`
`
`
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`____________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`____________
`
`HEWLETT-PACKARD ENTERPRISE CO.; HP ENTERPRISE
`SERVICES LLC; and TERADATA OPERATIONS, INC.,
`Petitioner,
`
`v.
`
`REALTIME DATA LLC,
`Patent Owner.
`____________
`
`Case IPR2016-00783
`Patent 6,597,812 B1
`____________
`
`Record of Oral Hearing
`Held: June 30, 2017
`____________
`
`
`
`
`BEFORE: GEORGIANNA W. BRADEN, J. JOHN LEE, and
`JASON J. CHUNG, Administrative Patent Judges.
`
`
`
`
`
`
`
`
`
`
`
`
`
`Case IPR2016-00783
`Patent 6,597,812 B1
`
`APPEARANCES:
`
`ON BEHALF OF THE PETITIONER:
`J. CHRISTOPHER CARRAWAY, ESQUIRE
`Klarquist Sparkman, LLP
`One World Trade Center
`121 SW Salmon Street, Suite 1600
`Portland, Oregon 97204
`
`and
`
`JAMIE R. LYNN, ESQUIRE
`Baker Botts, LLP
`1299 Pennsylvania Avenue, N.W.
`Washington, D.C. 20004
`
`ON BEHALF OF PATENT OWNER:
`KAYVAN B. NOROOZI, ESQUIRE
`WILLIAM P. ROTHWELL, ESQUIRE
`Noroozi, P.C.
`1299 Ocean Avenue, Suite 450
`Santa Monica, California 90401
`
`
`The above-entitled matter came on for hearing on Friday,
`
`June 30, 2017, commencing at 1:01 p.m., at the U.S. Patent and
`Trademark Office, 600 Dulany Street, Alexandria, Virginia.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
` 2
`
`
`
`Case IPR2016-00783
`Patent 6,597,812 B1
`
`
`P R O C E E D I N G S
`- - - - -
`JUDGE CHUNG: This is the hearing for case
`IPR2016-00783 pertaining to the U.S. patent number 6,597,812.
`With me on the panel are Judge Georgianna Braden, who is
`sitting remote, and Judge John Lee, and myself, Jason Chung.
`Who do we have for petitioner?
`MR. CARRAWAY: Petitioner, Your Honor, we have
`lead counsel, Jamie Lynne, and I am backup counsel, Chris
`Carraway.
`JUDGE CHUNG: Who do we have for patent owner?
`MR. NOROOZI: We have lead counsel, Mr. William
`Rothwell, and myself, Kayvan Noroozi.
`JUDGE CHUNG: Each side will have 30 minutes to
`present their argument. Petitioner will go first. Petitioner may
`reserve part of that 30 minutes as rebuttal time. Would petitioner
`like to reserve any of that rebuttal time at this moment?
`MR. CARRAWAY: Yes, Your Honor, seven minutes,
`
`please.
`
`JUDGE CHUNG: And the panel would like to remind
`the parties respectfully to speak into the microphone clearly
`because one of the judges is remote, and also identify the slide
`number when referring to -- when the slide is in the presentation
`so that our remote judge can follow along to the conversation.
`At this moment, the petitioner may begin.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`
`
`
`
` 3
`
`
`
`Case IPR2016-00783
`Patent 6,597,812 B1
`
`
`MR. CARRAWAY: May it please the Board, I am
`again Chris Carraway, counsel for the three petitioners,
`Hewlett-Packard Enterprise, Hewlett-Packard Enterprise
`Services, and Teradata. On slide 2, we have the grounds raised
`by this petition and on which the Board has instituted the IPR for
`the '812 patent. Ground 1 addresses method claims 1 to 4 and 8
`and system claim 28 using O'Brien and Nelson. Ground 2
`addresses claims 14 to 17 and 21, which are CRM claims for
`software that performs the same steps as method claims 1 to 4
`and 8.
`
`The addition of the Welch prior art for ground 2 is
`because O'Brien uses hardware circuitry to perform its
`compression method, and Welch shows that this type of
`compression can be implemented on either hardware or software.
`Because these two software claims track the method claims, most
`of the disputes for ground 2 are going to be the same as ground 1.
`We'll move to slide 5, please. On slide 5, I want to start
`by just quickly doing an overview of the '812 patent and the
`O'Brien patent, which is the primary prior art reference at issue.
