`571-272-7822
`
`
`
`
`Paper No. 40
`Entered: December 9, 2021
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`_________________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`_________________
`
`DISH NETWORK, L.L.C.,
`DISH TECHNOLOGIES L.L.C., and
`SLING TV, L.L.C.,
`Petitioner,
`
`v.
`
`SOUND VIEW INNOVATIONS, LLC,
`Patent Owner.
`_________________
`
`IPR2020-01041
`Patent 6,725,456 B1
`_________________
`
`Record of Oral Appeal Hearing
`Held: October 19, 2021
`
`
`
`
`Before DEBRA K. STEPHENS, DANIEL J. GALLIGAN, and
`JOHN A. HUDALLA, Administrative Patent Judges.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`IPR2020-01041
`Patent 6,725,456 B1
`
`
`
`APPEARANCES:
`
`ON BEHALF OF THE PETITIONER:
`
`
`MATTHEW W. HINDMAN
`BROCK S. WEBER
`Pillsbury Winthrop Shaw Pittman LLP
`matthew.hindman@pillsburylaw.com
`
`
`
`ON BEHALF OF THE PATENT OWNER:
`
`
`NATHAN LOWENSTEIN
`PARHAM HENDIFAR
`Lowenstein & Weatherwax LLP
`lowenstein@lowensteinweatherwax.com
`hendifar@lowensteinweatherwax.com
`
`
`
`
`The above entitled matter came on for hearing on Tuesday, October
`
`19, 2021, commencing at 10:00 a.m. EST, by video/by telephone.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`2
`
`
`
`IPR2020-01041
`Patent 6,725,456 B1
`
`
`P R O C E E D I N G S
`- - - - -
` JUDGE GALLIGAN: Good morning. This is a hearing for
`IPR2020-01041. Petitioners are Dish Network, LLC; Dish
`Technologies, LLC; and Sling TV, LLC. Patent Owner is Sound View
`Innovations, LLC. The patent is 6,725,456 B1.
` I'm Administrative Patent Judge Galligan and with me are
`Administrative Patent Judges Stephens and Hudalla.
` May I have appearances of counsel first starting for
`Petitioner?
` MR. WEBER: Good morning, Your Honor. My name is Brock
`Weber, counsel for Petitioners. I'm with the law firm Pillsbury,
`Winthrop, Shaw, Pittman, and on the public audio line is lead
`counsel, Matt Hindman.
` JUDGE GALLIGAN: Thank you, Mr. Weber.
` For Patent Owner?
` MR. HENDIFAR: Good morning, Your Honor. Parham
`Hendifar of Lowenstein & Weatherwax, backup counsel for Patent
`Owner. Here with me, Mr. Nathan Lowenstein, also backup counsel
`for Patent Owner. And on video, lead counsel, Mr. Kenneth
`Weatherwax.
` JUDGE GALLIGAN: Thank you, Mr. Hendifar.
` As set forth in the oral hearing order, each party has
`one hour of argument time. Our first priority is your right to be
`heard, so if, at any time, during the hearing you encounter
`technical or other difficulties that you feel undermine your
`
`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
`26
`
`3
`
`
`
`IPR2020-01041
`Patent 6,725,456 B1
`
`ability to adequately represent your client, please let us know
`immediately, and you may need to do this by contacting the team
`who provided you with the connection information.
` And if you stop hearing and drop off, if you'd just try
`to note what was discussed so we can figure out where to pick up,
`we'll also, you know, try to see if someone drops off as well on
`the video.
` Please mute your microphone and only unmute when you're
`speaking. Please identify yourself when you speak so the
`transcript accurately reflects the speaker.
` When referring to an item in the record, please do so
`with specificity so that the panel can follow along and for
`clarity in the transcript. We have all the papers and the
`demonstratives, so if you can identify the demonstratives by slide
`number, for instance, the paper by -- the paper you're referring
`and the page of the paper.
