Riverbed Technology Inc.v.Silver Peak Systems, Inc.

Patent Trial and Appeal Board

Dec 30, 2014

11202697 (P.T.A.B. Dec. 30, 2014)

Trials@uspto.gov Paper 33
571-272-7822 Entered: December 30, 2014

UNITED STATES PATENT AND TRADEMARK OFFICE
____________

BEFORE THE PATENT TRIAL AND APPEAL BOARD
____________

RIVERBED TECHNOLOGY, INC.,
Petitioner,

v.

SILVER PEAK SYSTEMS, INC.,
Patent Owner.
____________

Case IPR2013-00403
Patent 8,370,583 B2

Before DENISE M. POTHIER, JUSTIN T. ARBES, and HYUN J. JUNG,
Administrative Patent Judges.

ARBES, Administrative Patent Judge.

FINAL WRITTEN DECISION
35 U.S.C. § 318(a) and 37 C.F.R. § 42.73
IPR2013-00403
Patent 8,370,583 B2

2
I. BACKGROUND
Petitioner Riverbed Technology, Inc. filed a Petition (Paper 2, “Pet.”)
seeking inter partes review of claims 1–27 of U.S. Patent No. 8,370,583 B2
(“the ’583 patent”) pursuant to 35 U.S.C. § 311. On January 2, 2014, we
instituted an inter partes review of claims 1–27 on two grounds of
unpatentability (Paper 14, “Dec. on Inst.”). Patent Owner Silver Peak
Systems, Inc. did not file a Patent Owner Response, and instead filed a
Motion to Amend (Paper 23, “Mot.”) seeking to cancel claims 1–27 and
substitute claims 28–33 in their place. Petitioner filed an Opposition
(Paper 24, “Opp.”) to the Motion to Amend, and Patent Owner filed a Reply
(Paper 25, “Reply”). A combined oral hearing for the instant proceeding
and related Case IPR2013-00402 was held on September 30, 2014.
A transcript (Paper 32, “Tr.”) of the hearing is included in the record.
We have jurisdiction under 35 U.S.C. § 6(c). This final written
decision is issued pursuant to 35 U.S.C. § 318(a) and 37 C.F.R. § 42.73.
For the reasons that follow, we grant Patent Owner’s Motion to
Amend to the extent it requests to cancel claims 1–27 of the ’583 patent, and
with respect to proposed substitute claims 28 and 30. We determine that
Patent Owner has not met its burden with respect to proposed substitute
claims 29 and 31–33. The Motion to Amend, therefore, is granted-in-part.

A. The ’583 Patent
The ’583 patent
1
relates to a network memory system. Ex. 1001,
col. 1, ll. 8–9. According to the ’583 patent, prior art information systems

1
The ’583 patent issued from U.S. Patent Application No. 11/202,697 (“the
’697 application”), filed on August 12, 2005. U.S. Patent No. 8,312,226 B2
IPR2013-00403
Patent 8,370,583 B2

3
were implemented with either centralized servers or distributed servers. Id.
at col. 1, ll. 11–14. In a centralized server system, shown in Figure 1, data
are stored in central servers and provided to remote computers over a
communication network upon request. Id. at col. 1, ll. 14–63. Such a
system can be managed easily from the central location, but potentially
causes a “bandwidth bottleneck” for multiple computers attempting to
retrieve data over the network. Id. In a distributed server system, shown in
Figure 2, by contrast, remote computers can access data from branch servers
rather than the central servers. Id. at col. 1, l. 64–col. 2, l. 58. Doing so
reduces the amount of data that must travel over the network, but requires
complex procedures to ensure the replication and synchronization of data
between the branch servers and central servers. Id.
The ’583 patent also describes the use of caching to reduce network
traffic, and problems with caching at the time. Id. at col. 2, ll. 59–67. For
example, because web caching is based on Uniform Resource Locator
(URL) or filename, not content of the document, a document may be present
in the cache but not retrieved because it was stored with a different URL or
filename. Id. at col. 3, ll. 14–20. Further, web caching is binary, such that
even a small change to a document will result in the cache not being utilized.
Id. at col. 3, ll. 20–24.

(“the ’226 patent”) issued from U.S. Patent Application No. 11/240,110,
filed on September 29, 2005, which is a continuation of the ’697 application.
The ’226 patent is the subject of related Case IPR2013-00402. Patent
Owner’s Motion to Amend in Case IPR2013-00402 also is granted-in-part in
the Final Written Decision entered concurrently with this Decision in its
corresponding record.
IPR2013-00403
Patent 8,370,583 B2

4
The ’583 patent describes a network memory system that reduces
network traffic by using appliances linked to the servers and remote
computers, where each appliance is capable of determining whether a
portion of requested data is locally accessible to the other appliance. Id.
at col. 5, ll. 12–44. Figure 3 of the ’583 patent is reproduced below.

Figure 3 depicts network memory system 300 comprising branch office 310
with branch appliance 350 and central office 320 with central appliance 380.
Id. at col. 5, ll. 31–44. Branch appliance 350 and central appliance 380
transparently intercept network traffic between computers 340 and central
servers 370. Id. at col. 6, ll. 63–66. Central appliance 380 comprises
hardware and/or software elements configured to (1) receive a request for
data (e.g., email or files) sent by computer 340 to branch appliance 350;
(2) receive the data locally; (3) determine whether a portion of the data is
locally accessible to branch appliance 350; (4) generate an instruction based
on the determination; and (5) transfer the instruction to branch appliance
350. Id. at col. 5, l. 12–col. 6, l. 18; col. 6, ll. 47–59.
IPR2013-00403
Patent 8,370,583 B2

