`
`____________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`____________
`
`
`
`
`
`APPLE INC., HTC CORPORATION, HTC AMERICA, INC., SAMSUNG
`ELECTRONICS CO. LTD, SAMSUNG ELECTRONICS AMERICA, INC., AND
`AMAZON.COM, INC.,
`
`Petitioners,
`
`v.
`
`MEMORY INTEGRITY, LLC,
`
`Patent Owner
`
`____________
`
`
`
`Case IPR2015-00159
`Patent 7,296,121
`____________
`
`
`
`PETITIONER’S MOTION FOR OBSERVATIONS
`
`
`
`Proceeding No. IPR2015-00159
`Attorney Docket No. 39521-0007IP1
`Observation #1 In the transcript of the January 8, 2016 deposition of Dr. Vojin
`
`Oklobdzija (Ex. 1032) at page 20, line 9 to page 22, line 4, Dr. Oklobdzija agreed
`
`that “it is a matter of semantics” whether “[o]ne can consider a node to be the one
`
`that contains the hub and both processors main memory or one can consider the
`
`processing node just to be the same as the processor.” This testimony is relevant to
`
`paragraph 9 of Dr. Oklobdzija’s Reply Declaration (Ex. 2042) and page 8 of MI’s
`
`Reply in Support of Motion to Amend (Paper 37) where Dr. Oklobdzija asserts that
`
`Petitioners “are vague with respect to whether they contend that the ‘hub’ is outside
`
`of or is subsumed within a ‘processing node.’” This testimony is relevant, because
`
`it proves that the Opposition’s application of the individual R10000 processor in the
`
`Origin system to the claimed “processing node” is valid.
`
`Observation #2 In Ex. 1032 at page 28, line 8 to page 29, line 20, Dr. Oklobdzija
`
`testified that he did not consider “where the probe in a hub-to-hub transmission
`
`originates from.” This testimony is relevant to paragraph 13 of Dr. Oklobdzija’s
`
`Reply Declaration and pages 9-10 of MI’s Reply in Support of Motion to Amend,
`
`where Dr. Oklobdzija asserts that “there is no reason to believe from the teachings
`
`of Culler and Laudon that any message originating from a processor in a request
`
`node is the same as the alleged ‘probe’ received by the hub in a home node such that
`
`it could be said that the ‘probe filtering unit . . . receive probes . . . from the
`
`processing nodes.’” This testimony is relevant because it demonstrates that Dr.
`
`
`
`1
`
`
`
` Case IPR2015-00159
`Attorney Docket No: 39521-0007IP1
`
`Oklobdzija did not identify an alternative source for a probe in the Origin system
`
`other than a requesting processor.
`
`Observation #3 In Ex. 1032 at page 52, line 10 to page 53, line 12, Dr. Oklobdzija
`
`testified that he did “not rely[] upon any language in the '121 patent regarding the
`
`message format of a probe as supporting [his] opinion in paragraph 13” and that “the
`
`substitute claims do not recite any specific message formats of the probe.” This
`
`testimony is relevant to paragraph 13 of Dr. Oklobdzija’s Reply Declaration and
`
`pages 9-10 of MI’s Reply in Support of Motion to Amend, where Dr. Oklobdzija
`
`cites to Culler’s teaching that the Origin system relies on varying message formats
`
`to transport data to support his contention that “messaging between hubs is also
`
`significantly different from the messaging within hubs.” This testimony is relevant
`
`because it demonstrates that, contrary to Dr. Oklobdzija’s implication, a change in a
`
`probe’s message format is not relevant to the limitations of the substitute claims.
`
`Observation #4 In Ex. 1032 at page 73, line 18 to page 75, line 9, Dr. Oklobdzija
`
`testifies that the ’121 Patent describes an embodiment in which a probe traveling
`
`between clusters traverses communication links using different protocols that
`
`require different message formats. This testimony is relevant to paragraph 13 of Dr.