` And there is a public audio line, and there wasn't
`any -- I don't know if anyone is listening to that. I know,
`Mr. Hendifar, you mentioned that one of the counsel's on there. I
`don't know if anyone else is on there, but just, you know, I don't
`think there's any public information or confidential information
`to be discussed. If there's any concern, I think the parties are
`supposed to raise that before.
` So, with that, Petitioner, you'll proceed first, and you
`may reserve time for rebuttal, and Patent Owner, Petitioner
`reserves time for rebuttal. You may have a sur-rebuttal.
`
`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
`26
`
`4
`
`
`
`IPR2020-01041
`Patent 6,725,456 B1
`
` Petitioner, would you like to reserve any time for
`rebuttal?
` MR. WEBER: Thank you, Your Honor. Yes, I'd like to
`reserve 20 minutes for rebuttal.
` JUDGE GALLIGAN: Okay. So it's about 10:04, and you may
`begin. And so about 10:44, you know, you'll be up at the first 40
`minutes. I'll kind of keep a general clock but you might want to
`keep one as well. Go ahead. Thank you.
` MR. WEBER: Thank you, Your Honor. Again, this is Brock
`Weber, counsel for Petitioners Dish Network, Dish Technologies,
`and Sling TV. And may it please the Board, is it best to proceed
`by sharing my desktop so you can see the demonstratives, even
`though you do have a copy yourselves?
` JUDGE GALLIGAN: No, I would just present and then refer to
`the demonstratives and we'll follow along.
` MR. WEBER: Thank you. Thank you, Your Honors. As the
`Board knows, we're here to discuss the invalidity of the '456
`patent, in particular based on the two grounds presented in our
`petition. The first ground is an obviousness ground over the
`Durand patent -- that's Exhibit 1006 -- in view of the Bennett
`paper. That's Exhibit 1008.
` And then ground two is obviousness over the Goyal paper.
`That's Exhibit 1007. And also in view of the Bennett paper,
`Exhibit 1008.
` And the Board is presented today with only one claim
`from this patent, Claim 13. As established in the petition, claim
`
`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
`26
`
`5
`
`
`
`IPR2020-01041
`Patent 6,725,456 B1
`
`13 is obvious in view of both of Petitioner's grounds one and two.
` The Patent Owner has responded to uphold claim 13 in
`view of secondary considerations, arguments about motivation to
`combine that Durand and Goyal references with Bennett and with
`views of the claimed limitations that are based on both
`misconceptions of the petition and unduly narrow claim
`constructions at this late stage.
` I'll address each of Patent Owner's arguments in turn
`and show why the petition still establishes that claim 13 is
`obvious in view of Durand over -- in further view of Bennett,
`ground 1, as well as obvious in view of Goyal in further view of
`Bennett.
` I'll begin with a short overview of the '456 patent.
`The '456 patent, as I'm referring to slide 5 here of my
`demonstrative exhibits, is directed to providing you a desired
`quality of service for an application running in computer system.
` The '456 patent correlates the phrase "desired quality
`of service" also with "quality of service guarantees," which we
`may also refer to in this proceeding as “QoS guarantees”; that's
`capital Q, lower case O, capital S, the shorthand for quality of
`service. That's a frequent shorthand also used in the prior art.
` Turning to demonstrative slide 6, the '456 patent states
`that applications running in a computer system often compete for a
`number of resources. Those resources can include CPU resources,
`memory resources, disk resources, and network bandwidth resources.
`These are all within the purview of the purported problem that 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
`26
`
`6
`
`
`
`IPR2020-01041
`Patent 6,725,456 B1
`
`'456 patent is attempting to address.
` In the background section of the '456 patent -- and now
`I'm referring to demonstrative slide 7 -- the '456 patent
`recognizes that numerous resource management techniques existed in
`the prior art. Here on slide 6, I have highlighted the '456
`patent's discussion of conventional time-sharing operating systems
`and the '456 patent uses the UNIX operating system as an example
`of such a time-sharing operating system as exemplified in the
`McKusick paper. The McKusick paper is also part of the petition
`that's Exhibit 1009. In particular, it provides details of the
`UNIX operating system that would necessarily be present.
