`
`
`
`
`
`
`
`
`
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`
`____________________
`
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`____________________
`
`APPLE INC.,
`Petitioner,
`
`v.
`
`LIMESTONE MEMORY SYSTEMS LLC,
`Patent Owner.
`____________________
`
`Case IPR2016-01561
`U.S. Patent No. 6,233,181
`____________________
`
`PETITIONER’S REPLY TO PATENT OWNER’S RESPONSE
`
`Mail Stop PATENT BOARD
`Patent Trial and Appeal Board
`U.S. Patent and Trademark Office
`P.O. Box 1450
`Alexandria, VA 22313-1450
`Submitted Electronically via the Patent Review Processing System
`
`DC01:1156335.6
`
`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`
`
`
` IPR2016-01561
`
`I.
`II.
`
`B.
`
`C.
`
`TABLE OF CONTENTS
`INTRODUCTION .............................................................................................. 1
`ARGUMENT ...................................................................................................... 7
`Sukegawa Does Not Teach Away from the Any-to-Any
`A.
`Redundant DRAM Claimed in Claims 1 and 2 ..................................... 7
`The Motivation to Use Fujishima’s Sense Amplifiers Is
`Found in Fujishima Itself .................................................................... 11
`Claim 3 Does Not Mention Control Circuitry, and
`Fujishima’s Control Circuitry Is Irrelevant to the
`Analysis of Claim 3 ............................................................................. 14
`D. Dr. Mazumder Properly Considered Obviousness
`Through the Lens of a POSITA at the Time of Invention,
`and the Dispute Between the Parties as to the Level of
`Skill Is Irrelevant Here ........................................................................ 15
`There is No General “Teaching Away” from the
`Combination of Sukegawa and Fujishima .......................................... 16
`Sukegawa and Fujishima Disclose All Elements of Claim
`5 Except for Simultaneously Driving Memory in
`Separate Arrays, Which Is Taught by Walck ...................................... 20
`G. Dr. Mazumder Clearly Describes the Motivations for
`Combining Sukegawa, Fujishima, and Walck to Arrive at
`the Memory Device in Claim 5 ........................................................... 22
`III. CONCLUSION ................................................................................................. 24
`
`
`
`E.
`
`F.
`
`
`
`DC01:1156335.6
`
`i
`
`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`
`
`
` IPR2016-01561
`
`UPDATED TABLE OF EXHIBITS
`
`Exhibit #
`
`Exhibit Description
`
`
`1001* Declaration of Dr. Pinaki Mazumder
`
`1002*
`
`Curriculum Vitae of Dr. Pinaki Mazumder
`
`1003* U.S. Patent No. 6,233,181
`
`1004*
`
`File History for U.S. Patent No. 6,233,181
`
`1005*
`
`U.S. Patent No. 5,487,040 to Sukegawa
`
`1006* U.S. Patent No. 5,267,214 to Fujishima
`
`1007*
`
`U.S. Patent No. 4,967,397 to Walck
`
`1008*
`
`U.S. Patent No. 5,956,285 to Watanabe
`
`1009*
`
`1010*
`
`Masashi Horiguchi et al., A Flexible Redundancy Technique for High-
`Density DRAM’s, IEEE JOURNAL OF SOLID-STATE CIRCUITS, VOL.
`26, NO. 1, Jan. 1991, at 12-17
`
`Kazutami Arimoto et al., A 60-ns 3.3-V-Only 16 Mbit DRAM with
`Multipurpose Register, IEEE JOURNAL OF SOLID-STATE CIRCUITS,
`VOL. 24, NO. 5, Oct. 1989, at 1184-90
`
`1011*
`
`U.S. Patent No. 5,687,123 to Hidaka
`
`1012*
`
`U.S. Patent No. 5,726,946 to Yamagata
`
`1013*
`
`U.S. Patent No. 6,003,148 to Yamauchi
`
`1014*
`
`U.S. Patent No. 6,075,743 to Barth
`
`1015*
`
`1016*
`
`Inter Partes Review No. IPR2016-00096, Decision Granting
`Institution filed April 21, 2016
`
`Inter Partes Review No. IPR2016-00096, Judgment Granting
`Request for Adverse Judgment filed August 3, 2016
`
`DC01:1156335.6
`
`ii
`
`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`
`
`
` IPR2016-01561
`
`1017* Affidavit of Michael N. Zachary in Support of Motion for Pro Hac
`Vice Admission
`1018* Affidavit of Rose Cordero Prey in Support of Motion for Pro Hac
`Vice Admission
`
`1019
`
`Transcript of July 28, 2017 Deposition of Dr. Sunil Khatri
`
`1020
`
`European Patent Application Publication No. 0 499 131 A1 to
`Sukegawa
`
`* Previously filed.