`The '812 patent is directed to a combination of two compression
`techniques, run length encoding and dictionary encoding. Run
`length encoding and codes are a run of characters like five As in a
`row. Dictionary compression instead maps characters and strings
`to an index or code in the dictionary. Now, the '812 patent admits
`that both of these compression techniques were old and then
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`
`
`
`
` 4
`
`
`
`Case IPR2016-00783
`Patent 6,597,812 B1
`
`purports to try to claim the combination, but the combination was
`old as well. That is shown by the O'Brien patent.
`Slide 5 shows overview diagrams for these two patents
`in both. An input is monitored by a run length encoder which is
`shaded orange on slide 5. Both of these look for runs of a certain
`number of characters. In O'Brien it's three. And in both if such a
`run is encountered, it is encoded using three components, a code
`telling the character that is being repeated, a code indicating that
`the -- telling the decoder that what is coming is going to be a run
`length and should be treated that way, and three, the number of
`times to repeat. And then in both --
`JUDGE CHUNG: Excuse me. Can we spend some
`time on the second thing that you discussed, the code telling the
`run length that a run length is approaching. Would you be able to
`point to me in your petition where you map reference value to the
`control code word and how reference value indicates that a run is
`approaching. Reference value from Table 8 that is the O'Brien
`reference.
`MR. CARRAWAY: Yes, Your Honor. I'm pulling up
`Exhibit 1005, which is the first Creusere declaration filed with
`our petition. Your Honor, this is page 51, paragraph 80 of
`Dr. Creusere's declaration. In it Dr. Creusere explains how
`O'Brien's compression system encodes a run length by three
`things: A reference value that represents the character being
`repeated, a run length reference value selected from table A
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`
`
`
`
` 5
`
`
`
`Case IPR2016-00783
`Patent 6,597,812 B1
`
`which indicates a run length that's been detected and indicates the
`number of bits used for the repeat count and then the repeat count
`for the run.
`Now, if you look down, we cited to column 5, lines 45
`to 50. And in our demonstratives we cited to column 3 for this
`very same point. It's -- the sentence is repeated. I just used the
`wrong location from where it was in our petition, which was 5:45
`to 50.
`
`JUDGE CHUNG: So in the demonstrative you cited to
`the new column 3 is what you're saying?
`MR. CARRAWAY: I accidentally used the first time
`the same sentence appears instead of the second time the sentence
`appears. It's the exact same sentence. In that it says runs of three
`or more repeated bytes are encoded using three things: A
`predetermined set of reserve reference values to indicate that the
`preceding character was repeated a number of times specified by
`the repeat code. That's three values, all of which are identified in
`O'Brien as being used.
`I want to go back to -- go into more detail about this
`particular element. On slide 18 when it comes up, we have the
`first of those three components, the code word signifying the
`character being repeated. And column 9, lines 10 to 47 of
`O'Brien sets the stage by explaining how run length encoding
`starts. The run length encoder circuit 303 monitors three data
`bytes looking for a run. It monitors the data byte that's in the
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`
`
`
`
` 6
`
`
`
`Case IPR2016-00783
`Patent 6,597,812 B1
`
`reference value encoder, it monitors what's in the holding register
`behind that, and then it monitors what's in the path behind the
`holding register. When it identifies that there's three in a row,
`this is where this segment of the O'Brien patent that's on slide 18
`states when all of these inputs are the same, run length encoder
`circuit 303 transmits a run length detect signal to reference value
`encoder 304 via leads 313 which then encodes the last character it
`received and then transfers control of the run length encoder
`circuit 303. So that's where we know that the character has been
`encoded by the reference value.
`JUDGE CHUNG: So that detect signal from or via
`circuit 303 is what triggers the reference value?
`MR. CARRAWAY: The reference value for the
`character because of course the characters are stored.
`JUDGE CHUNG: And the reference value being the
`control code word, mapped to the claim control code word?
`MR. CARRAWAY: That's the code word for the value
`in the dictionary for the character. So again there's three words
`that are required by the claim. One is a control code word
`indicating that a run length is coming. That's telling the decoder.
`There is the character itself and there is the repeat code. Now, in
`O'Brien the order is switched.
`JUDGE CHUNG: So you are saying that the claim, if I
`may paraphrase, is ABC and then in O'Brien it's BAC?