5
If the determination is that the data are locally accessible to branch
appliance 350, central appliance 380 sends a retrieve instruction (retrieve
instruction 640 in Figure 6), and branch appliance 350 retrieves the data
locally and forwards the data to computer 340 (without the data having to be
transferred over the network). Id. at col. 11, ll. 25–43. If the determination
is that the data are not locally accessible, however, central appliance 380
sends the data itself with a store instruction (store instruction 440 in
Figure 4), and branch appliance 350 stores a copy of the data locally for
future use. Id. at col. 10, ll. 41–61. In another embodiment, if the
determination is that only a portion of the data is locally accessible, central
appliance 380 sends the non-locally accessible portion with store and
retrieve instructions (instructions 740 in Figure 7A). Id. at col. 12, ll. 10–40.
According to the ’583 patent, the disclosed network memory “does
not cache the data in the traditional sense.” Id. at col. 7, ll. 56–57. Data can
be retrieved locally “even if the URL or filename for the data is different
because the data may be identified by a pattern for the data itself and not by
the URL or filename.” Id. at col. 7, ll. 57–60. The ’583 patent describes the
use of specific data structures to determine whether a portion of response
data is locally accessible. Id. at col. 8, ll. 51–63.
IPR2013-00403
Patent 8,370,583 B2

6
Figure 5 of the ’583 patent is reproduced below.

Figure 5 depicts an exemplary set of data structures for the disclosed
network memory system. When data are stored, for example, at the central
and branch appliances, the central appliance “calculates hashes at every byte
boundary of [the] data flow.” Id. at col. 8, l. 64–col. 9, l. 1. Each calculated
hash is filtered by fine filter 560 and coarse filter 565 to determine fine and
coarse sync-points (i.e., locations in the data flow that meet the fine and
coarse filter criteria, with the coarse filter criteria being “more restrictive”).
Id. at col. 9, ll. 20–34. The hashes are stored in entries of the fine signature
hash table (SHT) 505 and coarse SHT 525. Id. at col. 9, ll. 46–57. Each
entry in the tables is indexed to the response data stored in flow history
pages (FHPs) 545. Id. at col. 10, ll. 4–40. Page fields 515/535 point to
pages in FHPs 545, and byte offsets 520/540 point to the start of fine and
coarse sync-points in the data, respectively. Id.
Upon receiving a request for data from the branch appliance, the
central appliance retrieves the data locally and determines whether the data
are locally accessible to the branch appliance by calculating hashes for the
IPR2013-00403
Patent 8,370,583 B2

7
data, filtering them through fine and coarse filters 560/565 to determine fine
and coarse sync-points, and comparing the hashes to check fields 510/530 in
SHTs 505/525. Id. at col. 9, l. 58–col. 10, l. 3. If there is no match, the data
are not locally accessible. Id. If any of the hashes matches, however,
“additional checks (such as direct forward comparisons and backward
memory comparisons between the response data 625 and the data 555 in the
FHPs 545) may also be made to determine the size of the matching region.”
Id. at col. 11, ll. 6–24. The additional checks may indicate that only a
portion of the response data is locally accessible, in which case the central
appliance (1) stores the generated hashes for the non-locally accessible
portion (called “deltas”) in SHTs 505/525; (2) stores the deltas in FHPs 545;
and (3) sends the deltas with retrieve and store instructions to the branch
appliance. Id. at col. 11, l. 55–col. 12, l. 40.

B. Status of the Claims
The challenged claims of the ’583 patent can be broken down into
three sets of claims, each with similar limitations. Claims 1–11 recite a
network memory system, claims 12–22 recite a method, and claims 23–27
recite a network memory system. Claim 1, for example, recites:
1. A network memory system comprising:
a source-site appliance comprising a first processor and a
first memory device, and configured to be coupled to a
source-site computer via a source-site local area network; and
a destination-site appliance comprising a second
processor and a second memory device, and configured to be
coupled to a destination-site computer via a destination-site
local area network, the source-site computer in communication
with the destination-site computer via a wide area network;
IPR2013-00403
Patent 8,370,583 B2

8
wherein the source-site appliance is configured to
identify locally accessible data of the destination-site appliance,
to intercept transmitted data sent from the source-site computer
and directed to the destination-site computer, to perform a
determination of whether a portion of the transmitted data
corresponds to the locally accessible data of the destination-site
appliance, to generate an instruction based on the
determination, and to send the instruction to the destination-site
appliance over the wide area network; and
wherein the destination-site appliance is configured to
receive the instruction from the source-site appliance over the
wide area network, to process the instruction to obtain the
transmitted data locally if the transmitted data corresponds to
the locally accessible data of the destination-site appliance, and
to transfer the transmitted data to the destination-site computer.
In its Motion to Amend, Patent Owner proposes six substitute claims,
with two proposed substitute claims, each with similar limitations, for each
set. Mot. 1–6. Proposed substitute claims 28 and 29, corresponding to the
first set of claims directed to a network memory system, recite (with
underlined material indicating language added to the original patent claims
and material in brackets indicating language removed from the original
patent claims):
28. (proposed substitute for claim 2) The network
memory system of claim 1 wherein the source-site appliance is
configured determine whether the portion of the transmitted
data corresponds to the locally accessible data via a process
comprising:
identifying sync points in the transmitted data having
matches in the locally accessible data by (i) determining hash
values corresponding to different byte locations of the
transmitted data, (ii) finely filtering the hash values using a fine
filter to determine a finely filtered set of the hash values
corresponding to fine sync points, and coarsely filtering the
hash values using a coarse filter to determine a coarsely filtered
IPR2013-00403
Patent 8,370,583 B2