`
`Oklobdzija’s Reply Declaration and pages 9-10 of MI’s Reply in Support of Motion
`
`to Amend, where Dr. Oklobdzija cites to Culler’s teaching that the Origin system
`
`relies on varying message formats to transport data to support his contention that
`
`2
`
`
`
` Case IPR2015-00159
`Attorney Docket No: 39521-0007IP1
`
`“messaging between hubs is also significantly different from the messaging within
`
`hubs.” This testimony is relevant because it demonstrates that, contrary to Dr.
`
`Oklobdzija’s implication, a change in a probe’s message format is not relevant to the
`
`limitations of the substitute claims.
`
`Observation #5 In Ex. 1032 at page 78, lines 2 to 6, after taking time to review the
`
`Culler reference, Dr. Oklobdzija agrees that “processors issue read requests to the
`
`hub chip through the PI” (i.e., the processor interface of the hub chip). Moreover,
`
`in Ex. 1032 at page 175, lines 17 to 25 and page 176, line 9 to page 176, line 19, Dr.
`
`Oklobdzija admits that the processors issue read requests, which are passed through
`
`the hub, and numbered in order to keep track of where the read requests are in the
`
`system. This testimony is relevant to paragraph 13 of Dr. Oklobdzija’s Reply
`
`Declaration and pages 9-10 of MI’s Reply in Support of Motion to Amend, where
`
`Dr. Oklobdzija asserts that “it would not be reasonable to assume that requests
`
`received by a hub in a home node in SGI Origin is the same or even a modified
`
`version of a message sent by a requesting processor attached to a different hub.”
`
`This testimony is relevant because Dr. Oklobdzija’s Reply Declaration does not
`
`present any evidence that the read request received by the hub in the home node is
`
`substantively different from the read request sent by the processor, nor could Dr.
`
`Oklobdzija testify to any changes. See, e.g., Ex. 1032 at 56:25-57:24; 65:22-66:5;
`
`68:21-69:3; 69:12-21; 167:8-168:7.
`
`3
`
`
`
` Case IPR2015-00159
`Attorney Docket No: 39521-0007IP1
`
`Observation #6 In Ex. 1032 at page 59, line 5 to page 60, line 5, Dr. Oklobdzija
`
`testifies that the substitute claims’ requirement that a probe received by the probe
`
`filtering unit “has to be the same probe which corresponds to memory lines from the
`
`processing node” is “not talking about a specific probe filtering format that you're
`
`talking about, but the information basically that is contained in the probe filtering
`
`message should be the same on all PFU.” Moreover, at page 67, line 15 to page 68,
`
`line 4, Dr. Oklobdzija testifies that as long as any modification to a probe does not
`
`affect the response solicited from the system, it is the same probe for purposes of the
`
`substitute claims. This testimony is relevant to paragraph 13 of Dr. Oklobdzija’s
`
`Reply Declaration and pages 9-10 of MI’s Reply in Support of Motion to Amend,
`
`where Dr. Oklobdzija asserts that “it would not be reasonable to assume that requests
`
`received by a hub in a home node in SGI Origin is the same or even a modified
`
`version of a message sent by a requesting processor attached to a different hub.”
`
`This testimony is relevant because Dr. Oklobdzija’s Reply Declaration does not
`
`assert that the read request received by the hub in a home node does not solicit the
`
`same response as the read request issued by the requesting processor.
`
`Observation #7 In Ex. 1032 at page 45, line 2 to page 47, line 5 and page 49, line
`
`19 to page 50, line 4, Dr. Oklobdzija admits that a probe received by a cache
`
`coherence controller containing the claimed probe filtering unit from a processor in
`
`a remote cluster would “receive” that probe from the processor in accordance with
`
`4
`
`
`
` Case IPR2015-00159
`Attorney Docket No: 39521-0007IP1
`
`the substitute claims, even though the cache coherence controller of the remote
`
`cluster would have to “forward the request – the probe.” This testimony is relevant
`
`to paragraph 14 of Dr. Oklobdzija’s Reply Declaration and pages 10-11 of MI’s
`
`Reply in Support of Motion to Amend, where Dr. Oklobdzija asserts that “‘[i]n a
`
`point-to-point architecture,’ there are ‘multiple processors directly connected to each
`
`other through point-to-point links[,]’” so “processors in separate nodes in the SGI
`
`Origin are not directly connected to each other, as required by the substitute claims
`
`and the express teachings of the ’121 Patent” because “they are connected through,
`
`at a minimum, two hub chips” (emphasis in original). This testimony is relevant
`
`because it contradicts Dr. Oklobdzija’s assertion that the ’121 Patent requires the
`
`processing nodes to be “directly” connected, as opposed to connected through two
`
`probe filtering units (i.e., hub chips).