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`
`
` 7
`
`
`
`Case IPR2016-00783
`Patent 6,597,812 B1
`
`
`MR. CARRAWAY: I think that's right, yes, Your
`Honor. The first thing that's encoded is shown here in column 9.
`And that is the character, it's an instruction from the run length
`encoder to the reference value encoder saying encode that
`character that you currently have in the encoder.
`JUDGE CHUNG: And the reference value is the first
`column of table A, and you are mapping that reference value to
`element A, if I may state -- the control word.
`MR. CARRAWAY: Yes, Your Honor. While we are
`on this outputting, I want to go through the other ones. So now
`we get to what in the claim is A, the control code word. And
`what we have here is that looking at slide 19, the next two words
`required are the control code word and then the repeat code.
`Now, O'Brien has a table A at column 13 which O'Brien
`described as an example of a series of reserved reference values.
`Those reserved reference values are, again -- the word reserve
`reference value was discussed in the portion of column 5 that we
`saw at the beginning. Now, conveniently enough, there they are.
`They say reference values, so we know those are the reference
`values in table A. And they go 1, 2, 3, 4, all the way down to 10,
`as an example.
`Now, the way that table A works, first of all, it's very
`clear there's two values here. There's a reference value in
`column 1 and a repeat value in column 2. Each reference value
`indicates that the repeat will be in a certain range using a certain
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`
`
`
`
` 8
`
`
`
`Case IPR2016-00783
`Patent 6,597,812 B1
`
`number of bits to encode. And the bits correspond to the
`reference value in this example.
`So just looking at the first row in table A, we have for a
`reference value of 1, that can be used to encode a repeat of 2 or 3.
`And that's in the far right column of table A using repeat codes of
`0 or 1. And so specifically the way that would work would be if
`you have a run of -- a repeat of 3, you would have a reference
`value of 1, a repeat code of 1, and that would encode a repeat of 3
`using two different codes.
`Now, it's more complicated in the '812 patent. Now, the
`way the '812 patent does it is they use one code and one code
`only. It's listed as D[1] in the '812 patent. That just indicates a
`run period. It's not a run of any range like the O'Brien patent
`does it. O'Brien explains why it does these multiple references
`instead of using one. And that's because it allows you to reduce
`the number of bits needed to encode a run. So for a more
`common repeat like 2 or 3, you can encode that using only
`one-bit number, 0 or 1, while less common repeats like 8 to 15
`would require three bits. So that way for the most common
`situations, you are going to be using fewer bits.
`Now, patent owner asserts that O'Brien only outputs
`two words. Not three. And that's just not true. And the only way
`you can get to that is by misreading O'Brien. The way that patent
`owner seems to get to that conclusion appears to be by assuming
`that that reference value in the first column of table A is the
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`
`
`
`
` 9
`
`
`
`Case IPR2016-00783
`Patent 6,597,812 B1
`
`character that's being encoded. That can't be right because as we
`saw in column 5, it stated that the reference values are reserve
`reference values that indicate that the preceding character is going
`to be repeated with a repeat code.
`So there's also the question of the order. It hasn't really
`been pushed in the demonstratives, but it's fully briefed in our
`petition and in the reply. So I don't want to belabor that point.
`Your Honor, if we can go to the other issue which is --
`the other main issue which is maintaining a dictionary -- actually,
`before we go to maintaining a dictionary, I did want to hit a
`threshold point that has been raised in the demonstratives by the
`patent owner, and that is the question of obviousness in general
`versus anticipation.
`Petitioner -- or patent owner seems to say that if there
`are no differences between the challenged claims and one of the
`references in the obviousness ground, that there can't be
`obviousness. And that's just not true. First of all, I want to
`explain what we did in the petition. In the petition, we cited
`O'Brien for all the elements except for the dictionary. And the
`reason why we didn't actually use that for a dictionary was
`because, and we were very clear about this in the petition, it has a
`dictionary. It just doesn't label it a dictionary. And what we used
`the Nelson for which is the secondary reference was for that
`label. It was for a teaching to one of ordinary skill in the art that
`what was in O'Brien is actually a dictionary.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`
`
`
`
` 10
`
`
`
`Case IPR2016-00783
`Patent 6,597,812 B1
`
`
`We were not arguing primarily in our petition that we
`somehow need to put together O'Brien and Nelson into some kind
`of a combination. Instead, we were using the teachings of the
`two. That's all.