` So, then, the '456 patent here is discussing the
`time-sharing operating systems can manage resources in a
`time-sharing manner where a certain process or job will be given
`its share of time for a CPU cycle, for instance.
` A further example is given in the '456 patent, and now
`I'm referring to slide -- demonstrative slide 8. The '456 patent
`discusses conventional proportional share schedulers. Examples of
`proportional share schedulers known in the art, according to the
`'456 patent, include those described in the Bennett paper,
`hierarchical packet fair queueing algorithms, which are similar in
`disclosure to the Bennett reference used in the invalidity grounds
`of the petition, Exhibit 1008.
` The '456 patent also discusses the Goyal paper as
`disclosing known proportional share scheduling algorithms and
`mechanisms, that like Goyal, of course, is also part of the petition's
`
`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
`26
`
`7
`
`
`
`IPR2020-01041
`Patent 6,725,456 B1
`
`invalidity grounds. That's Ground 2.
` And the '456 patent discusses those conventional
`proportional share schedulers were known to provide QoS
`guarantees. And they're called proportional share schedulers in
`essence because a group of applications or processes contending
`for resources in a computer system will be given a proportional
`share of the available resources. It can be fairly, which means
`they're all given the same share, or weighted fairly (or weighted
`fairness) in which case certain processes are given a weight and
`they are given a share of the available resources according to
`that weight.
` JUDGE GALLIGAN: Counsel, Mr. Weber, this is APJ
`Galligan. I just want to jump ahead to -- it kind of touches on
`this subject of the proportional share and your contentions --
`Petitioner's contentions for Durand. Is Petitioner's position
`that by setting out a portion of a resource –– that's a quality of
`service guarantee?
` MR. WEBER: Well, as the petition discusses, including
`that's Paper 2 on pages 36 through 44, the quality of service
`guarantee provided by Durand is more than that. It's -- it's the
`system's adjustment of the execution priority of some jobs within
`the dimension based on a given relative weight. And so the gold
`quality of service guarantee is when the jobs are processes that
`are executing within a dimension of Durand receive the allocation of
`resources according to the weight. And so it's not the weight
`itself. It's based on the weight that then addresses 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
`26
`
`8
`
`
`
`IPR2020-01041
`Patent 6,725,456 B1
`
`execution priority for those processes within the dimension.
` JUDGE GALLIGAN: I mean, so it's -- and I'm trying to get
`sort of -- there seems to be a dispute between the parties on this
`“wherein the quality of service guarantees comprise” limitation --
`recitation. And is it -- is it Petitioner's position that if
`you -- if you specify a portion of resource set aside for
`exclusive use by a process, then that's providing a quality of
`service guarantee?
` MR. WEBER: I mean, yes, that's an example of such a QoS
`guarantee in Durand. It's the -- using those parameters to adjust
`the execution priorities of the processes. And so Durand
`discloses that they will receive that share of resources that have
`been designated for them even in a heavily loaded system. And so
`Durand is disclosing that those processes running within the
`dimension will receive those resources for the exclusive use.
` JUDGE GALLIGAN: Okay. Thank you.
` MR. WEBER: Given that we have jumped a little bit
`ahead, I'll just quickly finish up on the review of the '456
`patent. I think we're all fairly familiar with it. But here on
`slide 9, I have excerpted and highlighted the way that the Board
`described in an exemplary embodiment of the '456 patent that uses
`what the '456 patent calls a uniform application programming
`interface for hierarchical proportional resource sharing. And
`this allows that the applications in subdivide resource
`reservations hierarchically and share resources of the respective
`resource -- fractions of the respective resource fairly. And so
`
`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
`26
`
`9
`
`
`
`IPR2020-01041
`Patent 6,725,456 B1
`
`this is, again, tying in with the concept of '456 patent is
`building upon and discussing proportional share schedulers.
` Here on slide 11 is the challenged claim 13 broken down
`into its constituent parts, labeled by element, as discussed in the
`petition. Claim 13 is a method claim for ensuring a particular
`quality of service for an application in a computer system.