9
set of the hash values corresponding to coarse sync points, and
(iii) determining from the finely filtered set of the hash values
and the coarsely filtered set of the hash values, a plurality of
hash values matching hash values of the locally accessible data;
and
performing for byte locations with matching hash values,
a forward and backward memory comparison between the
transmitted data and data representing the locally accessible
data, the forward and backward comparison to identify a size of
a matching region of the transmitted data with the data
representing the locally accessible data, and performing based
on results of the forward and backward memory comparison, a
determination of a locally accessible portion of the transmitted
data corresponding to the matching region that is locally
accessible at the destination-site appliance and a non-locally
accessible portion of the transmitted data that is not locally
accessible at the destination-site appliance; and
wherein the source-site appliance is further configured to
generate a plurality of instructions based on the determination,
wherein the plurality of instructions comprise a retrieve
instruction and a store instruction, the retrieve instruction
indicating to the destination-site appliance to retrieve the locally
accessible portion of the transmitted data from a database
locally accessible to the destination-site appliance, and the store
instruction indicating to the destination–site appliance to store
the non-locally accessible portion of the transmitted data.
29. (proposed substitute for claim 11) The network
memory system of claim [1] 28 wherein the store instruction
further indicates to the destination-site appliance an index
within [a] the database for storing another copy of the locally
accessible portion of the transmitted data together with the
non-locally accessible portion of the data.

C. Prior Art
The pending grounds of unpatentability in the instant inter partes
review are based on the following prior art:
IPR2013-00403
Patent 8,370,583 B2

10
1. U.S. Patent No. 6,667,700 B1, issued December 23,
2003 (Ex. 1004, “McCanne ’700”);
2. U.S. Patent No. 8,069,225 B2, filed August 12, 2003,
issued November 29, 2011 (Ex. 1005, “McCanne ’225”); and
3. U.S. Patent Application Publication No. 2004/0088376
A1, published May 6, 2004 (Ex. 1003, “McCanne ’376”).2
In their papers, the parties also refer to the following references:
4. U.S. Patent Application Publication No. 2005/0091234
A1, filed October 23, 2003, published April 28, 2005
(Ex. 1007, “Hsu”);
5. Silver Peak Systems, Inc., “The Benefits of Byte-Level
WAN Deduplication: WAN Deduplication–Doing More with
Less,” 1–4 (2008) (Ex. 2001, “Silver Peak White Paper”); and
6. Neil T. Spring & David Wetherall,
“A Protocol-Independent Technique for Eliminating Redundant
Network Traffic,” SIGCOMM ’00, 87–95 (2000) (Ex. 2002,
“Spring”).

D. Pending Grounds of Unpatentability
The instant inter partes review involves the following grounds of
unpatentability:
Reference(s) Basis Claims
McCanne ’376 35 U.S.C. § 102(b) 1–27
McCanne ’700 and
McCanne ’225
35 U.S.C. § 103(a) 1–27
Dec. on Inst. 21.

2
We refer to McCanne ’376, McCanne ’700, and McCanne ’225
collectively as the “McCanne references.”
IPR2013-00403
Patent 8,370,583 B2

11
II. ANALYSIS
A. Challenged Claims
As noted above, Patent Owner did not file a Patent Owner Response
to the Petition. In its Motion to Amend, Patent Owner “moves to cancel
Claims 1–27 and to substitute Claims 28–33 in their place.” Mot. 1; see
35 U.S.C. § 316(d)(1); 37 C.F.R. § 42.121(a)(3) (“A motion to amend may
cancel a challenged claim or propose a reasonable number of substitute
claims.”). Patent Owner’s request to cancel claims 1–27 is not contingent on
the claims being determined to be unpatentable. We grant the request and
turn to the proposed substitute claims in the Motion to Amend.

B. Proposed Substitute Claims
As the moving party, Patent Owner bears the burden of proof to
establish that it is entitled to the relief requested. See 37 C.F.R. § 42.20(c).
Entry of proposed amendments is not automatic, but occurs only upon the
patent owner having met the requirements of 37 C.F.R. § 42.121 and
demonstrated, by a preponderance of the evidence, the patentability of the
proposed substitute claims. See Idle Free Systems, Inc. v. Bergstrom, Inc.,
IPR2012-00027, slip op. at 7–8 (PTAB June 11, 2013) (Paper 26, “Idle
Free”) (informative); Toyota Motor Corp. v. American Vehicular Sciences
LLC, IPR2013-00419, slip op. at 4–5 (PTAB Mar. 7, 2014) (Paper 32,
“Toyota”). For the reasons explained below, we conclude that Patent Owner
has met its burden with respect to claims 28 and 30, but not claims 29 and
31–33.
IPR2013-00403
Patent 8,370,583 B2

12
1. No Broadening of Scope
Proposed substitute claims in an inter partes review “may not enlarge
the scope of the claims of the patent.” 35 U.S.C. § 316(d)(3); see 37 C.F.R.
§ 42.121(a)(2)(ii). In its Motion to Amend, Patent Owner proposes claim 28
as a substitute for claim 2, claim 29 as a substitute for claim 11, claim 30 as
a substitute for claim 13, claim 31 as a substitute for claim 22, claim 32 as a
substitute for claim 24, and claim 33 as a substitute for claim 27. Mot. 1–6.
Each claim includes all of the limitations of the corresponding claim for
which it is a substitute, and adds additional limitations. No limitations are
removed. Petitioner in its Opposition does not dispute Patent Owner’s
assertion that the proposed substitute claims do not enlarge the scope of the
claims of the ’583 patent. See Mot. 8. We are persuaded that the proposed
substitute claims do not enlarge the scope of the original patent claims.