`
`
`
`DC01:1156335.6
`
`iii
`
`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`I.
`
`INTRODUCTION
`
`
`
`IPR2016-01561
`
`With respect to claim 3, the Response (Paper 13, “Resp.”) of Patent Owner
`
`Limestone Memory Systems LLC (“Limestone”) attacks only the motivation to
`
`combine the amplifiers of U.S. Patent No. 5,267,214 to Fujishima et al.
`
`(“Fujishima”) with the dynamic random access memory (“DRAM”) of U.S. Patent
`
`No. 5,487,040 to Sukegawa et al. (“Sukegawa”). Limestone does not dispute that:
`
`• Sukegawa discloses each of the limitations found in claims 1 and 2 of
`U.S. Patent No. 6,233,181 (“the ‘181 patent”).
`• Fujishima discloses each of the additional sense amplifier limitations
`found in claim 3 of the ‘181 patent.
`• The alternate shared sense amplifiers claimed in claim 3 in fact have the
`known benefits set forth in Fujishima.
`
`By admitting that the alternate shared sense amplifier arrangement of
`
`Fujishima possesses the benefits described therein (Paper 13, at 27-28), Limestone
`
`has effectively conceded that claim 3 of the ‘181 patent represents the
`
`incorporation of a known sense amplifier scheme having known benefits into the
`
`any-to-any redundant DRAM of Sukegawa to achieve the expected result. This is
`
`the epitome of an obvious combination. KSR Intern. Co. v. Teleflex Inc., 550 U.S.
`
`398, 401 (2007) (“If a person of ordinary skill in the art can implement a
`
`predictable variation, and would see the benefit of doing so, §103 likely bars its
`
`patentability”). Fujishima itself provides the motivation to use the alternate shared
`
`sense amplifiers claimed in claim 3.
`
`DC01:1156335.6
`
`1
`
`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`
`
`
`IPR2016-01561
`
`Limestone nonetheless argues that there is no motivation to use the
`
`admittedly beneficial alternate shared sense amplifier because Fujishima describes
`
`another shared sense amplifier embodiment that has the same benefits. However,
`
`if a prior art reference provides two embodiments and describes them both as
`
`beneficial it provides a motivation to use either of them, not neither of them. In re
`
`Fulton, 391 F.3d 1195, 1200 (Fed. Cir. 2004) (“[O]ur case law does not require
`
`that a particular combination must be the preferred, or the most desirable,
`
`combination described in the prior art in order to provide motivation for the current
`
`invention”). Fujishima describes the shared sense amplifier claimed in claim 3,
`
`and describes the benefits of such an arrangement. The fact that Fujishima also
`
`describes another shared sense amplifier arrangement as being beneficial does not
`
`erase or negate the favorable teachings with respect to the amplifiers of claim 3 in
`
`Fujishima.
`
`Limestone’s other arguments are likewise based on fundamental
`
`misunderstandings of the law of obviousness. Limestone spends much of its brief
`
`arguing that a POSITA would not use the alternate shared sense amplifier scheme
`
`of Fujishima with Sukegawa’s any-to-any redundant DRAM because a POSITA
`
`would not want to use the any-to-any redundancy scheme in the first instance.
