`
`571-272-7822
`
`Paper No. 49
`
`Entered: March 3, 2015
`
`RECORD OF ORAL HEARING
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`
`- - - - - -
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`- - - - - -
`
`
`
`INTERNATIONAL BUSINESS MACHINES CORPORATION,
`
`Petitioner,
`
`v.
`
`INTELLECTUAL VENTURES II LLC,
`
`Patent Owner.
`
`- - - - - - -
`
`Case IPR2014-00180
`
`Patent 7,634,666
`
`- - - - - -
`
`Oral Hearing Held: Tuesday, January 13, 2015
`
`
`
`Before MIRIAM QUINN (via video conference), DAVID McKONE (via
`
`video conference), and JAMES TARTAL, Administrative Patent Judges.
`
`
`
`The above-entitled matter came on for hearing on Tuesday, January 13,
`
`2015, at 1:31 p.m., in Hearing Room B, taken at the U.S. Patent and
`
`Trademark Office, 600 Dulany Street, Alexandria, Virginia.
`
`
`
`Case IPR2014-00180
`Patent 7,634,666
`
`APPEARANCES:
`
`ON BEHALF OF PETITIONER:
`
`
`
`
`
`
`
`
`
`
`
`KENNETH R. ADAMO, ESQ.
`EUGENE GORYUNOV, ESQ.
`JOEL R. MERKIN, ESQ.
`Kirkland & Ellis LLP
`300 North LaSalle
`Chicago, Illinois 60654
`312-862-2000
`
`
`
`
`
`
`
`
`ON BEHALF OF PATENT OWNER:
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`HERBERT D. HART, III, ESQ.
`PETER J. McANDREWS, ESQ.
`McAndrews Held & Malloy Ltd.
`500 West Madison Street
`34th Floor
`Chicago, Illinois 60661
`312-775-8000
`
`DON COULMAN, PhD, ESQ.
`Intellectual Ventures
`3150 139th Avenue SE
`Bellevue, Washington 98005
`425-677-2973
`
`2
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`Case IPR2014-00180
`Patent 7,634,666
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`P R O C E E D I N G S
`
`(1:31 p. m.)
`
`JUDGE QUINN: This is the h e aring, oral
`
`argu ment , on IPR 2014 -00180 concerning U. S. Paten t
`
`7,634,666, IBM C orporation versus Intellectual Vent ures II
`
`LLC.
`
`The Judges presid ing on this Panel are Judge David
`
`McKone, presiding over our satellit e office in Detroi t; Judge
`
`Ja mes Tartal , wh o is present in the courtroo m in Ale xandria;
`
`and myself , Judge Miria m Quinn, presiding over fro m our
`
`satellite office in Dallas.
`
`At this point I wo uld like the partie s to state your
`
`na mes for the r ec ord and appearan ce in the hearing starting
`
`with Petit ioner.
`
`MR. ADAMO: Good afternoon, Your Honor. Ken
`
`Ada mo, lead counsel for IBM . Wit h me he re toda y i s Eugene
`
`Gor yunov, who is backup counsel, and also Joel Mer kin.
`
`JUDGE QUINN: Would Patent Owner proceed to
`
`the podium and st ate an appearance for the re cord.
`
`MR. HAR T: Yes, Your Honor. Thi s is Herb Hart ,
`
`lead counsel for Patent Owner . With me toda y ar e backup
`
`counsels Peter M cAndre ws and Do n Coul man .
`
`JUDGE QUINN: Have the parties s ub mitted a cop y
`
`of the de monstrat ives to the court r eporter?
`
`
`
`3
`
`
`
`Case IPR2014-00180
`Patent 7,634,666
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`MR. ADAMO: Yes, Your Honor, both parties
`
`have.
`
`JUDGE QUINN: Oka y. Just a couple of
`
`instructions that were sub mitted in our order for the request of
`
`oral hearing: The parties will have each 40 minutes for their
`
`argu ment ti me. P etitioner will go f irst an d ma y reser ve ti me
`
`for rebuttal.
`
`Patent Owner will respond to Petit ioner's case . No
`
`rebuttal ti me ma y be allotted at this ti me for Pat ent Owne r.
`
`This hearing is o pen to the public. So , ther efore , I
`
`don't think that there was confidential infor mation to discuss,
`
`but the parties ar e alerted that this hearing is open t o the
`
`public.