2. Written Description Support
Pursuant to 37 C.F.R. § 42.121(b), a motion to amend in an inter
partes review must set forth “[t]he support in the original disclosure of the
patent for each claim that is added or amended,” and “[t]he support in an
earlier-filed disclosure for each claim for which benefit of the filing date of
the earlier filed disclosure is sought.”
In its Motion to Amend, Patent Owner explains how the subject
matter of proposed substitute claims 28–31
3
have written description support
in the Specification of the ’697 application, which issued as the ’583 patent,

3
We address proposed substitute claims 32 and 33 separately below. See
infra Section II.B.6.
IPR2013-00403
Patent 8,370,583 B2

13
as filed. Mot. 7–8 & n.2; see Ex. 1009, 403–60.
4
Regarding the added
limitations in proposed substitute claims 28 and 30, Patent Owner relies on
Figures 5 and 7A, and paragraphs 52–59, 63, and 66–69, of the ’697
application, describing the recited filtering, forward and backward memory
comparison, and generation of both store and receive instructions. Mot. 7–8.
Regarding the added limitations in proposed substitute claims 29 and 31,
Patent Owner relies on Figure 7A and paragraphs 69 and 70. Id. Patent
Owner also cites support in the ’697 application, as filed, for the other
limitations of proposed substitute claims 28–31, as shown in the figures of
the application. Id. Petitioner in its Opposition does not argue that the
claims lack sufficient written description support.
Upon review of Patent Owner’s arguments and the disclosure of the
’697 application, we conclude that Patent Owner has made a sufficient
showing that proposed substitute claims 28–31, as a whole, have written
description support in the disclosure of the ’697 application as filed.

3. Claim Interpretation
The patent owner bears the burden in a motion to amend to show a
patentable distinction of each proposed substitute claim over the prior art.
See 37 C.F.R. § 42.20(c). Accordingly, a “patent owner should identify
specifically the feature or features added to each substitute claim, as
compared to the challenged claim it replaces, and come forward with

4
Petitioner filed a copy of the prosecution history of the ’583 patent as three
documents each numbered Exhibit 1002. To ensure a clear record, we
renumber the document labeled “Part 2 of 3” as Exhibit 1008, and renumber
the document labeled “Part 3 of 3” as Exhibit 1009. See 37 C.F.R.
§ 42.63(c) (“exhibits must be uniquely numbered sequentially”).
IPR2013-00403
Patent 8,370,583 B2

14
technical facts and reasoning about those feature(s).” Idle Free at 7. This
includes “construction of new claim terms, sufficient to persuade the Board
that the proposed substitute claim is patentable over the prior art of record,
and over prior art not of record but known to the patent owner.” Id.; Toyota
at 5. Further, consistent with the statute and legislative history of the
Leahy-Smith America Invents Act, Pub. L. No. 112-29, 125 Stat. 284 (2011)
(“AIA”), the Board interprets claims using the “broadest reasonable
construction in light of the specification of the patent in which [they]
appear[].” 37 C.F.R. § 42.100(b); see also Office Patent Trial Practice
Guide, 77 Fed. Reg. 48,756, 48,766 (Aug. 14, 2012).
In its Motion to Amend, Patent Owner provides a proposed
interpretation for three terms in each of the proposed substitute claims:
Term Proposed Interpretation
“sync point” “a location in a data flow having data
meeting a specified criteri[on]”
“fine sync point” “a location in a data flow having data
meeting a specified fine criteri[on] that is
less restrictive than a coarse criteri[on]”
“coarse sync point” “a location in a data flow having data
meeting a specified coarse criteri[on]
more restrictive than a fine criteri[on]”
Mot. 6–7 (citing Ex. 1001, col. 9, ll. 20–57). Petitioner in its Opposition
does not argue that Patent Owner’s proposed interpretations are incorrect.
We agree that Patent Owner’s proposed interpretations represent the
broadest reasonable interpretation of the claims in light of the Specification
of the ’583 patent. The Specification discloses fine filter 560 and coarse
filter 565, where “[t]he appliance designates the locations in the data flow
which meet the fine and coarse filter criteria as fine and coarse sync-points,
IPR2013-00403
Patent 8,370,583 B2

15
respectively,” and the filter criteria for coarse filter 565 are “more
restrictive” than the filter criteria for fine filter 560. Ex. 1001, col. 9,
ll. 20–38. The Specification also describes an example where “the filter
criteri[on] declares a fine sync-point when the top five bits of the hashes are
all zeros and a coarse filter criteri[on] which stores or compares hashes when
the top ten bits of the hashes are all zeros” (i.e., the filter criterion for the
coarse filter is more restrictive). Id. at col. 9, ll. 30–34. Accordingly, we
interpret “sync point,” “fine sync point,” and “coarse sync point” as set forth
above.
Also, in the Decision on Institution, we interpreted various claim
terms of the independent claims of the ’583 patent as follows:
Term Interpretation
“network memory” (claims 1, 12,
and 23)
device(s) in a network for
storing information
“appliance” (claims 1, 12, and 23) hardware and/or software
elements applied to a particular
use
“instruction” (claims 1, 12, and
23)
a message or signal that
indicates, explicitly or
implicitly, an action to perform
See Dec. on Inst. 6–10. Patent Owner’s proposed substitute claims depend
from the original independent claims of the ’583 patent, and the parties do
not dispute these interpretations in their papers. We do not perceive any
reason or evidence that now compels any deviation from these
interpretations. Accordingly, for purposes of assessing the proposed
substitute claims, we incorporate our previous analysis of the three terms
above. See id.
IPR2013-00403
Patent 8,370,583 B2

16
We also interpreted various means-plus-function limitations in
independent claim 23 in the Decision on Institution. See id. at 10–14.
We address those limitations in the discussion below regarding proposed
substitute claims 32 and 33. See infra Section II.B.6.

