slot, that's where the data flows, and you have addressing and controlling occurring through a different circuit.

[00:07:03] Speaker 03:
Jedila teaches that when you test, you should test in the same way as normal operation.

[00:07:08] Speaker 03:
Otherwise, you're not really testing the memory module.

[00:07:11] Speaker 03:
And so the data for testing should be in the data path.

[00:07:14] Speaker 00:
So maybe the problem is, I mean, even in your brief to us, I mean, look, this is the figure you gave us for what easy to amount to their, you know, this one here?

[00:07:27] Speaker 00:
Yes.

[00:07:28] Speaker 00:
That's their figure, yeah.

[00:07:32] Speaker 00:
I tend to agree with Judge Chen that even in your presentation to us, you made it look anything but simple.

[00:07:41] Speaker 00:
And given that this is a substantial evidence review on our part, I don't see how we overturn the board's conclusion that you are asking to have the teachings of Jettilo

[00:07:54] Speaker 00:
chopped up and redistributed within Ellsbury.

[00:07:57] Speaker 00:
And that's really what they said.

[00:07:58] Speaker 00:
That's what they held.

[00:08:00] Speaker 00:
And that seems completely consistent with what you've argued to us in your brief.

[00:08:04] Speaker 00:
So given that, it's hard for me to conclude that there isn't substantial evidence for their conclusion, since I'm not a skilled artisan and I review for substantial evidence.

[00:08:14] Speaker 00:
They concluded that all of this was too much.

[00:08:20] Speaker 00:
Looks like a lot.

[00:08:22] Speaker 03:
Well, respectfully, Your Honor,

[00:08:23] Speaker 03:
There is a difference between disassembling and dividing circuitry and instead taking functionality that the person or a skill only understands and implementing that using what is an agreed level of skill that includes circuit design and built-in self-test.

[00:08:39] Speaker 03:
The board did the former.

[00:08:41] Speaker 03:
And this court has a long line of cases saying,

[00:08:44] Speaker 03:
You don't require that the various circuitry, the bodily incorporation.

[00:08:49] Speaker 00:
Where did they say the circuitry?

[00:08:52] Speaker 03:
That would be on page 40 of the appendix, Your Honor.

[00:08:54] Speaker 00:
Portions of general self-test module cannot be disassembled and distributed across Ellsbury's

[00:09:01] Speaker 00:
memory module as proposed by petitioner.

[00:09:03] Speaker 00:
But that's exactly what you submitted in your petition and actually in your brief to us.

[00:09:09] Speaker 00:
This picture, that's why I held it up.

[00:09:11] Speaker 00:
You show in your brief to us the chopping up of those components and inserting them in different places.

[00:09:19] Speaker 00:
So why would the board have erred in assuming that that was the argument that you intended since it's in your petition, in your pleadings to the board, in your brief to us?

[00:09:28] Speaker 00:
It seems like

[00:09:29] Speaker 00:
You have to live and die by the argument you chose to make.

[00:09:33] Speaker 03:
Respectfully, Your Honor, as I pointed out in the petition on page 123, we were very specific that it was the functionality, not the circuitry, that needed to be incorporated.

[00:09:44] Speaker 03:
And we said that twice, once with respect to the functionality going into the data path, and once with respect to the functionality going into the control and address path.

[00:09:55] Speaker 03:
So on page 123, we talk about the functionality.

[00:09:58] Speaker 02:
I guess the question I have is, why would it be wrong for the board and the patent owner to respond to your proposal

[00:10:10] Speaker 02:
insert different functional blocks into different places in Ellsbury by saying, well, it's unclear why a skilled artisan would have the instinct to make all of these different choices when a skilled artisan would recognize that

[00:10:28] Speaker 02:
there would be a lot more to it than just a simple plug and play of different functional blocks.

[00:10:34] Speaker 02:
There would be timing and synchronization issues once you disaggregate all the pieces of Jettilo and move them into different places in Ellsbury.

[00:10:45] Speaker 02:
And likewise, you have to do further

[00:10:49] Speaker 02:
redesigns to the ASIC switches in Ellsbury, and then also you have to modify certain pathways inside of Ellsbury in order to do all these different things, to isolate the system controller from the memory devices during test mode.

[00:11:08] Speaker 02:
And so therefore, it starts to feel more and more like we're using the claim itself as a roadmap in order

[00:11:15] Speaker 02:
to make all these different iterative choices.

[00:11:20] Speaker 03:
Respectfully, Your Honor, the roadmap here is definitely Ellsbury.

[00:11:24] Speaker 03:
Ellsbury identified a memory module that had both the data with a path and then a separate control address path.

[00:11:34] Speaker 03:
And all Samsung suggested was that the person who was still in the yard who has knowledge of built-in self-tests would say, okay, we have a data module here.