`
`Observation #8 In Ex. 1032 at page 79, line 6 to page 80, line 8, Dr. Oklobdzija
`
`admits that the substitute claims do not require processors that are directly connected
`
`through point-to-point links. Moreover, at page 44, line 5 to page 45, line 18 and
`
`page 79, line 10 to page 81, line 10, Dr. Oklobdzija admits that the point-to-point
`
`architecture in the ‘121 patent’s Figure 2 includes processors that are not directly
`
`connected. This testimony is relevant to paragraph 14 of Dr. Oklobdzija’s Reply
`
`Declaration and pages 10-11 of MI’s Reply in Support of Motion to Amend, where
`
`5
`
`
`
` Case IPR2015-00159
`Attorney Docket No: 39521-0007IP1
`
`Dr. Oklobdzija asserts that “‘[i]n a point-to-point architecture,’ there are ‘multiple
`
`processors directly connected to each other through point-to-point links.”
`
`Observation #9 In Ex. 1032 at page 90, lines 3 to 6 and page 92, line 2 to page 93,
`
`line 24, Dr. Oklobdzija admits that there are different types of buses, agrees that “the
`
`’121 patent is comparing a conventional shared bus that is used in a system that
`
`performs snooping against a point-to-point architecture,” and agrees that the
`
`processors of the Origin system do not perform snooping on the SysAD bus. This
`
`testimony is relevant to paragraph 14 of Dr. Oklobdzija’s Reply Declaration and
`
`pages 10-11 of MI’s Reply in Support of Motion to Amend, where Dr. Oklobdzija
`
`asserts that “’121 patent expressly distinguishes” buses like the Origin system’s
`
`SysAD bus “from point-to-point interfaces.” This testimony is relevant because it
`
`demonstrates that the Origin system’s SysAD bus is not the same type of bus
`
`distinguished by the ’121 patent.
`
`Observation #10 In Ex. 1032 at page 102, line 7 to page 103, line 19, Dr.
`
`Oklobdzija acknowledged that, in the implementation of the Origin system relied
`
`upon by Petitioner’s, “there would only be two endpoints for this path connecting
`
`processor A and the hub” (i.e., the SysAD bus), and that all communications “would
`
`only go through these two points.” This testimony is relevant to paragraph 14 of Dr.
`
`Oklobdzija’s Reply Declaration and pages 10-11 of MI’s Reply in Support of Motion
`
`to Amend, where Dr. Oklobdzija asserts that, “within a local node in SGI Origin,
`
`6
`
`
`
` Case IPR2015-00159
`Attorney Docket No: 39521-0007IP1
`
`individual processors are connected via a ‘SysAD bus,’ which is also used to connect
`
`the processors to the hub chip,” which the “’121 patent expressly distinguishes such
`
`buses from point-to- point interfaces.” This testimony is relevant, because it
`
`demonstrates that the SysAD bus in Origin is a point-to-point link, as opposed to a
`
`snooping-based shared medium.
`
`Observation #11 In Ex. 1032 at page 94, line 4 to page 98, line 16, Dr. Oklobdzija
`
`acknowledges that the goal of using point-to-point links in the ’121 Patent, as
`
`compared to a conventional shared bus that performs snooping, is the same as the
`
`goal of the manner in which the Origin system utilizes the SysAD bus (i.e.,
`
`“reduc[ing] latency and increase[ing] bandwith”). This testimony is relevant to
`
`paragraph 14 of Dr. Oklobdzija’s Reply Declaration and pages 10-11 of MI’s Reply
`
`in Support of Motion to Amend, where Dr. Oklobdzija asserts that, “within a local
`
`node in SGI Origin, individual processors are connected via a ‘SysAD bus,’ which
`
`is also used to connect the processors to the hub chip,” which the “’121 patent
`
`expressly distinguishes such buses from point-to- point interfaces.” This testimony
`
`is relevant, because it demonstrates that the SysAD bus in Origin serves the same
`
`goals as the point-to-point architecture in the ’121 Patent, and, even if the SysAD
`
`bus were not considered a point-to-point link, it would have been obvious to “just
`
`use[] point to point.”