4. Claims 28 and 30
Having determined that Patent Owner’s proposed substitute claims do
not enlarge the scope of the claims of the ’583 patent, determined that
proposed substitute claims 28–31 have sufficient written description support,
and interpreted the language of the claims, we turn to each set of claims
specifically to determine if Patent Owner has met its burden of proof,
beginning with proposed substitute claims 28 and 30.
In a motion to amend, the patent owner bears the burden of proof to
demonstrate patentability of its proposed substitute claims over the prior art,
and, thus, entitlement to the claims. Idle Free at 7. This does not mean that
the patent owner is assumed to be aware of every item of prior art known to
a person of ordinary skill in the art. However, the patent owner should
explain in its motion why the proposed substitute claims are patentable over
not just the prior art of record, but also prior art not of record but known to
the patent owner:
A patent owner should identify specifically the feature or
features added to each substitute claim, as compared to the
challenged claim it replaces, and come forward with technical
facts and reasoning about those feature(s), including
construction of new claim terms, sufficient to persuade the
Board that the proposed substitute claim is patentable over the
prior art of record, and over prior art not of record but known to
the patent owner. The burden is not on the petitioner to show
unpatentability, but on the patent owner to show patentable
IPR2013-00403
Patent 8,370,583 B2

17
distinction over the prior art of record and also prior art known
to the patent owner. Some representation should be made about
the specific technical disclosure of the closest prior art known
to the patent owner, and not just a conclusory remark that no
prior art known to the patent owner renders obvious the
proposed substitute claims.
Id. This includes addressing the basic knowledge and skill set possessed by
a person of ordinary skill in the art even without reliance on any particular
item of prior art. Id. at 7–8; Toyota at 4–5. The petitioner then has the
opportunity, in its opposition, to argue any deficiency in the patent owner’s
motion and “come forward with specific evidence and reasoning, including
citation and submission of any applicable prior art,” to rebut the patent
owner’s position on patentability. Idle Free at 8.
Proposed substitute claims 28 and 30 add a number of limitations to
the corresponding claims for which they are substitutes, which we refer to
for convenience as follows:
(1) “fine and coarse filtering of hash values” limitations:
“identifying sync points in the transmitted data having matches
in the locally accessible data by (i) determining hash values
corresponding to different byte locations of the transmitted
data, (ii) finely filtering the hash values using a fine filter to
determine a finely filtered set of the hash values corresponding
to fine sync points, and coarsely filtering the hash values using
a coarse filter to determine a coarsely filtered set of the hash
values corresponding to coarse sync points, and (iii)
determining from the finely filtered set of the hash values and
the coarsely filtered set of the hash values, a plurality of hash
values matching hash values of the locally accessible data”; and
(2) “forward and backward memory comparison”
limitations: “performing for byte locations with matching hash
values, a forward and backward memory comparison between
the transmitted data and data representing the locally accessible
data, the forward and backward comparison to identify a size of
IPR2013-00403
Patent 8,370,583 B2

18
a matching region of the transmitted data with the data
representing the locally accessible data, and performing based
on results of the forward and backward memory comparison, a
determination of a locally accessible portion of the transmitted
data corresponding to the matching region that is locally
accessible at the destination-site appliance and a non-locally
accessible portion of the transmitted data that is not locally
accessible at the destination-site appliance . . . wherein the
plurality of instructions comprise a retrieve instruction and a
store instruction, the retrieve instruction indicating to the
destination-site appliance to retrieve the locally accessible
portion of the transmitted data from a database locally
accessible to the destination-site appliance, and the store
instruction indicating to the destination-site appliance to store
the non-locally accessible portion of the transmitted data.”
See Mot. 1–4.
Patent Owner argues that pre-existing data de-duplication systems
reduced network traffic by dividing data into “segments” (or “chunks”) and
transmitting a “reference” (or “token”), rather than the data segment itself,
when the destination device already has a copy of the data segment. Id. at 9.
Patent Owner cites the Silver Peak White Paper’s
5
description of such
systems, as well as the McCanne references, as examples of that approach.
Id. (citing Exs. 1003–1005, 2001). The McCanne references, for example,
describe segmenting input data by “identifying ‘cut points,’ such as offsets
in the input data where one segment ends and the next segment begins.”
Ex. 1004, col. 6, ll. 59–64. The segmentation is based on the content of the
data, such that “segment boundaries should always appear in the same place
for the same sequences of data, regardless of the context in which that data
appeared,” and both the source and destination devices determine segments

5
The Silver Peak White Paper was provided as an exhibit with Patent
Owner’s Motion to Amend. Ex. 2001.
IPR2013-00403
Patent 8,370,583 B2

19
the same way. Id. at col. 6, l. 65–col. 7, l. 26; Ex. 1003 ¶ 47; Ex. 1005,
col. 12, ll. 27–39. After the data are segmented, the source device performs
a segment-by-segment lookup against all data segments in memory.
Ex. 1004, col. 8, ll. 53–55. If a match is found (indicating that the
destination device has a copy of a data segment), the source device replaces
the data segment with a “reference” and sends just the reference instead,
causing the destination device to retrieve the data locally and reducing
network traffic between the devices. Id. at col. 9, ll. 20–32; Ex. 1003
¶¶ 69–70, 75.
Patent Owner contends that the proposed substitute claims describe a
very different approach. Mot. 10–14. Instead of relying on inflexible
“pre-segmented data” and inefficiently searching through all data segments,
as in the McCanne references, the proposed substitute claims recite a
specific method of (1) identifying sync points in the data by determining
hash values for different byte locations in the data, and finely and coarsely
filtering the hash values to locate matches; and (2) performing a forward and
backward memory comparison to identify a matching region of data of
optimum length. Id. at 10–11. Patent Owner explains:
Data deduplication systems like McCanne work well when
changes to the data are isolated to small contiguous regions of
the file, but perform poorly when changes are dispersed
throughout the file or when the changes themselves are small
(referred to as “dynamic data”). Because these prior systems
use pre-segmented data, they cannot operate at byte-level
granularity and cannot detect partial matches between
segments. Even if there is only a very small difference between
two segments (e.g., a 1% difference), these systems may be
unable to identify the matching portion and treat the whole
segment as non-matching. As a result, prior art systems have
limited efficiency when dynamic data is present.
IPR2013-00403
Patent 8,370,583 B2