` As we can see, it's discussing quality of service
`guarantees, and there's no mention of a specific scheduler or
`multiple schedulers or CP scheduling in particular. Again, it's a
`method for ensuring a particular quality of service for an
`application in the computer system.
` I'm discussing slide 13 now and the Patent Owner's first
`argument in its Patent Owner response appears to be that claim 13
`should be upheld due to objective indicia of non-obviousness or
`secondary considerations. And this position is without merit.
` Turning to Slide 14, the Patent Owner argues that
`Bell Labs is a renowned resource -- research institution and that
`Lucent Technologies, the assignee on the '456 patent, is a part of
`Bell Labs, and, apparently, therefore, based on the renown of Bell
`Labs, claim 13 should be upheld. This position is without merit
`for a couple of reasons. Number one, Patent Owner cited to
`petitioner's expert, Dr. Negus, in support of this, but when
`Dr. Negus testified to -- more fully, you know, in the full
`context of his deposition was that Bell Labs was, of course, you
`know, well known and is still well known but its prominence,
`especially in this space, was no more than any other research
`
`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
`26
`
`10
`
`
`
`IPR2020-01041
`Patent 6,725,456 B1
`
`department, including others that were more active in this base,
`HP, IBM, Motorola, for instance.
` And, further, there is no tying of this alleged renowned
`for Bell Labs or any of the mentors to claim 13 in particular or
`even the '456 patent in particular. And under controlling case
`law, that's insufficient to provide secondary considerations to
`uphold a challenged claim.
` In addition, Patent Owner alleges that praise for a
`USENIX paper that corresponds to the '456 patent shows secondary
`considerations of non-obviousness. And this argument also fails
`for a couple of reasons. Here on slide 15 of my demonstratives,
`I'm discussing that Patent Owner's expert, what they're relying on
`for this position, testified that he was unsure how USENIX viewed
`the conference paper -- the USENIX conference paper tied to the
`'456 patent. How it viewed the USENIX's paper contributions over
`Goyal and Bennett. So there can't be any actual basis for
`alleging that the '456 USENIX paper constituted any type of
`remarkable achievement over the prior art references in the
`petition.
` Turning now to demonstrative slide 16, the '456 patent
`also talks about the fact that the prior art discloses the
`elements 13(a) and 13(b), which are the only elements that Patent
`Owner attempted to tie to this alleged industry phrase. In
`particular, the '456 patent discusses, as exemplified here in
`slide 16, that proportional share schedulers, such as Goyal and
`Bennett, provide -- can provide an API for the resource allegation
`
`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
`26
`
`11
`
`
`
`IPR2020-01041
`Patent 6,725,456 B1
`
`mechanisms and that they can provide QoS guarantees.
` Further, turning now to slide 17, the back
`(indiscernible) citations that Patent Owner relies on to show that
`the '456 patent -- well, the USENIX '456 paper received industry
`pays, these other references merely just discuss aspects of that
`paper, not particularly praising it, and then, other instances,
`criticizing it.
` For instance, here on slide 17, the Reumann paper,
`discusses that, unlike VS and other mechanism in the prior art,
`Eclipse/BSD, which is what's discussed in the USENIX '456 paper,
`does not infer the tag or request in the absence of application
`support and does not exploit the scheduling of an application.
`And furthermore, the precursors of this work include Reference 9
`in the hierarchical scheduler in Reference 9. Reference 9 is the
`Goyal paper presented in the petition.
` And so under precedential precedent at the Board, Patent
`Owners failed to make its case under secondary considerations for
`a couple of reasons. Number one, based on the renown of Bell Labs
`or the industry praise, they have not shown that that is any way
`tied or is coextensive in scope to Claim 13. Their arguments in
`that regard are contrary to the evidence, but not only that, even
`if you accept the position that Claim 13(a) and 13(b) received
`such praise, 13(a) and 13(b) are not coextensive with the full
`scope of Claim 13.