`
`All objections to de monstratives ar e waived. I f
`
`there ar e an y issu es with de monstr atives that need t o be
`
`addressed during the argu ment, we ask that the parti es do so
`
`durin g their ti me that the y are allotted for their argu ment .
`
`Are there an y que stions regarding these
`
`instructions?
`
`MR. ADAMO: Your Honor, ma y I ask several
`
`housekeeping questions; would that be all right with the
`
`Court? This is Ken Ada mo for IB M.
`
`JUDGE QUINN: Yes.
`
`
`
`4
`
`
`
`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
`
`Case IPR2014-00180
`Patent 7,634,666
`
`
`MR. ADAMO: No questions as to the Court's
`
`instruction about an y ob jections be ing stated during one's
`
`ti me .
`
`The question wou ld beco me, if duri ng opposing
`
`counsel's argu me nt there is an objection, would you prefe r that
`
`we state it during their ti me or hol d it for rebuttal? Various
`
`panels of the Boa rd have done it different wa ys. Wh at is the
`
`Court's prefe renc e toda y, if I might ask you?
`
`JUDGE QUINN: There will be no speaking
`
`objections that interrupt the other side's argu ment. If you
`
`have an objection to make or so met hing you want to note
`
`regarding the other side's de monst ration and argu ment, you
`
`need to do so dur ing your own arg u ment.
`
`MR. ADAMO: Thank you, Your Honor.
`
`A further housekeeping matte r: Th ere are several
`
`motions to exclude pending. As I u nderstand the Boa rd's usual
`
`practice, you will rule on those mo tions as part of your final
`
`decision.
`
`Is it ac ceptable to the Court that we don't, even
`
`during our own ti me , repeat an y ob jections to the exhibits that
`
`are in issue in those motions to exc lude, and the und erstanding
`
`would be there is no waiving of an y objections si mply be cause
`
`it is not repeated during the trial h ere toda y?
`
`JUDGE QUINN: That's correct . Whatever
`
`objections you ha ve made in your motion to exclude are
`
`
`
`5
`
`
`
`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
`
`Case IPR2014-00180
`Patent 7,634,666
`
`preserv ed in those papers. No nee d to raise those d uring the
`
`argu ment .
`
`MR. ADAMO: One last point and then I will sit
`
`down: I had disc ussed with opposing counsel, as a wa y to
`
`mini mize ti me us age, that it would -- we suggest to the Board
`
`that, both of the decks of Po werPo ints, if I could use that,
`
`have got an exhibit nu mber on it, a marked for ident ification
`
`nu mber . It is 1040 is the nu mb er f or the I BM Power Points and
`
`it is 2018 for I V's deck.
`
`To save ti me, if we state that all o f the opening
`
`presentation is coming out of IBM ' s 1040, so that f r o m there
`
`on out we just ha ve to recite the sl ide nu mbers and, of course,
`
`IV would do the s a me during their opening presentation, would
`
`that be acc eptable to you to preserv e the r ecord in th e for m the
`
`Board would like it?
`
`JUDGE QUINN: Yes, as long as it is clear that,
`
`when you sa y I B M's 1040, you ar e actuall y re ferrin g to a slide.
`
`That's fine to r efe r to the m with sli de nu mbe rs. You don't
`
`have to repeat tha t it is an exhibit to the record .
`
`MR. ADAMO: What we were ho ping to do is not
`
`have to repeat the 1040 so that, unl ess we said other wise, it is
`
`all out of 1040 an d theirs would all be out of 2018 on the
`
`openings.
`
`JUDGE QUINN: That's correct . It is understood.
`
`MR. ADAMO: Oka y.
`
`
`
`6
`
`
`
`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
`
`Case IPR2014-00180
`Patent 7,634,666
`
`
`JUDGE QUINN: Thank you . Can I ask Pat ent
`
`Owne r if the y hav e an y questions?
`
`MR. HAR T: You r Honor, we have no questions
`
`about the procedure.
`
`JUDGE QUINN: Oka y.
`
`MR. ADAMO: The last point and then I will sit
`
`down: It is IB M' s intention, Your Honor, that Mr . Gor yunov
`
`will do 98 perc ent o f the argu ment today during both the
`
`opening argu ment and during the r e buttal.