20
Id. at 10 (citation omitted). Patent Owner’s explanation as to why the
proposed substitute claims are different from pre-existing data
de-duplication systems is persuasive and is supported by the evidence of
record. See Ex. 2001, 2–3 (describing problems with handling dynamic data
in token-based systems); Ex. 1004, col. 7, ll. 36–41 (recognizing that the
pre-segmented data approach “does not produce the optimal segmentation
(i.e., maximizing the size of segments while simultaneously maximizing the
number of repeated segments found)”); Ex. 1001, col. 9, ll. 40–44 (stating
that using fine and coarse hash tables helps to “optimize” the tradeoff
between maximizing the number of generated sync points and limiting the
size of the tables).
Patent Owner also accounts for the knowledge of a person of ordinary
skill in the art in two respects.

“Fine and Coarse Filtering of Hash Values” Limitations
Patent Owner states that “hash values and filters” were known in
“other contexts,” but the specific method recited in the proposed substitute
claims of determining hash values at byte locations of data, and identifying
sync points by finely and coarsely filtering the hash values, was not known.
Mot. 14. According to Patent Owner, the invention of the proposed
substitute claims allows for “rapid search, while preserving the ability to
match at the byte level, which none of the prior art of which the Patent
Owner is aware could provide at the time the invention was made either
alone or in combination.” Id.
Petitioner responds that Hsu, a reference cited during prosecution of
the ’583 patent, discloses the “fine and coarse filtering of hash values”
IPR2013-00403
Patent 8,370,583 B2

21
limitations recited in the proposed substitute claims. Opp. 4–7 (citing
Ex. 1007 ¶¶ 16, 60–63, Fig. 6). Hsu describes a “data chunking system” that
divides data into “fixed-sized chunks” so that duplicate data may be
identified and only unique data chunks are stored or transmitted. Ex. 1007
¶¶ 13–14, Abstract. The system “chunks data by selecting a chunk of fixed
size, then moving or ‘sliding’ a window of a fixed size, across the data
stream until a match to existing data is found.” Id. ¶ 16. Hsu describes one
way of determining whether data in the window matches a previously
remembered chunk (and is not a “false positive”). Id. ¶¶ 53, 60.
Figure 6 of Hsu is reproduced below.

Figure 6 depicts an “approximation test” at step 605 and a “more accurate
test” at step 620. The approximation test determines whether the data in the
IPR2013-00403
Patent 8,370,583 B2

22
window has “probably been seen before,” and may be “performed by
computing a rolling checksum of the data and looking up the computed
value in a hash table.” Id. ¶¶ 42, 60–61. The more accurate test may
involve “computing a cryptographic hash of the data in the window and
looking up the computed value in a hash table.” Id. ¶¶ 42, 62.
Petitioner equates the approximation test with the fine filtering recited
in the proposed substitute claims, and the more accurate test with the recited
coarse filtering. Opp. 5–7. As Patent Owner points out, however, Hsu is a
token-based system, like the McCanne references, that operates on
“fixed-size chunks.” Reply 2–3. Hsu calculates a hash value for the data in
the window and compares it with every value in a hash table to locate
potential matches. Ex. 1007 ¶¶ 42, 60–62. It does not identify sync points
in the data by “determining hash values corresponding to different byte
locations of the transmitted data,” “finely filtering the hash values . . . to
determine a finely filtered set of the hash values corresponding to fine sync
points,” “coarsely filtering the hash values . . . to determine a coarsely
filtered set of the hash values corresponding to coarse sync points,” and
“determining from the finely filtered set of the hash values and the coarsely
filtered set of the hash values, a plurality of hash values matching hash
values of the locally accessible data,” as recited in the proposed substitute
claims. Thus, although Hsu may be considered as teaching the general
concept of matching at different levels of granularity, we are not persuaded
that it teaches or suggests the specific “fine and coarse filtering of hash
values” limitations recited in the proposed substitute claims.

IPR2013-00403
Patent 8,370,583 B2

23
“Forward and Backward Memory Comparison” Limitations
Patent Owner acknowledges that “the general concept of performing
direct memory comparisons of data to determine a matching region . . . was
known in the context[] of network-layer caches,” citing Spring
6
as the
closest prior art on the subject. Mot. 11–12. Spring describes a method of
“identify[ing] repeated byte ranges between packets to avoid
retransmitting . . . redundant data” in a shared packet cache arrangement.
Ex. 2002, 1, 3. A cache holds the “most recent packets.” Id. at 2. For every
packet of an input stream, the system generates a set of “fingerprints”
(unique “integers generated by a one-way function applied to a set of bytes”)
and checks each fingerprint against the indexed fingerprints of the cache. Id.
If there is a match, the input packet likely has the same content as the packet
in the cache at the regions corresponding to the fingerprint. Id. The system
then compares the input and cached packets to verify the match, and “[t]hen
the matching region is expanded, both to the left and to the right,
byte-by-byte in each packet, to find the largest matching region.” Id. Once
a matching region is determined, the system can transmit tokens that allow
the data to be retrieved from the shared packet cache, rather than
transmitting the data itself. Id. at 1, 3.
Patent Owner argues that it would not have been obvious to
incorporate the technique of Spring into a network data de-duplication
system, as recited in the proposed substitute claims, because Spring “had
very limited practical application in the network memory context.” Mot. 12.
Patent Owner’s position is supported by the statement in McCanne ’376 that
due to Spring’s small cache and first-in first-out replacement policy, “the

