`(880 Patent)
`
`IPR2017-01409 (Intel)
`IPR2017-01736 (Cavium)
`IPR2018-00338 (Dell)
`IPR2017-01410 (Intel)
`IPR2017-1737 (Cavium)
`IPR2018-0339 (Dell)
`
`*All citations herein are to the IPR2017-01391 case unless otherwise noted.
`
`207
`
`
`
`880 Patent: Instituted Grounds
`
`• 2017IPR-01409, IPR2017-01736, IPR2018-00338
`• Ground 1: Thia (Ex. 1015) in view of Tanenbaum96 (Ex.1006)
`• Claims 1, 5-10, 12, 14, 16, 17, 20-23, 27, 28, 45, and 55
`• 2017IPR-01410, IPR2017-01737, IPR2018-00339
`• Ground 1: Thia (Ex. 1015) in view of Tanenbaum96 (Ex. 1006)
`• Claims 32, 34, 35, 39, 41, 42, and 43
`• Ground 2: Thia (Ex. 1015) in view of Tanenbaum96 (Ex. 1006) and
`Nahum (Ex. 1079)
`• Claims 37 and 38
`
`Ex. 1006 – Tanenbaum, Andrew S., Computer Networks (“Tanenbaum96”)
`Ex. 1015 – Tia, Y.H., Woodside, C.M. Publication (“Thia”)
`Ex. 1079 – Nahum, Erich, Professional Issues in Parallelized Network Protocols (“Nahum”)
`
`208
`
`
`
`880 Patent: Disputes
`
`1. A POSA would have combined Thia and
`Tanenbaum96 (and Nahum)
`
`2. Thia and Nahum are enabling
`
`3. The prior art combinations disclose the limitations of the
`challenged claims of the 880 Patent
`
`4. Motions to Amend 880 Patent should be denied
`
`209
`
`
`
`880 Patent: Disputes
`
`1. A POSA would have combined Thia and Tanenbaum96 (and
`Nahum)
`a. A POSA would have understood that Thia’s teachings
`are applicable to TCP/IP
`b. The trend towards TCP/IP in the 1990s would motivate
`combining Thia’s bypass architecture with TCP/IP
`c. Tanenbaum96 does not teach away from the combination
`d.
`It would have been obvious to combine Nahum with Thia
`and Tanenbaum96
`
`210
`
`
`
`Both disclose a bypass/fast-path based
`on TCP/IP header prediction
`
`Ex. 1015 (Thia) at .002;
`See also Paper 1 (1409 Petition) at 32-31; Paper 1 (1410 Petition) at 33-34, 36-37.
`
`Ex. 1006.585 (Tanenbaum96);
`See also Paper 1 (1409 Petition) at 32-31; Paper 1 (1410 Petition) at 33-34, 36-37.
`
`211
`
`
`
`Thia’s teachings are not limited to OSI
`
`Ex. 1015.001 (Thia); See also Paper 1 (1409 Petition) at 24-25, Paper 42 (1409 Reply) at 9; Ex.1223.016-.017 (1409 Lin Reply
`Decl.) at ¶ 26; Paper 1 (1410 Petition) at 25; Paper 42 (1410 Reply) at 7-8; Ex.1223.026-.027 (1410 Lin Reply Decl.) at ¶ 37.
`
`212
`
`
`
`Thia’s standard protocol stack (SPS) is
`a “multi-layer” stack, not an “OSI” stack
`
`Ex. 1015.003 (Thia); See also Paper 1 (1409 Petition) at 25, 30, 34, 35; Paper 42 (1409 Reply) at 9-
`11; Paper 1 (1410 Petition) at 26, 31-32, 35, 40; Paper 42 (1410 Reply) at 7-9.
`
`213
`
`
`
`Thia teaches that its bypass offload is
`for more than one multi-layer stack
`
`Ex. 1015.005 (Thia); See also Paper 42 (1409 Reply) at 9;
`Paper 42 (1410 Reply) at 7.
`
`214
`
`
`
`TCP/IP and OSI were widely understood
`to be very similar
`
`Ex. 1003.011 (IPR2017-1409 Lin Decl.); See also Paper 1 (1409 Petition) at 30-35; Ex.1003.068-.074 (1409 Lin
`Decl.); Paper 1 (1410 Petition) at 31-40; Ex.1003.069-.080 (1410 Lin Decl.).