`
`7
`
`
`
` Case IPR2015-00159
`Attorney Docket No: 39521-0007IP1
`
`Observation #12 In Ex. 1032 at page 113, lines 15-19 and page 179, line 19 to page
`
`180, line 16, after reviewing the ’121 Patent’s disclosure as a result of questioning
`
`on redirect, Dr. Oklobdzija acknowledges that his previous interpretation of what
`
`constitutes a “point-to-point architecture,” as that term is used in the ’121 Patent has
`
`been “incorrect.” After this realization, at page 181, line 7 to page, 182, line 2, Dr.
`
`Oklobdzija acknowledges that the cube architecture of the Origin system shown in
`
`FIG. 3 of the Laudon reference “is a point to point architecture.” This testimony is
`
`relevant to paragraph 14 of Dr. Oklobdzija’s Reply Declaration and pages 10-11 of
`
`MI’s Reply in Support of Motion to Amend, where Dr. Oklobdzija asserts that, “’[i]n
`
`a point-to-point architecture,’ there are ‘multiple processors directly connected to
`
`each other through point-to-point links.’” (emphasis in original). This testimony is
`
`relevant, because it demonstrates that cube architecture of the Origin system is a
`
`point-to-point architecture.
`
`Observation #13 In Ex. 1032 at page 125, line 23 to page 127, line 6, Dr.
`
`Oklobdzija admits that in a “system that communicates with a separate memory
`
`which is not cached and, therefore, has no coherency because it's not cached, then
`
`what you're trying to say, it fits the [Patent Owner’s proposed] definition of a
`
`noncoherent interface,” though he still distinguishes such a system from a
`
`noncoherent interface. This testimony is relevant to paragraph 15 of Dr.
`
`Oklobdzija’s Reply Declaration and pages 11-12 of MI’s Reply in Support of Motion
`
`8
`
`
`
` Case IPR2015-00159
`Attorney Docket No: 39521-0007IP1
`
`to Amend, where Dr. Oklobdzija notes that, “in discussing how Origin processes I/O
`
`(which is typically associated with non-coherent transactions), the [Culler] reference
`
`describes the ‘Xbow’ interface, which ‘connects the Hub to other I/O interfaces,’ as
`
`permitting not only non-coherent ‘uncached’ I/O operations, but also ‘coherent
`
`DMA operations.’” This testimony is relevant, because Dr. Oklobdzija admits that
`
`the “Xbow” interface fits “the words” of Patent Owner’s proposed definition of a
`
`non-coherent interface when the “Xbow” interface “communicates with a separate
`
`memory which is not cached.”
`
`Observation #14 In Ex. 1032 at page 130, line 13 to page 131, line 8, Dr.
`
`Oklobdzija acknowledges that a POSITA would have been able to convert a shared
`
`coherent and non-coherent interface into “separate noncoherent interface and
`
`coherent protocol interfaces.” This testimony is relevant to paragraph 15 of Dr.
`
`Oklobdzija’s Reply Declaration and pages 11-12 of MI’s Reply in Support of Motion
`
`to Amend, where Dr. Oklobdzija asserts that “Culler describes SGI Origin’s
`
`coherent and non-coherent operations as being part of a single ‘Origin protocol’
`
`occurring over the same ‘interfaces.’” This testimony is relevant because it
`
`demonstrates that a POSITA would have been able to convert the shared interface
`
`of the Origin protocol into “separate noncoherent interface and coherent protocol
`
`interfaces.”