6
Spring was cited during prosecution of the ’583 patent. Ex. 1001, 2.
IPR2013-00403
Patent 8,370,583 B2

24
efficacy of [Spring’s] approach is limited to detecting and exploiting
communication redundancies that are fairly localized in time.” Ex. 1003
¶ 20; see Mot. 12. The proposed substitute claims, which identify sync
points from which forward and backward memory comparisons are
performed, are not limited to only recently transmitted data.
Patent Owner also asserts that the specific combination of “fine and
coarse filtering of hash values” and “forward and backward memory
comparison” limitations recited in the proposed substitute claims represents
a practical solution to the “tradeoff” that exists with data de-duplication
systems, and is not solved by any of the prior art of record. Mot. 12–14.
Specifically, Patent Owner states that there is a “tradeoff” between the size
of an index and the likelihood that an indexed segment will be matched. Id.
Spring uses a small fixed-size cache and small packets, and, thus, works best
for only recently transmitted data. Id. By contrast, according to Patent
Owner, the proposed substitute claims provide a practical solution to the
tradeoff by operating at the byte-level and using a combination of fine and
coarse filters to identify sync points from which forward and backward
memory comparisons are performed. Id. Patent Owner’s explanations
regarding the prior art and why the proposed substitute claims allegedly
would not have been obvious are supported by the record and are persuasive.
We also note that although Petitioner addresses the “fine and coarse
filtering of hash values” limitations in its Opposition (arguing that they are
taught by Hsu), Petitioner does not argue that Hsu—or any other reference—
teaches or suggests the “forward and backward memory comparison”
limitations. Nor does Petitioner propose any specific combination of
references that allegedly would have rendered obvious the proposed
IPR2013-00403
Patent 8,370,583 B2

25
substitute claims as a whole. Thus, Patent Owner’s assertions regarding key
aspects of the prior art, such as Spring, essentially are unrebutted.

Petitioner’s Remaining Arguments
Petitioner makes two additional arguments in its Opposition regarding
proposed substitute claims 28 and 30. First, Petitioner argues that Patent
Owner only attempted to distinguish the proposed substitute claims over the
prior art references at issue in this proceeding—the three McCanne
references—not other prior art of which Patent Owner is aware. Opp. 3–4.
We disagree. In its Motion to Amend, Patent Owner addresses the prior art
of which it is aware generally, including the McCanne references, and
explains why it believes proposed substitute claims 28 and 30 are patentable
over another reference, Spring, and the prior art discussed in the Silver Peak
White Paper, alone or in combination with the McCanne references. See
Mot. 9–15; Reply 1–2. Under the particular factual circumstances of this
case, we determine that Patent Owner has provided a sufficient explanation
and technical reasoning for why proposed substitute claims 28 and 30 are
patentable over the prior art of record and prior art of which Patent Owner is
aware, and Petitioner has not rebutted that showing successfully.
Second, Petitioner argues that Patent Owner failed to provide a
declaration from a technical expert in support of its arguments regarding
patentability, and did not provide evidence as to “what it believes would
constitute the level of ‘ordinary’ skill in the art,” including “some indication
of education, training, and/or work experience.” Opp. 7–8. Testimony from
a technical expert certainly can be helpful to show what would have been
known to a person of ordinary skill in the art and explain the significance of
IPR2013-00403
Patent 8,370,583 B2

26
features added in a proposed substitute claim. See Idle Free at 7–8. It is not
a prerequisite for a motion to amend, however, just as it is not a prerequisite
for a petition seeking inter partes review. Every case is different and will
depend on its own facts. See Perfect Web Techs., Inc. v. InfoUSA, Inc.,
587 F.3d 1324, 1329 (Fed. Cir. 2009) (“Nor are expert opinions always a
prerequisite [for an obviousness analysis].”); Chore-Time Equip., Inc. v.
Cumberland Corp., 713 F.2d 774, 779 n.2 (Fed. Cir. 1983) (“an invention
may be held to have been either obvious (or nonobvious) without a specific
finding of a particular level of skill or the reception of expert testimony on
the level of skill where, as here, the prior art itself reflects an appropriate
level and a need for such expert testimony has not been shown”);
Tr. 37:23–38:16 (Petitioner acknowledging that a technical expert is not
required). Under the particular factual circumstances of this case, we
determine that Patent Owner has provided a sufficient explanation of why
proposed substitute claims 28 and 30 are patentable over the prior art,
without reliance on testimony from a technical expert.
We also disagree with Petitioner’s argument regarding the level of
ordinary skill in the art. Patent Owner explains in its Motion to Amend what
a person of ordinary skill in the art would have known at the time of the ’583
patent, and cites prior art references of record and other references of which
it is aware. See Mot. 9–15 (citing Exs. 1003–1005, 2001, 2002). Moreover,
Petitioner does not explain why an explicit definition of the level of ordinary
skill in the art is necessary under the circumstances, or argue for a particular
definition under which the proposed substitute claims would have been
obvious to a person of ordinary skill. We also note that Petitioner did not
define explicitly the level of ordinary skill in the art for the ’583 patent in its
IPR2013-00403
Patent 8,370,583 B2

27
Petition. Under the particular factual circumstances of this case, we
determine that the prior art references before us and Patent Owner’s
accompanying explanation provide sufficient guidance as to the level of
ordinary skill in the art. See Okajima v. Bourdeau, 261 F.3d 1350, 1355
(Fed. Cir. 2001) (the prior art itself can reflect the appropriate level of
ordinary skill in the art); Toyota at 4 (“[I]t would not be meaningful to say
that a person of ordinary skill in the art possesses this many years of
education and that many years of experience. Rather, the discussion should
be specific about the technical knowledge pertaining to the feature added.”).