`
`Limestone bases this argument on the fact that the any-to-any (inter-block)
`
`redundancy scheme of Sukegawa—and claims 1 and 2—can be more complex than
`
`DC01:1156335.6
`
`2
`
`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`an intra-block redundancy scheme. This argument fails for various reasons. First,
`
`IPR2016-01561
`
`
`
`to the extent that any-to-any redundancy schemes are more complicated than intra-
`
`block redundant schemes, that is an inherent feature of the claimed invention itself,
`
`and not a reason that claim 3 is valid. Under Limestone’s logic, claims to less-
`
`than-perfect inventions would be harder to invalidate than claims to optimal
`
`inventions. Second, Sukegawa teaches that an any-to-any redundancy scheme can
`
`be useful with DRAM, and Sukegawa thus is an appropriate starting reference
`
`when considering DRAM redundancy, the subject of the ‘181 patent. That
`
`Sukegawa fully teaches the redundancy scheme of claims 1 and 2 is a fact
`
`Limestone admitted when it conceded the obviousness of those claims. To the
`
`extent that Limestone now has implicitly changed its position and is arguing
`
`Sukegawa actually teaches away from an any-to-any redundancy scheme, it is
`
`simply wrong.1
`
`Limestone also argues that the alternate shared sense amplifier of Fujishima
`
`would not be used in Sukegawa because Fujishima discloses a complicated control
`
`system that could not be incorporated into Sukegawa without adding complexity.
`
`This argument fails for at least two reasons. First, Petitioner is not arguing that the
`
`particular control circuitry of Fujishima would be incorporated verbatim into
`
`
`Such an argument also would be estopped under 37 CFR § 42.73(d)(3).
`1
`
`DC01:1156335.6
`
`3
`
`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`Sukegawa, nor does it have to, because claim 3 requires no particular control
`
`IPR2016-01561
`
`
`
`circuitry. Claim 3 covers DRAMs with simple or complex control circuitry.
`
`Limestone cannot distinguish prior art based on unclaimed limitations. In re
`
`Hiniker Co., 150 F.3d 1362, 1368-69 (Fed. Cir. 1998); In re Self, 671 F.2d 1344,
`
`1348 (C.C.P.A. 1982). Second, it is black-letter law that circuitry from one
`
`reference does not have to be able to be incorporated without change into the
`
`circuitry of another reference in order for those two references to be combined.
`
`MCM Portfolio v. Hewlett-Packard, 812 F.3d 1284, 1294 (Fed. Cir. 2015) (“[T]he
`
`test for obviousness is not whether the features of a secondary reference may be
`
`bodily incorporated into the structure of the primary reference.”). Rather, the
`
`POSITA has the ability take beneficial circuit ideas from one reference and
`
`implement them into other circuits. Id.
`
`The remainder of Limestone’s arguments are equally fruitless. Limestone
`
`notes there are thousands of prior art DRAM references, and Petitioner did not
`
`provide a particular reason for choosing these two references out of those
`
`thousands. The law does not require such an exercise. The POSITA is assumed to
`
`be aware of ALL prior art, no matter how crowded, and can make whatever
`
`combinations are suggested from the art or the problem being faced. See
`
`generally, KSR, 550 U.S. at 419-21. A motivation to combine two references does
`
`not first require an explanation of how those two references are distinguished from
`
`DC01:1156335.6
`
`4
`
`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`every other reference not being relied upon for invalidity. Under Limestone’s
`
`IPR2016-01561
`
`
`
`illogical theory, a crowded art field would be less, not more, likely to lead to
`
`invalid patents, because of the difficulty of saying why a POSITA would use two
`
`particular patents out of the universe of potentially relevant patents.
`
`Limestone also criticizes Dr. Mazumder for not applying the correct
`
`standard for a POSITA, either because he applied his own expert understanding, or
`
`too high of a skill level. This also is wrong. Dr. Mazumder properly looked at the
`
`art through the eyes of a POSITA in considering the issue of obviousness.
`
`Lastly, Limestone argues that the POSITA would not use the alternate
`
`shared sense amplifier of Fujishima with Sukegawa’s DRAM because other
`
`references criticize the complexity of any-to-any redundant schemes. This
`
`argument also fails for multiple reasons. First, it is a repeat of the argument
`
`described above that any-to-any redundant schemes are disfavored. Petitioners are
`
`starting with Sukegawa’s any-to-any redundancy scheme, and the relevant inquiry
`
`is which types of amplifiers could be used with that scheme. A certain degree of
`
`complexity is inherent to such schemes and the claimed invention, and is not a
`
`reason that the claims are valid. Further, this argument relies on prior art
`
`references that are not describing Sukegawa, and therefore runs counter to the
`
`notion that the POSITA is aware of all of the prior art, including Sukegawa.