`
`With the Boa rd's per mission, it wa s my intention,
`
`if an y objections or other evidentia r y issues needed t o be
`
`stated, that I would do that. But ot her than that M r. Gor yunov
`
`will do ever ything.
`
`Is that oka y with the Board?
`
`JUDGE QUINN: Yes. You c an ha ve your backup
`
`counsel do so me argu ments and yo u can divide that as you
`
`wish.
`
`MR. ADAMO: All right. Then , Your Honor, if I
`
`might, I a m going to sit down and l et Mr . Gor yuno v start.
`
`JUDGE QUINN: Yes. The counsel for Petitioner
`
`who is going to a rgue can step up t o the podiu m and address
`
`the Board with how much ti me you want for your a rgu ment and
`
`how much for r ebuttal.
`
`MR. GOR YUNOV: Your Honor, good afternoon.
`
`Eugene Gor yunov on behalf of Peti tioner IBM . We would like
`
`
`
`7
`
`
`
`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
`
`Case IPR2014-00180
`Patent 7,634,666
`
`with your per miss ion to reserve 10 minutes for rebut tal and 10
`
`minutes for our d irect presentation toda y to you .
`
`Is that all right? I' m sor r y, 30 minutes for the
`
`direct presentation and 10 minutes for th e rebuttal.
`
`JUDGE QUINN: Yes. Please proc eed.
`
`MR. GOR YUNOV: Ma y it please the Board. We
`
`are he re toda y re garding the invalidity of the '666 p atent, the
`
`argu ments pr esented in various mo tions to exclude, and the
`
`proper construction of certain clai m ter ms .
`
`IBM's petition wa s filed in Nove mb er 2013 and
`
`was supported b y the declaration of Dr. Cetin Ka ya Koc, who
`
`is joining us here in the boardroo m. Dr. Koc is the l eading
`
`authority on hard ware i mple mentat ions of cr yptogra phy. He
`
`co-founded three of the most i mpor tant conferences on
`
`cr yptographic har dware and literall y wrote the book on
`
`cr yptographic alg orith ms and recon figurable hardwa r e.
`
`The Board r eviewed IBM 's petition and all of its
`
`evidence, including Dr . Koc's decl aration, in instituting its
`
`IPR t rial. The Board issued its tria l on all clai ms of the '666
`
`patent, whe re the independent claims of the '666 pat ent we re
`
`instituted on the co mbination of th e Matsuzaki and the
`
`Dworkin prior art ref erences .
`
`Matsuzaki is IBM 's pri mar y p rior a rt ref erence an d
`
`discloses all but two ele ments b y itself of all of the
`
`independent claims . The co mbina tion of Matsuzaki and
`
`
`
`8
`
`
`
`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
`
`Case IPR2014-00180
`Patent 7,634,666
`
`Dworkin, which i s IBM's secondar y prior art re fere nce,
`
`discloses all ele ments of the indepe ndent clai ms arr a nged as in
`
`the clai ms.
`
`Now, IV cont ends that this co mbination does not
`
`disclose just thre e ele ments of the various nu merous
`
`li mitations recite d in the independent clai ms. Those
`
`li mitations are fe eding back multip le outputs in claims 1 and
`
`4; a concatenate r and splitter for merging or split ting data
`
`widths of clai m 4 ; the cr yptographic controller gener ating
`
`status interrupt signals for the host; and also generat ing op
`
`code signals wher e the a rith metic u nit selects betwe e n RS A
`
`and EC C modes o f operation based on those op codes.
`
`IV does not dispute an y of the othe r ele ments of
`
`the independent clai ms . And as I B M has de monstrat ed in its
`
`petition as well a s its repl y papers , all of these ele ments are
`
`disclosed by its c o mbination of Ma tsuzaki and Dwor kin. The
`
`Board should find that the '666 pat e nt is invalid.
`
`Now, the first ele ment that I V challenges is the
`
`feedback of multiple outputs. Mat suzaki discloses three
`
`co mputational units of the '666 pate nt.
`
`As you can see , we're ref erring to s lide nu mber 11,
`
`which is depicting figure 17 of the Ma tsuzaki prior art, Exhibit
`
`1008. Each of the multiplier, the s ign inverting unit, and the
`
`three-input adder, have outputs. It is the line co ming out of
`
`the botto m of eac h one of those un its.