`
`215
`
`
`
`Layered protocols mean TCP/IP can be
`substituted for OSI
`
`Ex. 1006.045-.046 (Tanenbaum96);
`See also Paper 1 (1409 Petition) at 16, 34-35;
`Paper 1 (1410 Petition) at 16-17, 35-36, 39-40.
`
`216
`
`
`
`880 Patent: Disputes
`
`1. A POSA would have combined Thia and Tanenbaum96 (and
`Nahum)
`a. A POSA would have understood that Thia’s teachings are
`applicable to TCP/IP
`b. The trend towards TCP/IP in the 1990s would motivate
`combining Thia’s bypass architecture with TCP/IP
`c. Tanenbaum96 does not teach away from the combination
`d.
`It would have been obvious to combine Nahum with Thia
`and Tanenbaum96
`
`217
`
`
`
`By 1996 OSI protocol use vanished and
`TCP/IP became dominant
`
`Ex. 1006.016 (Tanenbaum96);
`See also Paper 1 (1409 Petition) at 28, 32-33; Paper 1 (1410 Petition) at 28, 34-35.
`
`218
`
`
`
`Thia’s hardware offload provides
`advantages over software alone
`
`Ex. 1015.013 (Thia);
`See also Paper 1 (1409 Petition) at 33-34, 41;
`Ex.1003.060, .072-.073 (1409 Lin Decl.);
`Paper 1 (1410 Petition) 34-35, 37-38, 61-62;
`Ex.1003.059-.060, .072-.074, .077-.078 (1410 Lin Decl.).
`
`219
`
`
`
`880 Patent: Disputes
`
`1. A POSA would have combined Thia and Tanenbaum96 (and
`Nahum)
`a. A POSA would have understood that Thia’s teachings are
`applicable to TCP/IP
`b. The trend towards TCP/IP in the 1990s would motivate
`combining Thia’s bypass architecture with TCP/IP
`c. Tanenbaum96 does not teach away from the
`combination
`It would have been obvious to combine Nahum with Thia
`and Tanenbaum96
`
`d.
`
`220
`
`
`
`Tanenbaum96 does not teach away
`from a combination with Thia
`Instead, it describes design
`preferences and tradeoffs
`
`Ex. 1006.588-.599 (Tanenbaum96);
`See also Paper 42 (1409 Reply) at 7-8; Ex. 1223.013-.016 (1409 Lin Reply Decl.);
`Paper 42 (1410 Reply) at 5-6; Ex. 1223.023-.025 (1410 Lin Reply Decl.).
`
`221
`
`
`
`Tanenbaum96 does not discourage
`offloading simple protocols
`
`Ex. 1006.588 (Tanenbaum96);
`See also Paper 42 (1409 Reply) at 7; Ex. 1223.014-.015 (1409 Lin Reply Decl.);
`Paper 42 (1410 Reply) at 5-6; Ex. 1223.024-.025 (1410 Lin Reply Decl.).
`
`222
`
`
`
`Tanenbaum96: Transport processing is
`“straightforward” in the “normal case”
`
`Ex. 1006.583 (Tanenbaum96);
`See also Paper 42 (1409 Reply) at 7; Ex. 1223.014-.015 (1409 Lin Reply Decl.);
`Paper 42 (1410 Reply) at 5-6; Ex. 1223.024-.025 (1410 Lin Reply Decl.).
`.
`
`Ex. 1006.585 (Tanenbaum96);
`See also Paper 42 (1409 Reply) at 7; Ex. 1223.014-.015 (1409 Lin Reply Decl.);
`Paper 42 (1410 Reply) at 5-6; Ex. 1223.024-.025 (1410 Lin Reply Decl.).
`
`223
`
`
`
`Thia also recognizes the difficulty of
`offloading a complex protocol stack
`
`Ex. 1015.002 (Thia);
`See also Paper 42 (1409 Reply) at 8;
`Ex. 1223.015-.016 (1409 Lin Reply Decl.); Paper 42 (1410 Reply) at 6;
`Ex. 1223.025 (1410 Lin Reply Decl.).
`
`224
`
`224
`
`
`
`Thia’s solution: “Fast path” offload
`based on header prediction
`
`Ex. 1015.002 (Thia);
`See also Paper 42 (1409 Reply) at 8; Ex. 1223.015-.016 (1409 Lin Reply Decl.);
`Paper 42 (1410 Reply) at 6; Ex. 1223.025 (1410 Lin Reply Decl.).