[00:11:43] Speaker 03:
One of the things under the standards that everyone agrees to is that we do have to do some testing of that, right?

[00:11:48] Speaker 03:
You don't want to send that out into the world without any way of making sure it works.

[00:11:53] Speaker 03:
Jettolo itself, as I pointed out in column three, says you've tested in accordance with its normal operation, separated data from control and addressing.

[00:12:04] Speaker 03:
So we took, you know, Samsung looked and said, okay, well if you take that functionality of Jettolo,

[00:12:09] Speaker 03:
and you do what Jettilo says, which is test in accordance with normal operation, you would take the portions of the functionality that go to the data, you put it with the data path, and you take the portions of the functionality that go to the addressing and command, and you put it where the addressing and command are.

[00:12:26] Speaker 03:
That followed from what Sam Brom proposed, which was a two-step motivation.

[00:12:32] Speaker 03:
You first have a motivation to add self-testing.

[00:12:37] Speaker 02:
adding the multiplexers and demultiplexers into the switches?

[00:12:42] Speaker 02:
I mean, why would someone do that?

[00:12:46] Speaker 03:
So respectfully, your honor, if you'll look at Jedlo, and this is specifically the figure that everyone looks at in Jedlo, which is on appendix 1176, right in the middle of the figure is a switch, and that's switch 260.

[00:13:00] Speaker 03:
Jettilo teaches specifically that that switch can be a multiplexer.

[00:13:05] Speaker 03:
It's one of the three embodiments that Jettilo gives for that switch.

[00:13:09] Speaker 03:
And that's on column seven at line 30 and 31.

[00:13:13] Speaker 03:
The switch can also be a set of multiplexers.

[00:13:16] Speaker 03:
So this is specific old element teaching that Jettilo already said.

[00:13:21] Speaker 03:
This is what you can do.

[00:13:23] Speaker 03:
So we were not creating anything new.

[00:13:25] Speaker 03:
We were not asking for any new circuitry.

[00:13:27] Speaker 03:
We were saying Jettilo teaches doing this.

[00:13:29] Speaker 03:
So you do it in accordance with normal operation, which Jettilo also teaches at Column 3.

[00:13:35] Speaker 03:
So this really was an old element set up.

[00:13:38] Speaker 03:
Yes, you have to take this old element and put it in normal operation where it goes.

[00:13:44] Speaker 03:
Data goes here.

[00:13:46] Speaker 03:
This element, you put in normal operation where it goes.

[00:13:48] Speaker 03:
Control and address go here.

[00:13:50] Speaker 03:
But for 40 years since Henry Eder, this court has been careful to say, focus on the claim subject matter.

[00:13:57] Speaker 03:
And the issues the court focused on

[00:13:59] Speaker 03:
They aren't claimed.

[00:14:00] Speaker 03:
They're not even disclosed in the 523 patent.

[00:14:03] Speaker 03:
So we have this situation where we have a patentee who's saying, oh, you can't do this because of timing and synchronization and the issues they raise, none of which they taught in their own patent.

[00:14:19] Speaker 03:
As this court put it well in Lockwood, why would you focus on issues that the patentee said were in the skill in the art

[00:14:27] Speaker 03:
in saying that an obvious miscombination is happening.

[00:14:29] Speaker 00:
Well one of your problems is they found your expert to be conclusory and they didn't trust his testimony.

[00:14:35] Speaker 00:
They made a clear credibility finding and they found the patentee's expert to be credible and the patentee's expert said what you're asking for creating these new pathways in the architecture in Ellsbury that would be required is outside

[00:14:50] Speaker 00:
It's too complex to be considered obvious.

[00:14:52] Speaker 00:
And they credited that.

[00:14:54] Speaker 00:
And that was a fact-minding based on credibility.

[00:14:56] Speaker 00:
They expressly found your expert not credible, and they expressly found the patentee's expert credible on whether these modifications take this outside the range of obviousness.

[00:15:06] Speaker 00:
We can't touch that on appeal.

[00:15:08] Speaker 00:
That's just not our business.

[00:15:09] Speaker 03:
Your Honor, we're not asking you to.

[00:15:12] Speaker 03:
because those findings are irrelevant to the legally appropriate question, which is whether the teachings match up with the claim subject matter.

[00:15:23] Speaker 03:
What the board was focusing on was findings.

[00:15:24] Speaker 00:
You've used all of your time in your rebuttal time.

[00:15:26] Speaker 00:
Why don't we save and give you, I'll restore a little bit of rebuttal time, but let's move on because you're definitely out of time.

[00:15:35] Speaker 00:
Mr. Milliken, please proceed.