`
`
`
`9
`
`
`
`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
`
`Case IPR2014-00180
`Patent 7,634,666
`
`
`Matsuzaki also discloses that the a rith metic unit
`
`outputs a carr y-u p signal. You c a n see that in slide 13. At the
`
`mo ment I a m refe rring to figure nu mber 11 which is on the
`
`right -hand side of Matsuzaki Exhibit 1008.
`
`A c arr y-up signal is a signal indicating that a
`
`re mainder has be en generated duri ng the modular
`
`multiplication operation, and that carr y -up signal is sent back
`
`to the controller.
`
`Now, Dworkin, I BM's secondar y p rior art
`
`referenc e, discloses the co mmon ar chitecture with a nu mber of
`
`co mputational units, with each co mputational unit having an
`
`output.
`
`Now, we have out lined these for th e Board's
`
`convenience in re d and the co mput ational units we h ave
`
`outlined on the bottom of what yo u see in front of you as slide
`
`14, and it is a dep iction of Dworkin's figure 2. And Dworkin
`
`is Exhibit 1012.
`
`Now, Dworkin al so discloses that each one of these
`
`outputs is fed back to the controller. And I V is wro ng when it
`
`argues that Dwor kin discloses feeding back up onl y one
`
`output. IV's a rgument is based, Yo ur Honors, on a
`
`funda mental misu nderstanding of the Dworkin re fe r ence.
`
`IV atte mpts to ta ke the si mple issue of feeding
`
`back multiple out puts, as you can see in our slide 14, and it
`
`makes an incorr e ct bit level a rgu ment based on incorrect
`
`
`
`10
`
`
`
`Case IPR2014-00180
`Patent 7,634,666
`
`reading of Dwork in's figure 6 rathe r than addressing the
`
`straightforward h ardwar e co mmuni cation lines that a re
`
`de monstrated wit hin Dworkin.
`
`IV's confusion about the disclosure of Dworkin
`
`only beca me appa rent afte r I V filed its Patent Owner response
`
`and its supporting expert's decla ra tion. IB M's expert and I BM
`
`took IV's argu me nt and explained wh y it was wrong . IBM
`
`even included a s eries of step -b y-s tep graphics to he lp explain
`
`wh y I V's interpr e tation of Dworkin was wrong.
`
`JUDGE Mc KONE: Mr . Gor yunov, I understand
`
`you ar e a rguing that you added this new argu ment i n to show
`
`that Patent Owne r was wrong.
`
`Wh y wasn't it inc u mbent on you in the petition to
`
`present this information to show th at Petitioner was correct?
`
`MR. GOR YUNOV: Your Honor, f irst of all, we
`
`disagree that this is new argu ment at all. In fa ct, during his
`
`deposition , Dr . Koc testified that a t least paragr aph 118 of his
`
`opening declaration, that's Exhibit 1001, addressed the
`
`feedback of multiple outputs.
`
`In his view this was a t rivial aspect of the Dworkin
`
`prior art refe renc e. It is in his vie w and based on th e re c ord,
`
`one of ordinar y s kill in the art wo uld have understood that this
`
`is just a trivial po int of the Dworki n refe rence . The r e was no
`
`need to explain 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
`
`
`
`11
`
`
`
`Case IPR2014-00180
`Patent 7,634,666
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`We didn't underst and that there wa s this level of
`
`confusion fro m I V's expert as well as I V until IV ha d filed its
`
`Patent Owner res ponse. We didn't have an y basis to believe
`
`that so mebod y co uld misunderstood so mething so
`
`straightforward.
`
`JUDGE Mc KONE: So you ad mit th at this wasn't
`
`presented in your petition; it is jus t your argu ment t hat you
`
`didn't thi nk you h ad to?
`
`MR. GOR YUNOV: Your Honor, we did pr esent
`
`the multiple feed back argu ment in our petition and Dr . Koc's
`
`declaration at le a st at paragr aph 118. We provided a
`
`step -by-step explanation of wh y I V's argu ment was incorrect
`
`in IBM's repl y as well as Dr. Koc's responsive repl y
`
`declaration.