`
`225
`
`
`
`Both disclose a bypass/fast-path based
`on TCP/IP header prediction
`
`Ex. 1015 (Thia) at .002;
`See also Paper 1 (1409 Petition) at 32-31; Paper 1 (1410 Petition) at 33-34, 36-37.
`
`Ex. 1006.585 (Tanenbaum96);
`See also Paper 1 (1409 Petition) at 32-31; Paper 1 (1410 Petition) at 33-34, 36-37.
`
`226
`
`
`
`880 Patent: Disputes
`
`1. A POSA would have combined Thia and Tanenbaum96 (and
`Nahum)
`a. A POSA would have understood that Thia’s teachings are
`applicable to TCP/IP
`b. The trend towards TCP/IP in the 1990s would motivate
`combining Thia’s bypass architecture with TCP/IP
`c. Tanenbaum96 does not teach away from the combination
`d.
`It would have been obvious to combine Nahum with
`Thia and Tanenbaum96
`
`227
`
`
`
`PO makes no additional arguments
`regarding a combination with Nahum
`
`Paper 32 (1410 Corrected Response) at 57-58 ;
`See also Paper 42 (1410 Reply) at 5.
`
`228
`
`
`
`880 Patent: Disputes
`
`1. A POSA would have combined Thia and Tanenbaum96
`(and Nahum)
`
`2. Thia and Nahum are enabling
`
`3. The prior art combinations disclose the limitations of the
`challenged claims of the 880 Patent
`
`4. Motions to Amend 880 Patent should be denied
`
`229
`
`
`
`PO fails to identify why Thia and Nahum
`are allegedly not enabling
`
`• Patent Owner contends that Thia is an “inoperable device” and is
`therefore a non-enabling reference
`Paper 32 (1409 Corrected Response) at 20-21; Paper 32 (1410 Corrected Response) at 16.
`• Patent Owner’s only support that Thia and Nahum are not enabling is
`Dr. Almeroth’s conclusory declaration
`
`See Paper 42 (1409 Reply) at 3-4; Paper 42 (1410 Reply) at 2-3.
`• But a non-enabling reference can be prior art “for all that it teaches”
`Id. (citing Beckman Instruments v. LKB Produkter AB, 892 F.2d 1547, 1551 (Fed. Cir. 1989)).
`
`230
`
`
`
`Thia is not a theoretical device
`
`Ex. 1015.008 (Thia).
`
`Ex. 1223.004-.005 (1409 Lin Reply Decl.) at ¶ 8; see also Ex. 1223.004-.005 (1410 Lin Reply Decl.) at ¶ 8;
`See also Paper 42 (1409 Reply) at 3-4; Paper 42 (1410 Reply) at 2-3.
`
`231
`
`
`
`Nahum Is enabling
`
`Ex. 1079.002 (Nahum);
`See also Paper 42 (1410 Reply) at 5.
`
`232
`
`
`
`880 Patent: Disputes
`
`1. A POSA would have combined Thia and Tanenbaum96
`(and Nahum)
`
`2. Thia and Nahum are enabling
`
`3. The prior art combinations disclose the limitations of
`the challenged claims of the 880 Patent
`
`4. Motions to Amend 880 Patents should be denied
`
`233
`
`
`
`880 Patent: Disputes
`
`3. The prior art combinations disclose the limitations of the
`challenged claims of the 880 Patent
`a. The prior art combination renders obvious “an
`operation code” (claims 1, 17, 32, 34, 45)
`b. Thia discloses a “re-assembler” on, or “re-assembly” by, a
`network interface (claims 32, 41, 43)
`c. A “flow key” that includes a “first hop medium access control
`(MAC) layer address” would have been obvious (claim 32)
`d. The prior art combination discloses storing the “header
`portion in a header buffer” if the “header conforms to the
`TCP protocol” (claim 32)
`e. The prior art combination discloses a “processor” for TCP
`processing (claims 1, 32, 41, 43)
`
`234
`
`
`
`880 Patent: Claims 1, 17, 32, 34, and 45
`
`Ex. 1001 (880 Patent), Claims 1, 17, 32, 34, 45.
`
`235
`
`
`
`PO told the patent office that a single bit
`can be an operation code
`
`Ex. 1002 (880 Patent File History) at .249;
`See also Paper 1 (1409 Petition) at 49 n.11;
`Paper 1 (1410 Petition) at 53 n.10.