`
`DC01:1156335.6
`
`5
`
`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`
`
`
`IPR2016-01561
`
`In claim 3, the applicants took an any-to-any redundancy scheme, which is
`
`admittedly disclosed in the prior art Sukagawa reference, and described using it
`
`with a shared sense amplifier scheme, the description of which in the ‘181 patent is
`
`copied nearly verbatim from the prior art, such as Fujishima. Paper 1, at 54; Ex.
`
`1001, ¶ 131. Further, the alternate shared sense amplifier scheme is admittedly
`
`described as being beneficial in Fujishima, and is, in fact, beneficial. Under these
`
`circumstances, the conclusion that claim 3 is obvious is inescapable. The Patent
`
`and Trial Appeal Board (“Board”) was correct to initiate this proceeding, and it
`
`should now find claim 3 unpatentable as obvious.
`
`
`
`As to claim 5, Limestone argues that Walck discusses driving multiple
`
`memory “banks” of DRAM into particular states, not memory “blocks” as required
`
`by claim 5, and therefore the combination allegedly does not teach the limitations
`
`of claim 5. This argument fails to address that Sukegawa discloses the requisite
`
`memory blocks; the only concept being taken from Walck is that multiple arrays of
`
`memory can be simultaneously driven during DRAM refresh. Thus, Limestone
`
`fails to address the references in combination. Further, to the extent Limestone is
`
`arguing that Walck does not teach simultaneously driving multiple memory blocks
`
`into a particular state during DRAM refresh, Limestone is wrong. Walck clearly
`
`discloses a refresh operation in which banks (and, therefore, the memory blocks
`
`included in the banks) are simultaneously driven.
`
`
`
`DC01:1156335.6
`
`6
`
`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`II. ARGUMENT
`A.
`Sukegawa Does Not Teach Away from the Any-to-Any Redundant
`DRAM Claimed in Claims 1 and 2
`
`IPR2016-01561
`
`
`
`Limestone, having previously conceded that Sukegawa renders claims 1 and
`
`2 obvious, does not argue that Sukegawa lacks any elements from those claims.
`
`As explained in the Petition (Paper 1), and found by the Board in its Decision on
`
`Institution (“Decision”), Sukegawa does disclose those elements. Sukegawa is the
`
`closest prior art and should be the starting point for any obviousness determination
`
`with respect to claim 3, which simply adds a shared sense amplifier scheme to
`
`claims 1 and 2.
`
`Limestone nevertheless argues that Sukegawa “teaches away” from its any-
`
`to-any redundancy scheme. This is wrong both factually, as Sukegawa does not
`
`state that the any-to-any scheme is disfavored (in fact, it embraces and claims it),
`
`and legally, since a reference that discloses multiple embodiments does not “teach
`
`away” from those embodiments (even embodiments that are not preferred) absent a
`
`clear discouragement from using the embodiment. Syntex (U.S.A.) LLC v. Apotex,
`
`Inc., 407 F.3d 1371, 1380 (Fed. Cir. 2005). See also Allergan, Inc. v. Apotex, Inc.
`
`754 F.3d 952, 964 (Fed. Cir. 2014) (“mere disclosure of alternative preferences”
`
`not a teaching away); In re Gurley, 27 F.3d 551, 553 (Fed. Cir. 1994) (“A known
`
`or obvious composition does not become patentable simply because it has been
`
`described as somewhat inferior to some other product . . . .”).