`
`We believed that we onl y needed t o do that afte r
`
`IV de monstrated an entirel y co mpl ete misunderstanding of the
`
`Dworkin prior art ref erence . Like I said, Dr. Koc be lieved one
`
`of ordinar y skill in the art would h ave no issues looking at
`
`Dworkin's disclosure and being abl e to identif y that there ar e
`
`multiple outputs fro m multiple co mputational units, as
`
`depicted on slide 14, and that e ach one of these outputs is fed
`
`back to the contr oller.
`
`Now, in an y case , I V's a rgu ment is wrong because
`
`if you take a look at Dworkin's pse udocode and figure 2 of
`
`
`
`12
`
`
`
`Case IPR2014-00180
`Patent 7,634,666
`
`Dworkin, which i ncludes a cross -s ection of figure 6, it
`
`discloses feeding back multiple out puts.
`
`You can see that on IBM's slide 15 at the ver y top.
`
`We have identifie d t hat for your co nvenience as grap hic 2.
`
`And this is an extract f ro m Dworkin at colu mn 4, lin es 21
`
`through 28, Exhibit 1012.
`
`The pseudocode o f Dworkin is responsible for
`
`finite field multiplication, and it do es so b y re ceiving as input
`
`the multiplier, th e multiplicand and modulus. It the n uses
`
`those values to ca lculate the - - to p erfor m finite fiel d
`
`multiplication.
`
`Now, to make it e asier for us to dis cuss all of these
`
`ele ments, I will r efer to ti me zero through time 1 , ti me 2 and
`
`so on, just so we can s ee what happens at each cloc k c ycle
`
`within Dworkin.
`
`Now, as an initial step, be fore an y co mputations
`
`are pe rfor med , th e values of A, B a nd M, which a re the
`
`multiplier, multiplicand and the mo dulus, are loaded into the
`
`special purpose r egisters, and spec i fic registers with in special
`
`purpose register 16. Those register s are A, B and M. You can
`
`see that in gr aphic 3 that we prese nt on slide 15.
`
`Now, first, at ti me ze ro, sub - ALU' s 18 perfor med
`
`the ver y first step of modular multiplication. And, you know,
`
`I apologize. I mi sspoke. It was not finite field. Modular
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`
`
`13
`
`
`
`Case IPR2014-00180
`Patent 7,634,666
`
`multiplication. And output the partial product C to r egister C,
`
`also contained wi thin the special purpose register 16.
`
`You can see that in the upper left -hand corner of
`
`IBM slide 16, which is an annotate d version of Dwo rkin's
`
`figure 2, Exhibit 1012. And that's the yello w highlighting that
`
`you see co ming o ut of the yellow s ub -ALU and going into the
`
`register C of the special purpose r egister 16.
`
`Second, at ti me 1, a fter this first p artial pro duct
`
`has been calculat ed, the entire firs t partial product i s shifted
`
`left. We depict this on the lower l eft -hand side of slide 16.
`
`JUDGE Mc KONE: And whe re is that disclosed in
`
`Dworkin?
`
`MR. GOR YUNOV: The shift left , Your Honor, is
`
`that your question ?
`
`JUDGE Mc KONE: The shift le ft.
`
`MR. GOR YUNOV: Yes, Your Ho nor. Dworkin
`
`discloses the shift left operation within its pseudocode. If you
`
`take a look again at slide 15, I beli eve.
`
`As you can tell, e ver y subsequent iteration of the
`
`pseudocode uses C j – 1. That is the value of the fed b ack
`
`ele ment that was fed back fro m spe cial purpose regis ter into
`
`the controller.
`
`JUDGE Mc KONE: And whe re doe s it show that
`
`C j – 1 is f ed back to the controller?
`
`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 IPR2014-00180
`Patent 7,634,666
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`MR. GOR YUNOV: Well , as an initial ma tter, if
`
`you take a look at graphic 3 on slid e 15, you can see the
`
`arrows co ming out of the control bit going into the controller.
`
`Also, IV's expert , Dr. Schau mont, does not dispute that C M bit
`
`is considered fee dback to the controller. So that's s o methin g
`
`that, at least in o ur understanding, there is no dispute on the
`
`record.
`
`Moreover, during his deposition Dr. Koc explained
`
`that the C register is a shift r egister and Dr . Koc the n pointed
`
`to, as an exa mple , Dwo rkin's clai m 9 which talks ab out C
`
`register being a shift register.