`
`The term “operation code” does
`not appear in the 880 Patent
`outside the claims
`
`236
`
`
`
`Result of the receive bypass test
`indicates if the packet is bypassable
`
`Ex. 1003.097 (1409 Lin Decl.);
`See also Paper 1 (1409 Petition) at 48; Paper 1 (1410 Petition) at 53; Ex. 1003.102 (1410 Lin Decl.).
`
`237
`
`
`
`A POSA would know that the receive
`bypass test results in an op code
`
`Ex. 1003.098-.099 (1409 Lin Decl.);
`See also Paper 1 (1409 Petition) at 49-50;
`Paper 1 (1410 Petition) at 53-54; Ex. 1003.103-.104 (1410 Lin Decl.).
`
`238
`
`
`
`Thia’s operation code: Flag used by the
`“no-in-transit PDU” test
`
` Thia’s flag indicates the status of the most
`recently-received packet – i.e. whether it
`will be processed on the bypass fast-path
`
`Ex.1015.004 (Thia); See also Paper 1 (1409 Petition) at 49-50; Ex. 1003.098-.099 (1409 Lin Decl.); Paper 42 (1409 Reply) at
`15-16; Paper 1 (1410 Petition) at 53-54; Ex. 1003.103-.104 (1410 Lin Decl.); Paper 42 (1410 Reply) at 15.
`
`239
`
`
`
`880 Patent: Disputes
`
`3. The prior art combinations disclose the limitations of the
`challenged claims of the 880 Patent
`a. The prior art combination renders obvious “an operation code” (claims 1,
`17, 32, 34, 45)
`b. Thia discloses a “re-assembler” on, or “re-assembly” by, a network
`interface (claims 32, 41, 43)
`c. A “flow key” that includes a “first hop medium access control (MAC) layer
`address” would have been obvious (claim 32)
`d. The prior art combination discloses storing the “header portion in a header
`buffer” if the “header conforms to the TCP protocol” (claim 32)
`e. The prior art combination discloses a “processor” for TCP processing
`(claims 1, 32, 41, 43)
`
`240
`
`
`
`880 Patent: Claim 32
`
`Ex. 1001 (880 Patent), Claim 32.
`
`241
`
`
`
`880 Patent: Claims 41 and 43
`
`Ex. 1001 (880 Patent), Claims 41, 43.
`
`242
`
`
`
`Unrebutted evidence that TCP
`reassembles segments into streams
`
`Ex. 1006.540-.541 (Tanenbaum96);
`See also Paper 1 (1410 Petition) at 57-58;
`Ex. 1003.109-.110 (1410 Lin Decl.).
`
`243
`
`
`
`Unrebutted evidence of re-assembler /
`re-assembly in Thia
`
`Ex. 1015.005 (Thia); See also Paper 1 (1410 Petition) at 77;
`Ex. 1003.138-.139 (1410 Lin Decl.).
`
`244
`
`
`
`“Segmentation/reassembly” refers to
`lower-layer fragmentation
`
`Ex. 1015.014 (Thia).
`
`Thia’s disclosure is discussing fragmentation
`and re-assembling those fragments at lower-
`layer protocols.
`See Ex. 1223.017-.020 (1410 Lin Reply) at ¶¶ 24-28; see also Paper 42 (1410
`Reply) at 17-18.
`
`245
`
`
`
`Network layer: “Segmentation” is
`“fragmentation”
`
`Ex. 1006.426 (Tanenbaum96);
`See also Ex. 1223.020 (1410 Lin Reply Decl.).
`
`246
`
`
`
`Thia teaches lower-layer
`segmentation/reassembly on the NIA
`
`No bypass –
`Host processes
`the packet
`
`Bypass –
`ROPE
`processes the
`packet and
`stores the data
`in host
`memory
`
`A packet is
`received at the
`NIA via
`Transmission
`Medium
`
`Ex. 1223.019-.020 (1410 Lin Reply) at ¶ 28 (excerpting and annotating Figure 2 from Ex. 1015.007 (Thia) with red, green,
`blue annotations and red shading); see also Paper 42 (1410 Reply) at 17-18.
`
`247
`
`
`
`Explanation of network layer (IP)
`fragmentation
`
`Ex. 1006.431 (Tanenbaum96); See also Paper 42 (1410
`Reply) at 17-18; Ex.1223.017-.018 (1410 Lin Reply
`Decl.).