`
`DC01:1156335.6
`
`7
`
`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`
`
`
`IPR2016-01561
`
`To begin, the argument that Sukegawa “explicitly teaches away” (Paper 13,
`
`at 18) from the any-to-any redundancy scheme disclosed and claimed therein is
`
`specious. In making this argument, Limestone refers to 2:47-51; 4:51-55; 5:53-57
`
`and 6:17-24, 29-31, and 36-39 of Sukegawa.2 Its expert relies solely on 2:47-51 in
`
`Sukegawa. Ex. 2004, ¶ 74. None of these sections teach away (i.e., clearly
`
`discourages) from any-to-any redundancy, or even state that it is not preferable. At
`
`most, Sukegawa notes that any-to-any redundancy may require the use of
`
`additional components. That is not a “teaching away.” See In re Dance, 160 F.3d
`
`1339, 1344 (Fed. Cir. 1998) (“simplicity of the prior art is rarely a characteristic
`
`that weighs against obviousness of a more complicated device with added
`
`function.”). Sukegawa explicitly embraces any-to-any redundancy as one of its
`
`preferred embodiments and discusses the benefits thereof:
`
`The row redundancy adopts a method that allows ANY TO ANY
`programming for realizing a high yield. When this ANY TO ANY
`redundant mechanism is adopted, the 64 redundant rows in one
`quadrant can be allotted selectively to all quadrants, including the
`present quadrant.
`Ex. 1005 at 2:30-35.
`
`The redundant memory placed in each quadrant can be replaced to the
`main memory of the other quadrants; hence increase in the data lines
`is not hampered.
`Id. at 2:51-54.
`
`2
`Limestone’s citations are tellingly selective—portions of text excised from
`these cites explicitly discuss benefits of the any-to-any redundancy scheme, and
`include it in the description of preferred embodiments.
`8
`
`DC01:1156335.6
`
`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`
`
`
`IPR2016-01561
`
`Here, attention should be paid to the fact that the ANY TO ANY
`method tolerates about four times the defective memory of those in
`the conventional case in the stage with yield over 80% . . . . by using
`the ANY TO ANY method, only 20% of the chips have to be disposed
`of . . . .
`Id. at 3:4-15.
`
`In the redundant mechanism shown in Fig. 8, connection is made
`between one memory quadrant and fuse decoders. In the case when
`the ANY TO ANY method for the redundant mechanism is used, fuse
`decoders 82-94 can function for replacing the defective memory in the
`other quadrants not shown in the figure.
`Id. at 6:29-35.
`
`Further, Sukegawa’s claims cover the any-to-any redundancy scheme.
`
`Claim 1 calls for a redundancy scheme having a plurality of memory blocks, but
`
`does not require that each block have its own redundant memory dedicated only to
`
`that block. Rather, claim 1 requires only that “at least some of the arrays of
`
`memory cells” include redundant memory. Id. at 8:3-6 (emphasis added).
`
`Similarly, claim 5 requires only that alternating memory blocks have the redundant
`
`memory. Id. at 8:49-55. In sum, Sukegawa describes any-to-any redundancy as a
`
`preferred embodiment that increases yield, and claims such redundancy schemes.
`
`This is not a teaching away.
`
`The argument that Fujishima would not be combined with Sukegawa’s any-
`
`to-any redundancy because it is more complicated than an intra-block redundancy
`
`scheme also ignores that any complications that arise from the use of any-to-any
`
`redundancy schemes are a part of the claimed invention. In short, if any-to-any
`
`DC01:1156335.6
`
`9
`
`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`redundancy schemes are more complicated than other possible redundancy
`
`IPR2016-01561
`
`
`
`schemes, the claims themselves embrace those complications, and they cannot be
`
`used to avoid the prior art. In re Mouttet, 686 F.3d 1322, 1334 (Fed. Cir. 2012).
`
`Indeed, Limestone’s analysis, which starts with Fujishima and discusses why
`
`a POSITA would not like the any-to-any redundancy of Sukegawa, is backwards.
`
`As noted above, Sukegawa is directed to the arrangement of redundant memory
`
`cells used to repair defective memory and teaches the redundancy scheme of claim
`
`3, which is the focus of the ‘181 patent, titled “Semiconductor Memory Device
`
`with Improved Flexible Redundancy Scheme.” Limestone has already conceded
`
`that Sukegawa alone renders claims 1 and 2 obvious. Sukegawa is an appropriate
`
`starting point for a POSITA considering redundancy schemes for DRAM. The
`
`only disclosure missing from it is the shared sense amplifier arrangement
`
`additionally claimed in claim 3. Sukegawa does not describe any particular sense
`
`amplifier arrangement. Thus, the question presented is whether a POSITA would
`
`know to use, for instance, an alternating shared sense amplifier such as those
`
`disclosed in Fujishima. Since the amplifiers are being used exactly as intended, the
`
`answer is yes. Paper 1, at 60-63; Ex. 1001, ¶¶ 46 and 141-43. KSR, 550 U.S. at
`
`416 (Where “a patent claims a structure already known in the prior art that is
`
`altered by the mere substitution of one element for another known in the field, the
`
`combination must do more than yield a predictable result.”).