`
`Now, at this point we have fed bac k -- i f we can go
`
`back to slide 16 - - the most signifi cant bit of the par tial
`
`product is fed bac k into the controller.
`
`Now, one of ordi nar y skill in the a rt would
`
`understand this happens, again, f or two reasons. One, as I
`
`mentioned, the va lue of the first pa rtial product that is fed
`
`back to the contr oller is used as input into the second iteration
`
`of the pseudocode. That's the C j – 1 .
`
`Secondl y, the mo st significant bit is a control bit
`
`that triggers the s ub -ALU's to perf or m the next ite ra tion of the
`
`multiplication process.
`
`Next, at ti me 2 su b -ALU 18 perfor ms the next
`
`phase of the pseu docode and gener ates the second p artial
`
`
`
`15
`
`
`
`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
`
`Case IPR2014-00180
`Patent 7,634,666
`
`product and, agai n, loads that into special purpose r egister 16 ,
`
`and specificall y r egister C within t hat special purpose register.
`
`Again, at ti me 3 t he entire se cond partial product
`
`is again shifted le ft. This is a one - bit shift where the most
`
`significant bit of the second partial product is shifte d into the
`
`contro ller. It is f ed back into the c ontroller as a res ult of the
`
`shift left operatio n. And you c an s ee that proc ess de picted on
`
`IBM's slide 16, g raphic 6 and gr aphic 7. That's the purple
`
`highlighting out of the purple sub - ALU that we hav e colored
`
`in for the Board's convenience.
`
`Now, this process is repe ated until all of the values
`
`of the partial product are eventualit y shifted back into the
`
`controller fro m re gister C in the special purpose regi ster 16.
`
`Secondl y, Dworki n's feedback is the sa me as that
`
`disclosed in the ' 666 patent. Dwor kin teaches fe eding back the
`
`output of one sub -ALU at one cloc k c ycle until all t he outputs
`
`of the sub -ALU's are fed back to the controller. Thi s is
`
`exactl y what the ' 666 patent discloses.
`
`In fact , IV's expe rt, Dr . Schau mo nt, ad mitted that
`
`the output of the multiplication unit in the '666 pate nt is
`
`multiplexed with the output of the addition unit.
`
`Now, as we can s ee on I BM's slide 18, which
`
`de monstrates gr a phic 10, it is a co mparison of figur e 2 of the
`
`'666 patent, whic h is Exhibit 1005, and the Dworkin referenc e,
`
`figure 2, which a gain is Exhibit 1012.
`
`
`
`16
`
`
`
`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
`
`Case IPR2014-00180
`Patent 7,634,666
`
`
`And, again, as Dr . S chau mont testified during his
`
`deposition, the outputs in the '666 patent have to be
`
`multiplexed beca use, if the y were output at the sa me ti me ,
`
`the y wou ld collide. This is exactl y what Dworkin de scribes.
`
`If you co mpar e th e two figures that we have
`
`presented to you on slide 18, one h aving ordinar y sk ill in the
`
`art would see that there is one co mmunication line f ro m the
`
`co mmunicational -- f ro m the co mp ut ational units to the
`
`controller. This me ans that Dwork in outputs only o ne output
`
`at a ti me per cloc k c ycle.
`
`Now, the second ele ment that I V disputes is the --
`
`JUDGE Mc KONE: Be fore you mo ve on to that,
`
`precisel y how are you proposing to co mbine Dworki n and
`
`Matsuzaki?
`
`MR. GOR YUNOV: So, Your Honor, one having
`
`ordinar y skill in t he art reviewing Matsuzaki and Dworkin
`
`would see the ref erences are co mple mentar y. One h aving
`
`ordinar y skill would, therefore , be motivated to co mbine the
`
`referenc es.
`
`Now, IBM --
`
`JUDGE Mc KONE: How? What do you t ake f ro m
`
`Dworkin and put into Matsuzaki?
`
`MR. GOR YUNOV: Understood, Your Honor. We
`
`would take - - one with ordinar y ski ll in the art woul d take the
`
`multiple fe edbacks of the outputs of the co mputational units of
`
`
`
`17
`
`
`
`Case IPR2014-00180
`Patent 7,634,666
`
`Dwork in and add those to the retur ning back of the c arr y -up
`
`signal of the Mat suzaki refer ence.