`
`248
`
`
`
`Disclosure for transmitting a packet fails
`to rebut disclosure of re-assembly
`
`Ex. 1015.009 (Thia).
`
`See, e.g., Paper 42 (1410 Reply) at 18; Ex.1223.020-.022 (1410 Lin Reply Decl.) at ¶¶ 29-30.
`
`249
`
`Ex. 1015.007 (Thia) (Fig. 2).
`
`
`
`880 Patent: Disputes
`
`3. The prior art combinations disclose the limitations of the
`challenged claims of the 880 Patent
`a. The prior art combination renders obvious “an operation code” (claims 1,
`17, 32, 34, 45)
`b. Thia discloses a “re-assembler” on, or “re-assembly” by, a network
`interface (claims 32, 41, 43)
`c. A “flow key” that includes a “first hop medium access control (MAC)
`layer address” would have been obvious (claim 32)
`d. The prior art combination discloses storing the “header portion in a header
`buffer” if the “header conforms to the TCP protocol” (claim 32)
`e. The prior art combination discloses a “processor” for TCP processing
`(claims 1, 32, 41, 43)
`
`250
`
`
`
`880 Patent: Claim 32
`
`Ex. 1001 (880 Patent), Claim 32.
`
`251
`
`
`
`Tanenbaum96 discloses flow key
`comprising the TCP/IP socket pair
`
`Ex. 1006.585 (Tanenbaum96);
`See Paper 1 (1410 Petition) at 47-48;
`Ex. 1003.093-.095 (1410 Lin Decl.);
`See also Petition 1 (1410 Petition) at 30, 48.
`
`252
`
`
`
`It would be obvious to include header
`information relevant to the connection
`
`Ex. 1003.095 (1410 Lin Decl.);
`See also Paper 1 (1410 Petition) at 49.
`
`253
`
`
`
`MAC layer address is relevant to the
`connection
`
`Ex. 1223.014-.015 (1410 Lin Decl.) at ¶ 18;
`See also Paper 1 (1410 Petition) at 49;
`Ex. 1003.095 (1410 Lin Decl.); Paper 42 (1410 Reply) at 10-12.
`
`254
`
`
`
`The pool of fields to include in a flow key
`is finite and small
`
`Ex. 1013.125 (Stevens2) at Fig. 4.8;
`See also Paper 42 (1410 Reply) at 10-11;
`Ex. 1223.011-.015 (1410 Lin Reply Decl.).
`
`255
`
`
`
`The pool of fields to include in a flow key
`is finite and small
`
`Ex. 1008.058 (Stevens1) at Fig. 3.1;
`See also Paper 42 (1410 Reply) at 10-11; Ex. 1223.011-.015 (1410 Lin Reply Decl.).
`
`256
`
`
`
`The pool of fields to include in a flow key
`is finite and small
`
`Ex. 1008.249 (Stevens1) at Fig. 17.2;
`See also Paper 42 (1410 Reply) at 10-11; Ex. 1223.011-.015 (1410 Lin Reply Decl.).
`
`257
`
`
`
`Dr. Almeroth opined that a MAC layer
`address is not required to infringe
`
`Ex. 1249.005 (Almeroth Infrgmnt. Rpt.) at ¶ 61; See also
`Paper 42 (1410 Reply) at 13.
`
`258
`
`
`
`880 Patent: Disputes
`
`3. The prior art combinations disclose the limitations of the
`challenged claims of the 880 Patent
`a. The prior art combination renders obvious “an operation code” (claims 1,
`17, 32, 34, 45)
`b. Thia discloses a “re-assembler” on, or “re-assembly” by, a network
`interface (claims 32, 41, 43)
`c. A “flow key” that includes a “first hop medium access control (MAC) layer
`address” would have been obvious (claim 32)
`d. The prior art combination discloses storing the “header portion in a
`header buffer” if the “header conforms to the TCP protocol” (claim
`32)
`e. The prior art combination discloses a “processor” for TCP processing
`(claims 1, 32, 41, 43)
`
`259
`
`
`
`880 Patent: Claim 32
`
`Ex. 1001 (880 Patent), Claim 32.
`
`260
`
`
`
`880 Patent: Storing said header portion
`in a header buffer
`
`Ex. 1003.111 (1410 Lin Decl.); See also Paper 1 (1410 Petition) at 58-60.
`
`Ex. 1015.011 (Thia) at Fig. 4.