` And, furthermore, the record reflects that these are not
`characteristics unique to Claim 13 that weren't known in the prior
`
`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
`26
`
`12
`
`
`
`IPR2020-01041
`Patent 6,725,456 B1
`
`art, including the admitted prior art.
` I'll turn now to addressing Ground 1 in particular,
`which is that Claim 13 is obvious in view of the Durand and Bennett. My
`previous discussion with secondary considerations would apply
`to both grounds the Patent Owner is attempting to, obviously,
`uphold Claim 13 in view of secondary considerations across
`both grounds.
` Turning now to specifically Ground 1, the petition shows
`that a person of ordinary skill in the art, POSITA, would have
`been motivated and able to combine Bennett with Durand.
` I'm now discussing Slide 1 of the demonstratives, and,
`here, it is reflective that the petition showed that a POSITA
`would have been motivated to combine Durand and Bennett for
`numerous reasons.
` JUDGE STEPHENS: Counselor --
` MR. WEBER: And Patent Owner fails to --
` JUDGE STEPHENS: Counselor?
` MR. WEBER: Yes, Your Honor.
` JUDGE STEPHENS: Can we skip first to showing where
`Durand shows the limitations and then come back to this? I
`particularly want you to address the arguments about where the
`request is in the quality of service guarantee and then the
`resource reservation. So if we could start with the actual
`limitations that have been argued, I'd appreciate that.
` MR. WEBER: Sure, Your Honor, yes. That makes sense.
`No problem. Let me skip ahead to that.
`
`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
`26
`
`13
`
`
`
`IPR2020-01041
`Patent 6,725,456 B1
`
` So Durand discloses and renders obvious the claim
`limitation of providing a particular quality of service to a
`request. The Petitioner shows that Durand discloses providing a
`particular quality of service to a request. And, in this
`instance, I think the Patent Owner has misconstrued a mapping of
`claim elements in what the petition shows Durand is providing as
`the quality of service guarantee. And I touched upon this a bit
`earlier with Honorable Galligan.
` Turning to the petition, for instance, at page 40 --
`this is paper 2 at page 40 -- the petition states that Durand
`discloses an LRM API in a UNIX system that adjusts execution
`priorities of jobs within dimensions based upon relative weights
`of the dimensions, thereby establishing one or more quality of
`service guarantees as recited in Claim 13.
` So that process is what is provided in the quality of
`service guarantee in Durand. It is not simply the relative
`weights on their own. And, furthermore, the request that a user
`makes through that LRM API is provided that quality of service
`guarantee. The petitioner is not saying that the request or the
`user's request is based on relative weights or that -- it's not to
`our request that the quality of service is provided.
` The quality of service guarantee is the adjustment of
`execution priorities based on the relative weights, and that is
`provided to the user's request.
` JUDGE HUDALLA: Counselor, could I ask you a question?
`This is Judge Hudalla.
`
`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
`26
`
`14
`
`
`
`IPR2020-01041
`Patent 6,725,456 B1
`
` MR. WEBER: Yes, Your Honor.
` JUDGE HUDALLA: I was wondering if you could address
`Patent Owner's argument about dimensions in Durand. I think
`there's an argument that's been made about how quality of service
`in Durand applies to all processes in a dimension. How do you
`respond to that?
` MR. WEBER: So if -- as an initial matter, I mean, it is
`accurate that Durand discloses system in which all processes
`within a dimension will receive the quality of service guarantee
`set aside for that dimension. Some dimensions have a single
`process within them. That's also shown in the figures cited in
`the petition.
` And what the point of the combination with Bennett was
`is that, since there can be multiple jobs or processes running
`within a dimension, though they don't have to be or there can be a
`single process, that Bennett provides the benefit of a
`hierarchical resource allocation mechanism for those processes
`that would be running within a dimension. And that would be
`entirely compatible with each other because, as is conceded,
`Durand is a UNIX-based system which already has concepts of
`inheritance and a hierarchical arrangement of processes such that
`the child process will inherit the properties of its parent
`process.