`
`Also, as I will --
`
`JUDGE Mc KONE: It is si mpl y, it is just si mpl y
`
`adding the m back ?
`
`MR. GOR YUNOV: I' m sorr y?
`
`JUDGE Mc KONE: It 's si mpl y adding the
`
`feedback?
`
`MR. GOR YUNOV: Yes, Your Ho nor. That is the
`
`only ele ment that Matsuzaki does n ot expressl y discl ose.
`
`Matsuzaki discloses sending back a car r y -up signal. Dworkin
`
`discloses sending back multiple ou tputs. That is the onl y
`
`ele ment that one of ordinar y skill woul d look to Dworkin to
`
`add in order to invalidate clai m 1 .
`
`Now, as I will me ntion later on -- o r I can address
`
`that right now if you pref er -- one of skill in the art would also
`
`look to Dworkin t o afford the op -c ode signal selecting between
`
`RSA and ECC co m putations for th e last li mitation of clai m 4.
`
`JUDGE QUINN: Let's sta y with this issue right
`
`now. It see ms to be ver y pressing . And, that is, you contend
`
`that the multiple outputs is what Dworkin contributes to
`
`Matsuzaki.
`
`And I think the ques tion fro m my colleague is wh y
`
`would you put mu ltiple outputs into Matsuzaki? Wha t reason
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`
`
`18
`
`
`
`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
`
`Case IPR2014-00180
`Patent 7,634,666
`
`could a person of ordinar y skill hav e to make that mo dification
`
`of multiple outputs?
`
`MR. GOR YUNOV: So as we desc ribed in our
`
`petition, one having ordinar y skill wo uld look to the multiple
`
`outputs in order to create a multi -f unctional co mputational
`
`unit that has mult iple, that has -- a nd let me make sure I get
`
`my wording right -- that has granular control on how
`
`co mputations are perfor med.
`
`The benefit of -- if we ma y go to s lide 26, Your
`
`Honor, please , in IBM's de ck. Bot h Matsuzaki and Dworkin
`
`disclose -- the y a ddress the sa me proble m, opti mal h ardwar e
`
`i mple mentations of cr yptographic co -processors. All right.
`
`So the r eason wh y one of ordinar y skill would take
`
`the multiple outputs disclosed in Dworkin and add th e m to
`
`Matsuzaki is, aga in, to be able to h ave this opti mal hardwar e
`
`with multiple co mponents.
`
`JUDGE Mc KONE: Is that in your petition
`
`an ywhe re?
`
`MR. GOR YUNOV: Your Honor, t hat is addressed
`
`in our petiti on at page 31 and also in Dr . Koc's opening
`
`declaration on pa ragraph 145. And we have addressed that
`
`this is exactl y wh at one of ordinar y skill in the art would do.
`
`Now, IBM also --
`
`JUDGE Mc KONE: Well, you've sa id that
`
`Matsuzaki and Dworkin are si mila r . The y have si milar
`
`
`
`19
`
`
`
`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
`
`Case IPR2014-00180
`Patent 7,634,666
`
`technology for so lving simila r pro ble ms. But I don' t see
`
`an ywhe re in here an y desc ription of how Montgo mer y's
`
`circuitr y would b e changed or i mp r oved or an ything based on
`
`incorporation of certain featur es in Dworkin.
`
`MR. GOR YUNOV: Yo ur Honor - -
`
`JUDGE Mc KONE: Am I wrong?
`
`MR. GOR YUNOV: I' m sorr y, Yo ur Honor, I
`
`started speaking before I heard wh at you just said .
`
`JUDGE Mc KONE: That's all right. Pl ease go
`
`ahead.
`
`MR. GOR YUNOV: So in our peti tion we
`
`identified the mul tiple si milarit ies between M atsuzaki and
`
`Dworkin. And, a ccording to KS R, Your Honor, a co urt can
`
`look to interrelated teachings of multiple references , the
`
`effects of the de mands known to t he design co mmu nity or
`
`present in the ma rketplace, and the background knowledge
`
`possessed by pers ons having ordinar y skill in the art , all to
`
`deter mine whether there is a re ason to co mbine the r eferences .