`
`261
`
`
`
`Claims do not recite a “separate” header
`buffer
`
`880 Patent, Claim 32;
`See also Paper 42 (1410 Reply) at 16-17.
`
`Paper 32 (1410 Response) at 41.
`
`262
`
`
`
`880 Patent: Disputes
`
`3. The prior art combinations disclose the limitations of the challenged claims
`of the 880 Patent
`a. The prior art combination renders obvious “an operation code” (claims 1,
`17, 32, 34, 45)
`b. Thia discloses a “re-assembler” on, or “re-assembly” by, a network
`interface (claims 32, 41, 43)
`c. A “flow key” that includes a “first hop medium access control (MAC) layer
`address” would have been obvious (claim 32)
`d. The prior art combination discloses storing the “header portion in a header
`buffer” if the “header conforms to the TCP protocol” (claim 32)
`e. The prior art combination discloses a “processor” for TCP
`processing (claims 1, 32, 41, 43)
`
`263
`
`
`
`880 Patent: Claim 1
`
`Ex. 1001 (880 Patent), Claim 1.
`
`264
`
`
`
`880 Patent: Claims 41 and 43
`
`Ex. 1001 (880 Patent), Claims 41, 43.
`
`265
`
`
`
`The prior art combination renders
`obvious TCP processing
`
`Ex.1015 (Thia) at Fig. 2; See also Paper 1; (1409 Petition) 51-57; Paper 1 (1410 Petition) at 60-65.
`
`Paper 1 (1410 Petition) at 34. See also Paper 1 (1409 Petition) 32.
`
`266
`
`
`
`880 Patent: Disputes
`
`1. A POSA would have combined Thia and Tanenbaum96
`(and Nahum)
`
`2. Thia and Nahum are enabling
`
`3. The prior art combinations disclose the limitations of the
`challenged claims of the 880 Patent
`
`4. Motions to Amend 880 Patent should be denied
`
`267
`
`
`
`880 Patent: Motions to Amend
`
`• 2017IPR-01409, IPR2017-01736, IPR2018-00338: Amending
`all challenged claims except for claim 8, which is cancelled
`
`• 2017IPR-01410, IPR2017-01737, IPR2018-00339: Amending
`all challenged claims
`
`268
`
`
`
`880 Patent: Disputes
`
`4. Motions to Amend 880 Patents should be denied
`
`a) PO has not met its burden of production under 35
`U.S.C. § 316(d) due to its failure to provide
`adequate written description support
`
`b) The substitute claims are indefinite
`
`c) The prior art combinations disclose the limitations of the
`substitute claims
`
`269
`
`
`
`PO only provides string citations
`
`Paper 20 (1410 Motion to Amend) at ii.
`
`Paper 20 (1410 Motion to Amend) at vii.
`
`See Paper 38 (1409 Opp. to Motion to Amend) at 2-9; Paper 38 (1410 Opp. to Motion to Amend) at 2-8.
`
`270
`
`
`
`PO’s citations do not identify a “packet
`memory”; just general purpose RAM
`
`Ex. 2025.092 (880 App. Pub.) at ¶ [0838];
`See Paper 50 (1409 Sur-Reply to Motion to Amend) at 2-3;
`Paper 50 (1410 Sur-Reply to Motion to Amend) at 2-3.
`
`Ex. 2025.037 (880 App. Pub.) at Fig. 50.
`
`271
`
`
`
`PO’s citations do not identify a header
`buffer separate from packet memory
`
`Paper 20 (1410 Motion to Amend), App’x A at ii.
`
`Ex. 2025.059 (880 App. Pub.) at ¶ [0271].
`
`Ex. 2025.049 (880 App. Pub.) at ¶ [0115].
`
`Ex. 2025.002 (880 App. Pub.) at Fig. 2.
`See Paper 38 (1409 Opp. to Motion to Amend) at 5-6; Paper 50 (1409 Sur-Reply to Motion to Amend) at 4-5; Paper 38 (1410 Opp. to
`Motion to Amend) at 4-5; Paper 50 (1410 Sur-Reply to Motion to Amend) at 4-5.
`
`Ex. 2025 (880 App. Pub.) at cl. 33.
`
`272
`
`
`
`PO in its reply relies entirely on new
`evidence for support
`
`Paper 20 (1410 Motion to Amend), App’x A at ii.