`Patent owner now admits that O'Brien does have a
`dictionary. So that missing element that we were using Nelson to
`show how a person would understand it is now not necessary.
`The ground is still obviousness over O'Brien and Nelson. It's just
`that Nelson's teachings are not necessary to understand O'Brien.
`Now, patent owner argues that anticipation by O'Brien
`forecloses obviousness. That would lead the Board into legal
`error because it's just not the law.
`JUDGE CHUNG: Counselor, can we just move on to
`maintaining a dictionary. If you have time at the end, maybe you
`can revisit this.
`MR. CARRAWAY: Yes, Your Honor. If we can go to
`slide 10, maintaining a dictionary, the dispute between the parties
`is largely one of claim construction. It's whether -- well, the
`petitioners have provided a plain meaning, and patent owner is
`trying to read in a limitation from the embodiment.
`In our reply, as shown on slide 11, in our reply, the
`petitioners refuted that construction and expressly set forth the
`plain and ordinary meaning that Dr. Creusere had used just when
`applying the claim terms themselves in its original declaration.
`Specifically, Dr. Creusere's declaration did three important things
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`
`
`
`
` 11
`
`
`
`Case IPR2016-00783
`Patent 6,597,812 B1
`
`with respect to the term maintaining. He explained that it was not
`a term of art in the compression field and did not have a special
`meaning. He noted that the plain meaning in a standard
`dictionary was just this, to keep an existing state repair. And he
`analyzed how that term was used in the '812 patent to make sure
`he understood how it was actually being used in its plain and
`ordinary meaning.
`JUDGE CHUNG: So that wasn't a supplemental
`declaration, correct?
`MR. CARRAWAY: This was in the reply declaration,
`yes, Your Honor.
`JUDGE CHUNG: So filed after the initial declaration?
`MR. CARRAWAY: It was. In the original declaration,
`the petitioners identified four or five terms that appeared to need
`construction. Maintaining a dictionary, Dr. Creusere explained
`how that applied to O'Brien, but he didn't label a construction per
`se because it was not raised -- for the first time it was raised in the
`response.
`JUDGE CHUNG: Okay.
`MR. CARRAWAY: As shown on slide 12 --
`JUDGE CHUNG: If I may interject, if we were to --
`let's say hypothetically speaking, if we were to adopt patent
`owner's definition which, I mean, may be problematic because of
`dependent claims 4 and 5, but say for a moment hypothetically
`we were to adopt it. Would the O'Brien -- would the combination
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`
`
`
`
` 12
`
`
`
`Case IPR2016-00783
`Patent 6,597,812 B1
`
`of O'Brien and Nelson teach maintaining a dictionary pursuant to
`patent owner's definition?
`MR. CARRAWAY: Yes, Your Honor. And the reason
`being that the patent owner has staked its distinction of O'Brien
`from the '812 patent on the idea that O'Brien segments data
`before it compresses it. And the problem with that is that there
`are cases where the segment may actually be smaller -- sorry,
`larger than the input stream. If that's the case, then the input
`stream is completely done before the end of the segment. And it
`would be just like any other situation where you have an input
`stream and the full thing is compressed.
`JUDGE CHUNG: Okay. Because it's a 103 here, then
`the question is obviousness, okay.
`MR. CARRAWAY: Yes, Your Honor. So I want to
`just explain why the -- go over briefly why the patent owner's
`limitation is incorrect. If we can go to -- so the patent owner has
`proposed this preferred embodiment construction that requires
`that the maintaining a dictionary is only until either the input data
`stream has been compressed or the dictionary is full. There is no
`basis -- well, there's three reasons why this is wrong. First of all,
`there's no basis to jam the details of a preferred embodiment into
`this claim term maintaining a dictionary. The patent owner hasn't
`said there's any special meaning to the term. The patent owner
`hasn't identified anything in the file history or the specification
`that would be a disavowal or lexicography. And Federal Circuit
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`
`
`
`
` 13
`
`
`
`Case IPR2016-00783
`Patent 6,597,812 B1
`
`canon is clear that absent these two things, you just don't read
`preferred embodiments into the claim.
`It's also inconsistent with other claim language. Claims
`1 and 14 preamble recite a lifetime limitation already
`compressing input data comprising a plurality of data blocks,
`which is a plurality of bytes. And then you have the maintaining
`a dictionary element itself which states that it only requires
`maintaining a dictionary comprising a plurality of code words.