` And so that's the motivation -- one of the motivations
`to combine Bennett with Durand and how the processes that can be
`running within a dimension of Durand will receive the QoS
`
`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
`26
`
`15
`
`
`
`IPR2020-01041
`Patent 6,725,456 B1
`
`guarantees.
` JUDGE HUDALLA: Does your combination rely on the case
`where there is only one process running in the dimension?
` MR. WEBER: I don't think it's accurate to say that it
`relies on that embodiment, however, in that embodiment where a
`single process is running within the dimension, Durand will be
`deterministic, as explained in the petition in the section
`discussing the combination of Bennett, and that's because Durand
`discloses that a process will receive the resources set aside for
`its exclusive use, even if the system is heavily loaded.
` And so in the case in which one process is running
`within Durand, that will receive a minimum amount of resources set
`aside for its exclusive use.
` JUDGE HUDALLA: Would you agree that, in the case where
`there are, say, two or more processes running in a dimension, that
`there's no mechanism in Durand for giving different quality of
`service amounts or values to those processes?
` MR. WEBER: Durand on its own -- I believe that is true.
`In the combination with Bennett, Bennett would provide a
`guaranteed service rate that, I believe, could fluctuate as
`disclosed in Bennett and that those would be enforced in the
`system of Durand, including because, as I have mentioned, Durand
`explicitly discloses that the process will receive the resources
`set aside for its use.
` JUDGE HUDALLA: Okay. I think I understand your
`position.
`
`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
`26
`
`16
`
`
`
`IPR2020-01041
`Patent 6,725,456 B1
`
` MR. WEBER: Thank you, Your Honor.
` So turning back to --
` JUDGE STEPHENS: So, Counselor, then -- sorry, this is
`Judge Stephens.
` Counselor, so you're showing on this slide, 32, that
`this relative weight, by putting that in, that you get the quality
`of service which is shown in the current percentage?
` MR. WEBER: So, yes, this figure is -- on slide 17 is
`exemplary so their request of Durand is when a user inputs into
`the interface of 2(e) or 2(h) the information including desired
`relative weight for a dimension. That's the request.
` And then so that request, Durand provides a quality of
`service guarantee in the form of modifying the execution
`priorities of processes within the dimension based on the relative
`weight.
` JUDGE STEPHENS: So what is quality of service then?
`Does quality of service depend on the weight?
` MR. WEBER: The quality of service guarantee is the
`process disclosed in Durand of modifying the execution priorities
`of the processes based on the relative weight. So the relative
`weight helps the system determine what to and how much to allocate
`but it's the adjustment of the execution priorities, also based on
`the NICE value, that provides the quality of service guarantee.
`And it's based on the relative weight.
` And so it's not that the quality of service is based on
`a request. It's provided to a request. That's what's established
`
`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
`26
`
`17
`
`
`
`IPR2020-01041
`Patent 6,725,456 B1
`
`in the petition. And a user enters the request to interface by
`typing in or entering a relative weight. That's the request. And
`that request receives the quality of service guarantee when Durand
`readjusts the execution of a priority based on the relative
`weight.
` JUDGE HUDALLA: Okay. But what is quality of service?
`How does that -- I mean, just because you give something more
`weight doesn't necessarily mean quality of service is going to
`improve. Are they two separate things?
` MR. WEBER: So, in Durand, the quality of service,
`again, is the execution priorities that are adjusted based on the
`relative weight, and as disclosed in Durand, a system will
`readjust resources until that requested relative weight is met,
`and Durand says that will happen even if the system is heavily
`loaded.
` And so it's providing those resources to a dimension
`based on the relative weight that provide that quality of service.
`That's a quality of service request, and the request receives a
`quality of service in Durand.
` And the -- the '456 patent, as I went over earlier,
`expressly concedes that proportional share schedulers can provide
`such quality of service guarantees, and Durand, as discussed in
`the petition in the reply, is a system providing a proportional
`share scheduling or allocation scheme. Because, again, it's
`trying to provide fairness -- a weighted fairness.