`
`DC01:1156335.6
`
`10
`
`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`
`
`
`IPR2016-01561
`
`B. The Motivation to Use Fujishima’s Sense Amplifiers Is Found in
`Fujishima Itself
`
`Limestone completely discounts the benefits that Fujishima ascribes to its
`
`shared sense amplifiers because those benefits are present in the alternate shared
`
`sense amplifier array of Fig. 14, as well as the shared sense amplifier array of Fig.
`
`1. Limestone thus admits that the alternate shared sense amplifier of Figure 14 is
`
`beneficial, and has the benefits set forth in Fujishima. Paper 13, at 27-28. See also
`
`Ex. 1019 at 144:14-145:17, 93:9-94:19. This is fatal to its case.
`
`Limestone would have the Board hold that when a prior art reference
`
`discloses two beneficial embodiments, it provides no motivation to use either
`
`embodiment, because it does not particularly favor one over the other. That is
`
`nonsensical. Legions of cases hold that a prior art embodiment need not be
`
`preferred in order to be used successfully in an obviousness combination. See,
`
`e.g., In re Fulton, 391 F.3d at 1201 (“[T]he prior art’s mere disclosure of more
`
`than one alternative does not constitute a teaching away from any of these
`
`alternatives because such disclosure does not criticize, discredit, or otherwise
`
`discourage the solution claimed....”); Tyco Healthcare Group LP v. Ethican Endo-
`
`Surgery, Inc., 774 F.3d 968, 977 (Fed. Cir. 2014); Merck & Co. v. Biocraft Labs,
`
`Inc., 874 F.2d 804, 807 (Fed. Cir. 1989). A fortiori, a prior art embodiment need
`
`not be described as the only embodiment to have a particular benefit before there is
`
`a motivation to use that embodiment. Apple Inc. v. Samsung Elecs. Co., 816 F.3d
`
`DC01:1156335.6
`
`11
`
`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`788, 801 (Fed. Cir. 2016) (“[A] motivation to use the teachings of a particular prior
`
`IPR2016-01561
`
`
`
`art reference need not be supported by a finding that that feature be the ‘preferred,
`
`or the most desirable.’”). Limestone’s argument undermines one of the
`
`foundations of KSR, which is that a POSITA is not an automaton, and can use
`
`known embodiments for their intended purposes in various contexts. KSR, 550
`
`U.S. at 420-21.
`
`Further, Dr. Mazumder explained in his deposition that, in fact, alternate
`
`shared sense amplifiers do have additional benefits that other shared sense
`
`amplifiers do not. Ex. 2007 at 234:3-46:18, 172:6-20, 174:4-20. For instance, in
`
`an alternate shared sense amplifier arrangement, each band has half as many
`
`amplifiers, because the amplifiers are only sensing either the odd or even lines in
`
`the adjacent memory blocks. This provides greater room for each amplifier,
`
`reducing manufacturing errors and decreasing noise (parasitic capacitance). Id.
`
`Limestone also suggests that there is no motivation to combine because
`
`Petitioner did not explain why the POSITA would choose Fujishima to combine
`
`with Sukegawa from among the thousands of prior art references that are not being
`
`relied upon for invalidity. Paper 13, at 65. However, the POSITA is assumed to
`
`have knowledge of all prior art in a field, and can combine any two references that
`
`might be useful to combine. KSR, 550 U.S. at 420. The references need not be the
`
`two most optimal references in the entire universe of possible references. Taken to
`
`DC01:1156335.6
`
`12
`
`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`its conclusion, Limestone’s argument would mean that patent claims in crowded
`
`IPR2016-01561
`
`
`
`art fields are more likely to be valid than claims in wide open art fields, which is
`
`illogical. Similarly, claimed inventions that themselves incorporate shortcomings
`
`or are inferior in some way (such as being more complex) would be more likely to
`
`be held valid under Limestone’s logic, because the patent owner could in effect
`
`argue that the shortcomings of his own invention prevent the POSITA from ever
`
`wanting to make that invention.