`So the challenged claims don't require the dictionary to be large
`and they don't require it to be full. They only require
`compressing at least two data blocks and maintaining a dictionary
`of at least two code words.
`And then in addition, there is the claim differentiation
`problem for the plaintiff -- sorry, for the patent owner. And that
`is that claim 5 already includes this idea of retaining the
`dictionary until it's full and then resetting it.
`JUDGE CHUNG: With respect to your proposed
`definition, keeping a dictionary in a state of repair, I believe
`somewhere in one of your papers that you filed, you said that that
`argument was proffered initially implicitly, right? Or that's not --
`that's supported with your original petition that was filed
`implicitly?
`MR. CARRAWAY: Yes, Your Honor.
`JUDGE CHUNG: Can you point us to where, please.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`
`
` 14
`
`
`
`Case IPR2016-00783
`Patent 6,597,812 B1
`
`
`MR. CARRAWAY: So if we can go to slide 15, so
`applying this -- the way that it was done was by applying the
`maintaining a dictionary to O'Brien. In Dr. Creusere's original
`declaration at paragraphs 35 to 40 and 83 to 84, he explained that
`O'Brien maintains a dictionary by storing its string table and then
`mapping reference values in that string table to characters and
`strings as they are encountered. That's how he viewed O'Brien as
`meeting the maintaining a dictionary term.
`In his reply declaration, he again provided his express
`construction and also --
`JUDGE CHUNG: Let's just focus on the initial filing or
`the initial declaration. So you are saying that implicitly we can
`draw from that that that is keeping a dictionary in a state of
`repair?
`
`MR. CARRAWAY: Yes, Your Honor. And that's why
`I was pointing to the reply.
`JUDGE CHUNG: Where does a state of repair come
`
`in?
`
`MR. CARRAWAY: The state of repair is from the
`Webster's dictionary. It's a generic term or it's a general
`definition for maintaining. But the important thing about his
`construction is he also explained in the paragraphs of his reply
`declaration how that is used in the '812 patent and what that term
`encompasses as shown in the '812 patent. And that is in
`paragraphs 6 and 7 of his reply declaration where he talks about
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`
`
`
`
` 15
`
`
`
`Case IPR2016-00783
`Patent 6,597,812 B1
`
`as used in the '812 patent, retaining or maintaining a dictionary by
`retaining it in a state of repair is shown in the '812 patent as
`storing a dictionary and adding to the dictionary as terms are
`encompassed.
`JUDGE CHUNG: Okay. So --
`JUDGE LEE: So the dictionary definition, by
`dictionary here I mean the Webster's dictionary, that definition
`that you referred to and you've highlighted in slide 11 of your
`demonstratives says to keep in an existing state. And that says
`two things to me, assuming for the moment that we apply it. First
`that it keeps something, right. Isn't it a problem for you that in
`O'Brien new dictionaries are created for each segment? So isn't
`that a problem that it's not kept? It's a new dictionary all
`together?
`And then second, I know that you chose to go with state
`of repair, but it says to keep in an existing state, at least according
`to the dictionary definition that you are applying. How does that
`apply to the -- I guess I'm having trouble applying the general
`definition to this particular context because dictionaries change,
`the contents of the dictionaries change, they are reinitialized, et
`cetera. So is that really all that applicable? It's two parts.
`MR. CARRAWAY: First part, segments. The problem
`is again there is no requirement in the claim that this thing, that
`this compression be for more than 2048 segments -- sorry,
`2048 bytes. Instead, what it is is maintaining a dictionary of a
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`
`
`
`
` 16
`
`
`
`Case IPR2016-00783
`Patent 6,597,812 B1
`
`plurality of code words. That's not 2048, which is the exemplary
`segment size of O'Brien. Basically what O'Brien does is it starts
`before the whole compression starts, which is what the patent
`claims are about. It has broken up data into segments. Each of
`those segments meets the limitations of the claim. There's
`nothing in the claim that says you have to have nonsegmented
`data, that something hasn't been processed on the data before it
`starts. The claim is focused just on the compression, and that's
`what O'Brien meets for each segment.