` So just to recapture, I'm now on Slide 33 of 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
`26
`
`18
`
`
`
`IPR2020-01041
`Patent 6,725,456 B1
`
`demonstratives. Again, the petition shows the QoS guarantee in
`Durand is provided to a request. The request is when the user
`enters a relative weight and other information into the interfaces
`of Durand, and then the quality of service is provided to that
`request. In other words, Durand will then respond to -- directly
`to that request by adjusting execution priorities of processes
`within the dimension based on the relative weight.
` JUDGE STEPHENS: Can you address the argument that the
`Patent Owner has brought up as far as the quality of service
`guarantee's comprised resource reservations, looking at that
`Figure 2(h), I think it is? Is it 2(h), 2(e), that you had marked
`in your petition? Page 45.
` MR. WEBER: Yes. So Durand discloses quality of service
`guarantees that comprise resource reservations. So, again, this
`is kind of a recap of why I believe Patent Owner has misconstrued
`the mapping of elements. It's not just the relative weight by
`itself that constitutes the quality of service guarantee in Durand
`as we have discussed. Instead, as I'm referring to here in Slide
`56 of the demonstratives, the petition states that Durand
`discloses that adjusting the process's priority to provide a
`quality of service guarantee based on the activation indicator,
`that it should be active, and the weight associated with the
`process. And the weight is the relative weight -- relative weight
`that's entered by the user.
` And so that's the quality of service guarantee, is the
`adjusting of the process's priority based on the relative weight.
`
`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
`26
`
`19
`
`
`
`IPR2020-01041
`Patent 6,725,456 B1
`
`And that comprises, when the user enters this, the name of the --
`the name of process, a relative weight, and the activation of the
`dynamic control, as well as a list of applicable processes or
`commands for a particular dimension. That's the QoS comprises a
`resource reservation. So that additional -- those are additional
`parameters –– is what comprises the resource reservation in Durand.
` So this is reflected on, for instance, in the petition,
`which is Paper 2 on page 45, for example, in the annotation of
`Figure 2(e) where there's a graphical user interface for the user
`to enter a name of the dimension, the relative weight, an
`activation indicator, and applicable processes. Those are the
`command fields. And so that comprises the resource reservation of
`the cost guarantee.
` If it pleases the Board, I'll address Durand and
`Bennett's disclosure of utilizing an application program and
`interface to establish the cross-guarantee.
` First, as reflected here on Demonstrative 44, Durand
`discloses in API, under the Board's construction and the
`institution decision, that the claimed API is a set of routines
`used by an application program to direct the performance of
`procedures by the computer's operating system.
` Here on Slide 45 of the demonstratives is a discussion
`of the annotated Figure 1 from the petition, at pages 37 through
`40, and in particular, the mapping of the LRM API of Durand as
`the -- as an example of the application program interface. LRM
`stands for local resource management API.
`
`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
`26
`
`20
`
`
`
`IPR2020-01041
`Patent 6,725,456 B1
`
` And as discussed in the petition on pages 37 through 40,
`the application programming interface, 34, communicates with LRM
`demon or daemon process as well as the LRM shared memory.
` In addition, the LRM API 34 has an associated LRM GUI,
`graphical user interface. And the graphical user interface is
`just, you know, a typical GUI way for the user to access and
`provide parameters to the LRM API. And, specifically, the LRM and
`GUI is disclosed as enabling the interfaces through which the user
`enters relative weights and the resource reservations such as in
`Figures 2(e) and 2(g) of Durand.
` And Durand discloses that, once the user interacts with
`those interfaces and presses the confirm button, the LRM GUI will
`enable the storage of that information -- those resource
`reservations and the QoS guarantees into the JOR file database.
`And as you can see from Figure 1 here, that could only happen
`through the communication with the LRM shared memory and then
`going through the LRM API. There's no depicted direct connection
`from the LRM GUI into the JOR file database, which the Durand
`discloses as the LRM GUI communicating with.
` And so, obviously, a person of ordinary skill in the
`art, as Petitioner's expert has testified, would understand the
`text and the figures of Durand to be disclosing that the LRM GUI,
`which is associated with the LRM API