`
`9
`
`
`
` Case IPR2015-00159
`Attorney Docket No: 39521-0007IP1
`
`Observation #15 In Ex. 1032 at page 132, lines 14 to 20, Dr. Oklobdzija testifies
`
`that it’s his “opinion that the substitute claims require that at least some portion of
`
`the coherent protocol interface and a noncoherent protocol interface are separate
`
`structures.” This testimony is relevant to footnote 5 on page 25 of Patent Owner’s
`
`Motion to Amend, which states that “[i]t should be understood that Patent Owner
`
`does not intend to suggest that the coherent protocol interface and non-coherent
`
`protocol interface of the proposed substitute claims need be implemented in a way
`
`that requires particular separate structures. Nothing in the intrinsic record of the ’121
`
`Patent restricts the interfaces in that manner.” This testimony is relevant because it
`
`indicates Dr. Oklobdzija applied a claim construction inconsistent with the ’121
`
`Patent and the Patent Owner’s proposed claim construction when reaching his
`
`conclusion that “SGI Origin does not disclose the limitations of the proposed
`
`substitute claims that the ‘probe filtering unit is coupled to a coherent protocol
`
`interface and a non-coherent protocol interface.’”
`
`Observation #16 In Ex. 1032 at page 132, line 22 to page 133, line 10, Dr.
`
`Oklobdzija agrees “that a system that had logic for performing cached and uncached
`
`operations across an interface would meet the claim limitations of a coherent
`
`protocol interface and a noncoherent protocol interface in the substitute claims.”
`
`Next, at page 134, line 14 to page 135, line 6 and page 136, lines 15 to 23, Dr.
`
`Oklobdzija admits that the Origin system “must have” some logic for performing
`
`10
`
`
`
` Case IPR2015-00159
`Attorney Docket No: 39521-0007IP1
`
`uncached operations. Then, at page 137, line 8 to page 140, line 10, Dr. Oklobdzija
`
`admits that, in the Origin system, “uncached operations are the accesses to memory,
`
`and the hub is in between the processor and memory, so there has to have a logic to
`
`access the memory for cached and uncached operations.” This testimony is relevant
`
`to paragraph 15 of Dr. Oklobdzija’s Reply Declaration and pages 11-12 of MI’s
`
`Reply in Support of Motion to Amend, where Dr. Oklobdzija asserts that “the
`
`proposed claims require a probe filtering unit which is coupled to physically or
`
`logically distinct coherent and non-coherent protocol interfaces.” This testimony is
`
`relevant because it demonstrates that the Origin system has logic for performing
`
`cached and uncached operations across an interface, which Dr. Oklobdzija agrees
`
`meets the requirements of the proposed substitute claims.
`
`Respectfully submitted,
`
`
`
`
`
`/ Roberto J. Devoto /
`
`Roberto Devoto, Reg. No. 55,108
`Fish & Richardson P.C.
`P.O. Box 1022
`Minneapolis, MN 55440-1022
`T: 202-783-6830
`F: 202-783-2331
`Attorneys for Petitioner
`
`
`
`
`
`
`
`
`11
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`Dated: 1-11-16
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`Proceeding No. IPR2015-00159
`Attorney Docket No. 39521-0007IP1
`CERTIFICATE OF SERVICE
`
`Pursuant to 37 CFR §§ 42.6(e)(4) and 42.205(b), the undersigned certifies that
`
`on January 11, 2016, a complete and entire copy of this Petitioner’s Motion for
`
`Observations was provided via email to the Patent Owner by serving the
`
`correspondence email address of record as follows:
`
`
`Jonathan D. Baker
`Michael D. Saunders
`Gurtej Singh
`Farney Daniels PC
`411 Borel Avenue, Suite 350
`San Mateo, California 94402
`
`
`
`Email:
`
`jbaker@farneydaniels.com
`msaunders@farneydaniels.com
`tsingh@farneydaniels.com
`MemoryIntegrityIPR@farneydaniels.com
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`/Diana Bradley/
`Diana Bradley
`Fish & Richardson P.C.
`60 South Sixth Street, Suite 3200
`Minneapolis, MN 55402
`(858) 678-5667