`The second issue is the state of repair. Again,
`Dr. Creusere in his declaration, he said, yes, we go to Webster's
`dictionary, that's the place you start. You look at that and then
`you look at the patent itself and understand how that applies to
`this particular technology. And in paragraph 7 of his declaration,
`6 and 7 of his declaration, he looks at the portions of the patent
`that use the word "maintaining," and there when they talk about
`maintaining, can we go to maybe 14 -- this is column 3, lines 31
`to 36 --
`
`JUDGE BRADEN: Are you on slide 14?
`MR. CARRAWAY: I'm on slide 11, Judge Braden. I'm
`sorry. And it's column 3, lines 31 to 36. I see my time is about
`up.
`
`JUDGE CHUNG: Finish the question.
`MR. CARRAWAY: Thank you. There it says it's one
`of the only two places where the word "maintaining" is even used
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`
`
`
`
` 17
`
`
`
`Case IPR2016-00783
`Patent 6,597,812 B1
`
`in the '812 patent other than the claims. And it says it's
`maintained and updated by generating a new code word
`corresponding to a built data block string if the data block string
`does not match a unique data block string in the dictionary and
`then adding it to the dictionary. So in his view, it doesn't mean
`that it never changes because of course that's not the purpose of a
`compression dictionary. It's added to.
`JUDGE LEE: I understand that you are over time and
`we'll grant a little extra time. Here it says maintained and
`updated, right. So that's two different actions, in theory, going on
`there. And again, I think that highlights sort of the difficulty or
`maybe not difficulty, but the uncertainty I have in applying the
`just general definition in this context and how does that work
`because I'm not sure what it means for a dictionary to be in repair.
`Do you repair a dictionary? That's nothing that either party has
`ever addressed. It's not discussed in the patent. I'm not sure what
`that means to be in repair in this context.
`So again, I'm just wondering if the general definition
`really has much usefulness here. It almost seems to me that what
`you are trying to say is maintain means it exists for the purposes
`of a segment. It's there. You are using it for a segment. So thus
`it is maintained. Is that really all that that claim limitation
`means?
`
`MR. CARRAWAY: I think that's what it means, Your
`Honor, and that's why we did say in a state of repair. The
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`
`
`
`
` 18
`
`
`
`Case IPR2016-00783
`Patent 6,597,812 B1
`
`dictionary doesn't disappear after one word is encoded. That's
`just not what happens. The dictionary is kept as you are doing
`the compression that is required by the claim. I mean, that's why
`you don't see maintaining appearing much in the specification.
`It's just a word to indicate that it continues to exist which is
`something that happens when you are describing a method and
`how something works. They may put in it's maintained during
`this process.
`JUDGE LEE: If O'Brien segmented -- let's say a
`segment in O'Brien was one byte, so after each byte you would
`have a new dictionary. Would that still be maintained in your
`understanding?
`MR. CARRAWAY: I don't think so because the claim
`requires that you maintain a dictionary of at least two code words
`and that you compress a plurality of bytes. So again, the claim
`already includes a minimum requirement, and O'Brien provides
`an exemplary segment size of 2048 bytes.
`JUDGE LEE: You are already over time. So we'll stop
`
`there.
`
`JUDGE CHUNG: We will give patent owner
`33 minutes to present their argument.
`MR. NOROOZI: Thank you, Your Honors, and may it
`please the Court, I would like to begin by responding to some of
`the points that counsel for petitioner made with respect to what is
`the invention here, what is the nature of these claims and why is it
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`
`
`
`
` 19
`
`
`
`Case IPR2016-00783
`Patent 6,597,812 B1
`
`that O'Brien does not teach the claims. Now, I also want to make
`sure that we touch on what I think are really critical points about
`the lack of necessary evidence to meet an obviousness
`combination and the requirement that obviousness begin with a
`difference between the reference and the challenged claim. I
`think those are absolutely fundamental, and our slides set those
`forth. But I want to get into some of these substantive points in
`response to petitioner's comments first.
`JUDGE LEE: Just because you brought it up,
`Mr. Noroozi, your position is that a given claim cannot be both
`anticipated and obvious in light of the same prior art reference; is
`that your position?
`MR. NOROOZI: That's not exactly our position, Judge.