[00:15:46] Speaker 01:
Thank you, Your Honors.

[00:15:48] Speaker 01:
This is not a bodily incorporation case.

[00:15:51] Speaker 01:
Samsung lost it on the facts.

[00:15:54] Speaker 01:
Samsung did not have any prior art even suggesting testing that is aggregated in the way that the claims lay out.

[00:16:01] Speaker 01:
And so Samsung picked out Ellsbury, which doesn't talk about testing at all, and then tried to combine it with Jettolo, which discloses a unitary self-test module and, as the board found, has a completely different architecture than Ellsbury.

[00:16:15] Speaker 01:
Samsung then proposed a contorted theory, and it was very clear in its petition about what it was proposing.

[00:16:22] Speaker 01:
This is at appendix 122 to 23.

[00:16:24] Speaker 01:
This theory requires slicing up Jell-O's testing module, distributing the pieces across Ellsbury, and then adding circuitry that neither reference discloses, all to get to a memory module that reads on the claims.

[00:16:40] Speaker 01:
And the board, in a thorough analysis.

[00:16:42] Speaker 01:
Conceptually,

[00:16:45] Speaker 02:
What is wrong with what could be a common sense view of Ellsbury that, assuming you want to have a test mode in Ellsbury, and it seems like there's decent grounds to want to have a test mode in any memory module, why wouldn't you apply the

[00:17:08] Speaker 02:
the control address signals of Jettilo in the control unit of Ellsbury, and then the data signals for the test mode from Jettilo into the data module of Ellsbury.

[00:17:23] Speaker 02:
Just conceptually speaking, it seems to make some sense.

[00:17:27] Speaker 01:
You keep apples to apples and oranges with oranges.

[00:17:31] Speaker 01:
Right.

[00:17:31] Speaker 01:
Understood, Your Honor.

[00:17:33] Speaker 01:
So the major problem with that conceptual argument

[00:17:36] Speaker 01:
is that as the board found at appendix 51 to 54, the prior art did disclose disaggregated or distributed architectures like Ellsbury that implemented self-testing.

[00:17:48] Speaker 01:
But they did self-testing in a different way than the claims require.

[00:17:53] Speaker 01:
Namely, they either put all of the self-testing functionality in what's equivalent to Ellsbury's control ASIC,

[00:18:01] Speaker 01:
or they duplicated all of the self-testing functionality and put it in each one of what's equivalent to Ellsbury Switch A6.

[00:18:10] Speaker 01:
And so there were distributed systems out there that implemented self-testing.

[00:18:15] Speaker 01:
There just wasn't one that did it in the way the claimed invention does.

[00:18:18] Speaker 01:
And so that reinforces

[00:18:20] Speaker 01:
the board's finding that the only motivation that Samsung really came up with here was just pure hindsight.

[00:18:27] Speaker 01:
They were using the claims to try to piece together the prior art in a way that would read on them.

[00:18:33] Speaker 01:
And so the board found that there was sort of no motivation to begin with, and then beyond that, found that the skilled artisan who was defined here to have an engineering degree and one year of experience

[00:18:47] Speaker 01:
would simply not be motivated to do this because it would entail so many complications.

[00:18:54] Speaker 01:
Both experts agreed that this combination would require redesigning Ellsbury and adding new circuitry.

[00:19:00] Speaker 01:
Both experts also agreed that no one had ever tried to disaggregate testing functionality in this way, even though, as I was discussing earlier, Judge Chin,

[00:19:10] Speaker 01:
Both testing and distributed architectures have been around a long time and the board credited our experts testimony that this proposed redesign of Ellsbury would create intractable complications.

[00:19:22] Speaker 01:
The combination would require Ellsbury's control ASIC and switch ASICs to do things that they simply weren't designed to do.

[00:19:30] Speaker 01:
And the board found that this would cause all kinds of complications.

[00:19:35] Speaker 01:
It would introduce new signals going through new components and new directions.

[00:19:39] Speaker 02:
It would lead to... But all of that new stuff that would be added to Ellsbury, that's just a necessary byproduct of once you decide you want to have a test mode.

[00:19:51] Speaker 02:
in Ellsbury.

[00:19:52] Speaker 02:
Of course there's going to be new signals, of course there's going to be new elements and functionality inside of Ellsbury.

[00:19:59] Speaker 02:
You have to do that in order to have a test mode.

[00:20:02] Speaker 02:
So just saying you have to do new things to Ellsbury in order to get the test mode by itself is not enough without persuasive arguments.

[00:20:12] Speaker 01:
Well, the board found that if one were to want to implement testing in Ellsbury, one would have done it in one of the ways that the prior art taught, which the board found were simpler than the way that Samsung proposed.

[00:20:24] Speaker 01:
That's a factual issue.