`
`Here, Sukegawa discloses the any-to-any redundant scheme of claims 1 and
`
`2, and states that such a scheme can improve yield (which it does, by allowing for
`
`more flexible replacement of memory). These are the same goals the ‘181 patent is
`
`attempting to achieve with its inter-block memory replacement. The only thing
`
`missing from Sukegawa with respect to claim 3 is the shared sense amplifier
`
`scheme additionally disclosed in claim 3. But, as set forth in the Petition, the
`
`alternate shared sense amplifier shown in Figure 14 of Fujishima has certain
`
`known benefits that are set forth in that reference and thus provide to a POSITA a
`
`motivation to combine Fujishima with Sukegawa. The fact that another
`
`embodiment of Fujishima also has these benefits does not remove the motivation
`
`to use the embodiment of Figure 14. The shared sense amplifiers in the ‘181
`
`patent were known in the prior art to be beneficial and are being used as intended
`
`to sense the adjacent memory arrays. This is the definition of an obvious
`
`DC01:1156335.6
`
`13
`
`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`combination. KSR, 550 U.S. at 416 (“The combination of familiar elements
`
`IPR2016-01561
`
`
`
`according to known methods is likely to be obvious when it does no more than
`
`yield predictable results.”).
`
`C. Claim 3 Does Not Mention Control Circuitry, and Fujishima’s
`Control Circuitry Is Irrelevant to the Analysis of Claim 3
`
`Limestone argues that the alternate shared sense amplifier scheme of
`
`Fujishima could not be combined with Sukegawa because its allegedly
`
`complicated control circuitry for the amplifiers could not be imported into
`
`Sukegawa without being reworked. This argument fails for two simple reasons.
`
`First, claim 3 does not require any particular control circuitry—it equally covers
`
`devices with elegant control circuitry, and devices with cumbersome control
`
`circuitry. Thus, the fact that Fujishima’s control circuitry is less than ideal
`
`(according to Limestone) is irrelevant. See In re Self, 671 F.2d at 1348. Petitioner
`
`is not relying on Fujishima’s disclosure of control circuitry for its proposed
`
`combination, merely its disclosure of the shared sense amplifiers, which is all that
`
`is actually claimed by Claim 3.3
`
`
`3
`This is not changed by the fact that Fujishima’s claimed invention related to
`control circuitry. See KSR, 550 U.S. at 420; In re Lemelson, 397 F.2d 1006, 1009
`(C.C.P.A. 1968) (“The use of patents as references is not limited to what the
`patentees describe as their own inventions or to the problems with which they are
`concerned.”).
`
`DC01:1156335.6
`
`14
`
`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`
`
`
`IPR2016-01561
`
`Second, it is well established that circuitry from one reference (even claimed
`
`circuitry, which the control circuitry is not) need not be able to be physically
`
`incorporated directly into another reference’s circuit in order for the teachings of
`
`the references to be combined. See MCM, 812 F.3d at 1294; In re Mouttet, 686
`
`F.3d at 1332; In re Etter, 756 F.2d 852, 859 (Fed. Cir. 1985) (en banc); In re
`
`Sneed, 710 F.2d 1544, 1550 (Fed. Cir. 1983); In re Keller, 642 F.2d 413, 425
`
`(CCPA 1981). The fact that surrounding circuitry would have to be “reworked,” as
`
`opposed to copied rotely from one circuit into another, is not grounds for a finding
`
`of nonobviousness. A POSITA here is a memory designer with years of
`
`experience. Not even Dr. Khatri claims that a POSITA could not incorporate an
`
`alternate shared sense amplifier into Sukegawa’s DRAM.
`
`D. Dr. Mazumder Properly Considered Obviousness Through the
`Lens of a POSITA at the Time of Invention, and the Dispute
`Between the Parties as to the Level of Skill Is Irrelevant Here
`
`Dr. Mazumder properly considered the requirements of 35 U.S.C. § 103,
`
`especially as it related to a POSITA and the corresponding level of ordinary skill in
`
`the art. Ex. 1001, ¶¶ 23-25 and 28-30; and p. 17, n.4; Ex. 2007 at 73:18-87:14.