`
`Paper 43 (1410 Motion to Amend
`Reply) at 1-2;
`See Paper 50 (1409 Sur-Reply to
`Motion to Amend) at 4-5;
`Paper 50 (1410 Sur-Reply to Motion
`to Amend) at 4-5.
`
`273
`
`
`
`Amended limitations are not identical to
`the original, as-filed claims
`
`•
`
`•
`
`•
`
`•
`
`“wherein the header buffer is separate from the packet
`memory” (claim 61)
`“wherein the header buffer is separate from the packet
`memory” (claim 79)
`“wherein the header buffer is separate from said packet
`memory” (claim 85)
`“wherein the header buffer is separate from the memory”
`(claim 87)
`
`274
`
`
`
`880 Patent: Disputes
`
`4. Motions to Amend 880 Patents should be denied
`
`a) PO has not met its burden of production under 35
`U.S.C. § 316(d) due to its failure to provide adequate
`written description support
`
`b) The substitute claims are indefinite
`
`c) The prior art combinations disclose the limitations of the
`substitute claims
`
`275
`
`
`
`A POSA would not know what “separate
`from” means in this context
`
`•
`
`•
`
`•
`
`“… could mean that the header buffer and packet memory are
`located on the same memory device, but the physical
`location on the memory device where the header is stored is
`different from the physical location on the memory device
`whether the packets are stored”
`“… could refer to the memory device itself, such that the
`header buffer is on a different memory device than the packet
`memory”
`“… could mean that the virtual address for the header is
`separate from the virtual address for the packet”
`
`Ex. 1210.010 (1409 Lin Opp. Decl.) at ¶ 24; See also Paper 38 (1409 Opp. to Mot. to Amend) at 9-11; Paper 50 (1409 Sur-Reply to Mot. to
`Amend) at 7-8; Paper 38 (1410 Opp. to Mot. to Amend) at 8-10, Ex. 1210.010-.011 (1410 Lin Opp. Decl.) at ¶ 24, Paper 50 (1410 Sur-Reply to
`Mot. to Amend) at 7-8.
`
`276
`
`
`
`880 Patent: Disputes
`
`4. Motions to Amend 880 Patents should be denied
`
`c) The prior art combinations disclose the limitations of the
`substitute claims
`
`i.
`
`ii.
`
`“storing said header portion in a header buffer, wherein
`the header buffer is separate from the packet memory”
`(substitute claims 61, 79, 85, 87)
`
`“re-assembling [said/a/the] data portion” / “re-assembler”
`(substitute claims 61, 79, 85, 87)
`
`277
`
`
`
`Exemplary proposed claim 61
`
`Paper 20 (1409 Motion to Amend), App’x C at xix.
`
`278
`
`
`
`The “slower external memory” is a
`“packet memory” as claimed
`
`Ex. 1015.011 (Thia);
`See Paper 38 (1409 Opp. to Motion to Amend) at 13-14, 20-22;
`Ex. 1210.025, .053 (1409 Lin Opp. Decl.); Paper 50 (1409 Sur-Reply to Motion to
`Amend) at 9-10; Paper 38 (1410 Opp. to Motion to Amend) at 12, 16-17; Ex. 1210.045,
`.055 (1410 Lin Opp. Decl.); Paper 50 (1410 Sur-Reply to Motion to Amend) at 9-10.
`
`279
`
`
`
`“header buffer” in internal memory is
`separate from external “packet memory”
`Thia’s Internal Dual-Ported Memory
`Thia’s External
`Comprising Header Buffers
`Packet Memory
`Header
`portion
`
`Move for
`bypass
`processing
`
`Payload
`
`Header
`portion
`
`Payload
`
`Ex. 1210.052-.053 (1409 Lin Opp. Decl.) (excerpting Ex. 1015 (Thia) at .011);
`See also Paper 38 (1409 Opp. to Motion to Amend) at 22-23; Paper 50 (1409 Sur-Reply to Motion to Amend) at 10-11; Paper 38 (1410 Opp. to
`Motion to Amend) at 17; Ex. 1210.054-.055 (1410 Lin Opp. Decl.); Paper 50 (1410 Sur-Reply to Motion to Amend) at 10-11.
`
`280
`
`
`
`PO’s (and Its expert’s) rebuttal of
`petitioner’s obviousness argument is
`based on a incorrect premise
`
`Paper 43 (1410 Motion to Amend Reply) at 8;
`See Paper 50 (1409 Sur-Reply to Motion to Amend) at 9-11; Paper 50 (1410
`Sur-Reply to Motion to Amend) at 9-11.