`And I appreciate you asking that. Our position is the following:
`You can have theoretically an alternative theory where you say
`we think the reference anticipates, and here is why. But if, for
`example, you read the reference a certain way, then it would be
`different on this ground and therefore, we now have an
`obviousness combination. That issue of whether you could have
`both anticipation and obviousness is not at the core of our
`argument. The core of our argument is simply that the law is
`very clear and the logic is very clear that when you want to allege
`obviousness, there needs to be some difference between the
`primary prior art reference and the claims. Otherwise that's
`anticipation. And you need to define --
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`
`
`
`
` 20
`
`
`
`Case IPR2016-00783
`Patent 6,597,812 B1
`
`
`JUDGE BRADEN: Doesn't that say then anticipation is
`the epitome of obviousness? So why can that not be the case?
`MR. NOROOZI: Judge Braden, thank you for the
`question. The answer to that is that anticipation is not the
`epitome of obviousness. Anticipation is anticipation, and
`obviousness is obviousness. Congress set this forth in
`Section 102 and 103 separately because they are two separate
`things. And they have different legal requirements and they have
`different standards. So the grounds on which a petition is
`instituted define the basis on which unpatentability can be found,
`if at all.
`
`Here there is no anticipation grounds. The reason there
`is no anticipation grounds is because presumably, and we can
`only presume, petitioners didn't want to commit themselves to the
`full burden of what comes with staking your claim to anticipation.
`And whether they could have made an alternative showing or
`claim of obviousness is sort of beside the point. They didn't do it.
`And if they had done it, frankly, it would have undermined their
`anticipation claim.
`JUDGE LEE: Well, Mr. Noroozi, I understand your
`view is that anticipation is anticipation. It's not obviousness.
`And anticipation is not the epitome of obviousness. But the
`Federal Circuit seems to disagree with you. That's a direct quote
`from Federal Circuit case law. Don't we have to follow what the
`Federal Circuit tells us is the case?
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`
`
`
`
` 21
`
`
`
`Case IPR2016-00783
`Patent 6,597,812 B1
`
`
`MR. NOROOZI: Well, Judge, I'm not sure which case
`you are referring to and it's certainly not one that's been argued in
`this proceeding so far. But without the context of the case, it's
`hard for me to address it directly. What I can say is the mere
`phrase anticipation is the epitome of obviousness sounds like a
`logical statement about the fact that if something in the prior art
`directly and perfectly anticipates a claim, then certainly that claim
`was obviousness in common parlance because someone thought
`of that exact thing.
`But when we talk about anticipation and obviousness in
`the specific legal terms of Section 102 and 103, then, no, it is not
`the case that anticipation is the epitome of obviousness. They are
`two separate grounds and they come with two separate sets of
`legal requirements for a showing. So without knowing what that
`Federal Circuit case is about, I can't directly address whether it
`refutes the point I'm making right now. But the basis of my
`argument --
`JUDGE BRADEN: Just for your future reference, it's
`Jones versus Hardy from the Federal Circuit. It's 727 F.2d 1524.
`And as Judge Lee said, that is a direct quote from the Federal
`Circuit. They bas
Accessing this document will incur an additional charge of $.
After purchase, you can access this document again without charge.
Accept $ Charge
Still Working On It
This document is taking longer than usual to download. This can happen if we need to contact the court directly to obtain the document and their servers are running slowly.
Give it another minute or two to complete, and then try the refresh button.
A few More Minutes ... Still Working
It can take up to 5 minutes for us to download a document if the court servers are running slowly.
Thank you for your continued patience.
This document could not be displayed.
We could not find this document within its docket. Please go back to the docket page and check the link. If that does not work, go back to the docket and refresh it to pull the newest information.
Your account does not support viewing this document.
You need a Paid Account to view this document. Click here to change your account type.
Your account does not support viewing this document.
Set your membership
status to view this document.
With a Docket Alarm membership, you'll
get a whole lot more, including:
- Up-to-date information for this case.
- Email alerts whenever there is an update.
- Full text search for other cases.
- Get email alerts whenever a new case matches your search.
One Moment Please
The filing “” is large (MB) and is being downloaded.
Please refresh this page in a few minutes to see if the filing has been downloaded. The filing will also be emailed to you when the download completes.
Your document is on its way!
If you do not receive the document in five minutes, contact support at support@docketalarm.com.
Sealed Document
We are unable to display this document, it may be under a court ordered seal.
If you have proper credentials to access the file, you may proceed directly to the court's system using your government issued username and password.
Access Government Site