`
`Dr. Mazumder considered his own experience, for instance, in determining what
`
`the level of a POSITA was. He did not say that he undertook the obviousness
`
`analysis using himself as a POSITA. Further, both experts have said that their
`
`analysis is not affected by the minor dispute between the parties over the level of a
`
`DC01:1156335.6
`
`15
`
`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`POSITA. Ex. 1001, p. 17, n.4; Ex. 2004, ¶ 33. Dr. Mazumder did not
`
`
`
`IPR2016-01561
`
`misunderstand the nature of the obviousness inquiry.
`
`If either expert took a legally-improper or unsupported approach to
`
`obviousness, it was Dr. Khatri. He had no understanding of whether DRAM
`
`suppliers in the 1990’s were actually incorporating alternate shared sense
`
`amplifiers or inter-block memory redundancy into DRAMs. Ex. 1019 at 52:6-
`
`54:8, 56:3-11, 66:12-67:5, 75:16-78:17. He couldn’t say for sure whether he had
`
`even heard of alternate shared sense amplifiers or inter-block memory redundancy
`
`prior to June 1998. Id. at 75:16-78:17. The first time he had read any of the
`
`materials he considered was after he was retained in this matter. Id. at 40:4-14,
`
`42:1-18. He argued against the Petitioner’s combination of references because
`
`they allegedly did not solve the same problem as the ‘181 patent (Ex. 2004, ¶ 109)
`
`-- reasoning that was explicitly rejected in KSR. KSR, 550 U.S. at 420-21. And,
`
`as explained supra at 7-10 and infra at 17-19, in testifying that Sukegawa and
`
`Horiguchi’s 1997 article “teach away” from the proposed combination, he
`
`evidences no understanding of what a teaching away actually is.
`
`E.
`
`There is No General “Teaching Away” from the Combination of
`Sukegawa and Fujishima
`
`
`
`To underscore the ubiquity of alternate shared sense amplifiers, the Petition
`
`cited various articles describing such amplifiers, largely in the exact same manner
`
`DC01:1156335.6
`
`16
`
`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`that the ‘181 patent describes them. Paper 1, at 12 and 23. Dr. Mazumder likewise
`
`IPR2016-01561
`
`
`
`underscored the ubiquity of such amplifiers at the time of invention. Ex. 1001, ¶¶
`
`46, 104, 139, 143, at p. 56, n.21; Ex. 2007 at 174:4-176:3, 234:3-236:14.
`
`Limestone’s statement that Petitioner has offered only “attorney argument” to
`
`support the fact that alternate shared sense amplifiers were a well-known design
`
`choice (Paper 13, at 40) is wrong.
`
`While ignoring that the cited articles disclose the shared sense amplifier
`
`scheme claimed in claim 3, Limestone attempts to use these articles to its
`
`advantage to argue that they teach that such amplifiers should not be used with
`
`“disfavored” any-to-any redundancy schemes. 4 Limestone’s argument does not
`
`save claim 3.
`
`
`
`To begin, by 1997, as shown in the 1997 Horiguchi article (Ex. 2012 at 26-
`
`27), inter-block memory replacement was accepted in the prior art. See also Ex.
`
`2007 at 139:21-143:1, 221:19-222:9. This article explicitly discloses inter-block
`
`memory replacement and removes the 1991 language quoted by Limestone
`
`
`4
`Largely, the references are silent regarding whether any-to-any redundancy
`should be used with alternate shared sense amplifiers. As the Board recognized in
`its Decision (Paper 11, at 8), Limestone essentially argues that any reference
`failing to teach the combination being asserted “teaches away” from that
`combination.
`
`DC01:1156335.6
`
`17
`
`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`referring to such schemes as cumbersome.5 To the contrary, Horiguchi in 1997
`
`IPR2016-01561
`
`
`
`concludes:
`
`Enhancing the replacement flexibility between defective lines and
`spare lines through the flexible intra-subarray replacement or through
`inter-subarray replacement is effective for DRAMs of increased
`memory array division.
`
`Id. at 29 (emphasis added). This is in line with the inter-s