[00:20:25] Speaker 01:
The board just found as a matter of fact, if the skilled artisan were motivated to add testing, they wouldn't have done it in the way that Samsung's combination proposed.

[00:20:33] Speaker 01:
That finding enjoys substantial evidence support.

[00:20:38] Speaker 01:
I'll just go back briefly to bottling corporations since my friends spent a significant amount of time on it.

[00:20:45] Speaker 01:
This argument fails factually and legally.

[00:20:48] Speaker 01:
Factually, the board didn't reject Samsung's argument on the basis that you couldn't physically put these two things together without modifying them.

[00:20:55] Speaker 01:
If the board were going to do that, the board could have stopped its analysis at page 34 once it reached Samsung's concession that you did have to redesign things to do this.

[00:21:06] Speaker 01:
The board went on for 20 more pages and explained why this specific redesign would not be obvious because it involved too many complexities and technical difficulties.

[00:21:17] Speaker 01:
That's consistent with the law, which requires a showing of why a skilled artisan would not just could make a proposed combination.

[00:21:26] Speaker 01:
Bodily incorporation is not a magic wand that you can wave on appeal to save a tortured combination of references.

[00:21:33] Speaker 01:
The complexity of an obviousness argument is highly relevant to whether it will succeed.

[00:21:40] Speaker 01:
This court's bodily incorporation cases say that

[00:21:43] Speaker 01:
The absence of physical combinability standing alone doesn't defeat obviousness, but they decidedly do not say that any combination is obvious no matter how complicated and technically problematic if you can somehow piece it together with snippets of prior art.

[00:22:00] Speaker 01:
If the court has any questions, I'd be happy to answer them.

[00:22:03] Speaker 01:
Otherwise, I'll rest on the briefs and cede the balance of my time.

[00:22:07] Speaker 00:
Thank you, counsel.

[00:22:08] Speaker 00:
We'll restore two minutes of rebuttal time.

[00:22:17] Speaker 03:
Your Honor, we agree it's not a magic wand.

[00:22:20] Speaker 03:
Bodily incorporation needs to be applied to focus on the claim subject matter.

[00:22:26] Speaker 03:
The one thing you didn't hear at all in that argument was any reference to how any of the details that the board focused on are in the claims, not once.

[00:22:36] Speaker 03:
And there's a good reason that section 103 says it's the claim subject matter that needs to be obvious.

[00:22:42] Speaker 03:
This court's bodily incorporation law protects that requirement of the statute.

[00:22:47] Speaker 03:
says don't look at a bunch of unclaimed details, look at what the claims require.

[00:22:54] Speaker 03:
And that's why it's important that we focus on the fact that the board required that you disassemble and divide the actual circuitry.

[00:23:03] Speaker 03:
Because Samsung said no, you look at the functionality and that the person of our needs to go in the art, understanding that functionality, would be able to implement the functionality in Ellsbury.

[00:23:15] Speaker 03:
And I'll point one other thing,

[00:23:17] Speaker 03:
Collie said that the skilled artisan had one year of engineering experience.

[00:23:22] Speaker 03:
He left out the entire second half of the person of ordinary skill, which is on appendix page 10.

[00:23:28] Speaker 03:
And that is that that person would be familiar with module design, that that person would be familiar with built-in self-test.

[00:23:35] Speaker 03:
There were specific skills and knowledge that were included in that level of ordinary skill that you didn't hear referenced in the argument you just heard.

[00:23:46] Speaker 03:
Finally, I would point out that he started with the idea that, well, we don't have testing in prior art with the distributed data design.

[00:23:54] Speaker 03:
And we don't.

[00:23:54] Speaker 03:
This is not an anticipation case.

[00:23:57] Speaker 03:
But the reality is, if that's your test, there's never gonna be an obviousness case.

[00:24:01] Speaker 03:
Because only anticipation would have included testing in the distributed point with data, with the data, and with the control and address, with the control and address.

[00:24:12] Speaker 03:
Yes, if we had had that piece of prior art,

[00:24:14] Speaker 03:
We probably would have brought it, and it would have been an anticipation case.

[00:24:17] Speaker 03:
But the fact that there's no anticipation case does not mean that the board is entitled to require unclaimed elements as part of its analysis in terms of assessing whether a obviousness combination has a motivation to come on.

[00:24:33] Speaker 03:
That's not the test.

[00:24:35] Speaker 03:
Anticipation is not the test.

[00:24:37] Speaker 03:
Unclaimed details is not the test.

[00:24:39] Speaker 03:
Thank you for restoring my time, Your Honor.

[00:24:41] Speaker 00:
I thank both counsel.

[00:24:42] Speaker 00:
This case is taken under submission.