`
`281
`
`
`
`A “window” refers to the number of
`bytes, not the size of packets, that can
`be received
`
`Ex. 1006.545 (Tanenbaum96);
`See Paper 50 (1409 Sur-reply to Motion to Amend) at 9-11;
`Ex. 1210.056-.057 (1409 Lin Opp. Decl.);
`Paper 50 (1410 Sur-reply to Motion to Amend) at 9-11;
`Ex. 1210.058-.059 (1410 Lin Opp. Decl.).
`
`282
`
`
`
`PO’s expert agrees that a window does
`not refer to size of packets
`
`Q. Does – is the window referencing a TCP window?
`A. No.
`Q. What is it referencing?
`A. Within the GO-back-N retransmission strategy,
`there is a window size. And so it’s referencing that
`window size. And in that instance, it’s referencing
`how much data can be buffered on the receive
`side….
`
`Ex. 1254 (Almeroth Depo.) at 100:15-22;
`See Paper 50 (1409 Sur-reply to Motion to Amend) at 9-10;
`Paper 50 (1410 Sur-reply to Motion to Amend) at 9-10.
`
`283
`
`
`
`PO’s expert disagrees that Thia’s
`external memory is only for packets that
`are too large for internal memory
`
`Q.
`
`A.
`
`Is it your understanding that slower external
`memory could only be needed if the packets were
`larger than the on-chip buffer?
`I don’t
`think I would agree with the “only”
`characterization. As the first part of the sentence
`says,
`it says, “The on-chip buffer may not be
`sufficient
`to hold the unacknowledged data
`packets for
`retransmission.”
`I
`think there’s a
`variety of scenarios under which that might be the
`case….
`
`Ex. 1254 (Almeroth, Depo.) at 105:24-106:8;
`See Paper 50 (1409 Sur-reply to Motion to Amend) at 9-10;
`Paper 50 (1410 Sur-reply to Motion to Amend) at 9-10.
`
`284
`
`
`
`PO’s rebuttal is based on a faulty
`premise contradicted by its expert
`
`Q.
`
`A.
`
`So if you had a situation where the large window
`encompassed, let’s just say ten packets, for example, if you
`had a case where the large window encompassed ten
`packets, and together, those ten packets were bigger than
`the on-chip buffer,
`that would be another circumstance
`where the slower external memory would be needed, right?
`If your hypothetical asks me to assume that the ten packets
`are larger than what can be stored in the on-chip buffer, then
`I would agree that the slower external memory would be
`needed.
`I
`think that pretty much reads straight
`from the
`sentences we have been looking at on Page 11.
`And just to be clear, that’s not the individual packets are too
`large to store, but together, the ten packets are too large to
`store, right?
`[objection omitted]
`A.
`If that’s part of your hypothetical, then I think that’s fine.
`Ex. 1254 (Almeroth Depo.) at 107:14-108:8;
`See Paper 50 (1409 Sur-reply to Motion to Amend) at 9-10; Paper
`50 (1410 Sur-reply to Motion to Amend) at 9-10.
`
`Q.
`
`285
`
`
`
`880 Patent: Disputes
`
`4. Motions to Amend 880 Patents should be denied
`
`c) The prior art combinations disclose the limitations of the
`substitute claims
`
`i.
`
`“storing said header portion in a header buffer, wherein the
`header buffer is separate from the packet memory” (substitute
`claims 61, 79, 85, 87)
`
`ii. “re-assembling [said/a/the] data portion” / “re-assembler”
`(substitute claims 61, 79, 85, 87)
`
`286
`
`
`
`PO’s arguments on re-assembly/re-
`assembler are not new and are similarly
`wrong
`
`Paper 43 (1410 Motion to Amend Reply) at 10.
`
`Ex. 2305.012 (Almeroth Decl. ISO Reply) at ¶¶ 28-29.
`
`See Paper 50 (1409 Sur-reply to Motion to Amend) at 11-12;
`Paper 50 (1410 Sur-reply to Motion to Amend) at 11-12.
`
`287
`
`
`
`The evidence PO and its expert rely on
`is for transmitting, not receiving
`
`See Paper 50 (1409 Sur-reply to Motion to Amend) at 11-12;
`Paper 50 (1410 Sur-reply to Motion to Amend) at 11-12.
`
`Ex. 1015.009 (Thia).
`
`288
`
`