`
`And that is exa ctly what IBM did in its petition,
`
`and as well as Dr. Koc 's decla ration. We have ident ified
`
`nu merous r easons wh y o ne having ordinar y skill in t he art
`
`would take the multiple interrelated teachings of the two
`
`referenc es, as wel l as the knowledg e of a person havi ng skill
`
`in the art , and wo uld co mbine the t wo re ferenc es to arrive at
`
`the invention of the '666 patent.
`
`
`
`20
`
`
`
`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
`
`Case IPR2014-00180
`Patent 7,634,666
`
`
`JUD GE Mc KONE: Beside that si milarit y, what
`
`other reasons do you state in the p etition?
`
`MR. GOR YUNOV: I' m sorr y, Yo ur Honor?
`
`JUDGE Mc KONE: Besides si mila r ities between
`
`the two refe rence s, what other reas ons do you state i n your
`
`petition?
`
`MR. GOR YUNOV: We state that one having
`
`ordinar y skill in t he art would have co mbined the m i n order to
`
`create a proc essor that perfor ms both EC C and RS A e fficientl y
`
`and quickly.
`
`JUDGE Mc KONE: Wh ere is that i n your petition?
`
`MR. GOR YUNOV: One mo ment, Your Honor. So
`
`we mentioned that in, fo r exa mple , paragraph 148 of Dr. Koc 's
`
`declaration, Exhibit 1001. And then, again , we discussed the
`
`whole point of - - I' m sor r y, we aga in discussed this in Dr .
`
`Koc's de claration, 1001 at pa ragra ph 145. And then again - -
`
`JUDGE Mc KONE : I ' m sor r y, wher e in paragr aph
`
`148 do you make this point, or does Dr . Koc make t his point?
`
`JUDGE QUINN: And particularl y here what we
`
`are looking for is the state ment that you said to do t hings
`
`efficientl y, that language. We don't see it in either y our
`
`petition or your d eclaration so we want to mak e sur e we don't
`
`miss an ything her e.
`
`MR. GOR YUNOV: Understood, Your Honor. One
`
`mo ment , please .
`
`
`
`21
`
`
`
`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
`
`Case IPR2014-00180
`Patent 7,634,666
`
`
`Oka y. So , first o f all, if you take a look at
`
`paragraph 145 of Dr . Koc's opening declaration, we discuss
`
`both Matsuzaki and Dworkin addre ssing the sa me pr oble m.
`
`And this is what we've also depict ed for you on slid e 26 of
`
`IBM's de monstrat ives, and that sa me proble m is opti mal
`
`hardware i mple mentations of cr ypt ographic co -proce ssors with
`
`multiple co mpone nts.
`
`Now, the other r e ason wh y one of ordinar y skill in
`
`the art would have co mbined Mats uzaki and Dworki n is
`
`because the y both teach ha rdwa re i mple mentations o f
`
`Montgome r y r eduction. And I re alize that I V has di sputed that
`
`disclosure. Howe ver, as we've des crib ed in our brie fing, that
`
`is absolutely not the case.
`
`JUDGE Mc KONE: Well, let's assume that's
`
`correct. That 's ju st another argu me nt that Matsuzaki and
`
`Dworkin are si mi lar.
`
`Do you have an y other reason to co mbine, other
`
`than that the y are si mila r?
`
`MR. GOR YUNOV: Well , if you t ake a look at the
`
`end of paragr aph 146, we state that the co mmon disc losure of
`
`Montgome r y r eduction in Matsuzaki and Dworkin would have
`
`led one of ordinar y skill to conside r their te achings together.
`
`JUDGE Mc KONE: It doesn't sa y wh y.
`
`JUDGE QUINN: That is the conclusion. I think
`
`what we are tr ying to get at here is wher e do you pr ovide the
`
`
`
`22
`
`
`
`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
`
`Case IPR2014-00180
`Patent 7,634,666
`
`facts to support t he conclusion of obviousness. We need to
`
`know these ma rke t de mands or co mmon sense applications or
`
`the reasons that o ne of ordinar y ski ll in the art woul d have
`
`made that conclus ion.
`
`And all we have here, so fa r you h ave shown us, is
`
`that the ref erence s are si mila r, or that the y solve similar
`
`proble ms, but you need to go one more step. And we 're t r