Conclusion
Based on the evidence of record in this proceeding, Patent Owner has
shown, by a preponderance of the evidence, that proposed substitute claims
28 and 30 are patentable over the prior art. Patent Owner’s Motion to
Amend is granted as to proposed substitute claims 28 and 30.

5. Claims 29 and 31
A patent owner may propose a reasonable number of substitute claims
for each challenged claim. 35 U.S.C. § 316(d)(1)(B). Absent special
circumstances, it is presumed that only one substitute claim is needed to
replace each challenged claim. 37 C.F.R. § 42.121(a)(3). In addition,
amendments must be responsive to a ground of unpatentability involved in
the trial. 37 C.F.R. § 42.121(a)(2)(i). In Idle Free, the Board explained how
these requirements may be met in the situation where one proposed
substitute claim adds limitations to an original claim and a second,
dependent proposed substitute claim adds further limitations:
IPR2013-00403
Patent 8,370,583 B2

28
Even in the case of proposing only one substitute claim
for a particular challenged claim, if the substitute claim is
presented as patentable over prior art on the same basis that
another substitute claim on which it depends is patentable over
prior art, then the patent owner should provide meaningful
reasons for making the additional changes effected by that
dependent claim. For instance, where independent claim 1 and
dependent claims 2 and 3 are challenged, and claims 2 and 3
each depend on claim 1, a patent owner may propose three
substitute claims 4–6 and indicate that claim 4 replaces claim 1,
claim 5 replaces claim 2, and claim 6 replaces claim 3 . . . .
[I]f the patent owner also proposes to add further features
into proposed substitute claims 5 and 6, the patent owner should
provide meaningful reasons to establish a special circumstance
for adding those features. Without any explanation, at least
facially the insertion of those additional features would not be
responsive to an alleged ground of unpatentability. Showing a
patentable distinction between each of proposed substitute
claims 5 and 6, and proposed substitute claim 4, would be one
such special circumstance. Adding features for no meaningful
reason is generally inconsistent with proposing a reasonable
number of substitute claims, and also not responsive to an
alleged ground of unpatentability. Any such proposed
substitute claim may be denied entry by the Board. See
37 C.F.R. §§ 42.121(a)(2) and (a)(3).
Idle Free at 9–10.
Patent Owner proposes claim 29 as a “substitute for claim 11.” Mot.
2. Claim 11 depends from independent claim 1. Proposed substitute claim
29 modifies the dependency to depend from proposed substitute claim 28,
rather than claim 1, and adds the limitation that the store instruction
indicates to the destination-side appliance an index within the database for
storing “another copy of the locally accessible portion” of the transmitted
data “together with the non-locally accessible portion of the data.” Id.
A similar situation is presented with respect to proposed substitute claim 31,
IPR2013-00403
Patent 8,370,583 B2

29
which is a “substitute for claim 22.” Id. at 4. Claim 22 depends from
independent claim 12. Proposed substitute claim 31 modifies the
dependency to depend from proposed substitute claim 30, rather than claim
12, and adds the “another copy” limitation. Id.
Patent Owner argues that proposed substitute claims 28 and 30 are
patentable based on the limitations they add to the claims for which they are
substitutes (claims 2 and 13). Id. at 9–14. Patent Owner further argues that
proposed substitute claims 29 and 31, which depend from and add additional
limitations to proposed substitute claims 28 and 30, are “independently
patentable over the prior art and over claims 28 [and 30].” Id. at 14–15.
Thus, the issue is whether Patent Owner has shown a special circumstance
for making the additional changes in proposed substitute claims 29 and 31,
such as a patentable distinction over the parent proposed substitute claims.
We conclude that it has not.
Patent Owner acknowledges that “[a] person of ordinary skill in the
art would have recognized that storing multiple copies of data is common in
other contexts, such as for the purposes of improving data reliability,” but
contends that prior art systems did not require “storing data at any particular
index location (or together with any other data).” Id. Patent Owner further
argues that storing an additional copy of data would not have been
“necessarily desirable or an obvious feature to implement” in the data
de-duplication context where storage space is limited. Id.
To demonstrate a patentable distinction over parent proposed
substitute claims 28 and 30, however, we must assume the parent claims to
be prior art. See Idle Free at 9–10. Patent Owner’s analysis does not
account for the parent claims. In particular, parent proposed substitute
IPR2013-00403
Patent 8,370,583 B2

30
claims 28 and 30 recite a “retrieve instruction” instructing the
destination-side appliance to retrieve the “locally accessible portion” of the
data, and a “store instruction” instructing the destination-side appliance to
store the “non-locally accessible portion” of the data. Proposed substitute
claims 29 and 31 recite that the store instruction further indicates to the
destination-side appliance an index within the database for storing “another
copy of the locally accessible portion . . . together with the non-locally
accessible portion of the data.” Thus, the parent proposed substitute claims
already recite the basic feature of retrieving one, previously-stored portion of
the data, and proposed substitute claims 29 and 31 simply add the feature of
storing another copy of that portion. Patent Owner does not explain
sufficiently why storing an additional copy would have been non-obvious,
assuming the former to be prior art. See id. Nor does Patent Owner
distinguish sufficiently the instant situation from other contexts where, as
Patent Owner acknowledges, storing an extra copy of data would have been
“common.” See Mot. 15. Patent Owner also does not cite any evidence in
the record to support its assertion that storage space would have been such a
concern to a person of ordinary skill in the art that storing another copy of
the locally accessible portion of the data would not have been obvious. See
id.
Patent Owner also argues that the “prior art recognize[d] a desire to
link related data segments that are likely to appear together in a data
stream.” Mot. 14. For example, McCanne ’700 uses hierarchical
referencing to group segment references at different indices. Id. at 14–15
(citing Ex. 1004, col. 10, ll. 11–36). Patent Owner contends that, by doing
so, McCanne ’700 teaches away from storing another copy of the locally
IPR2013-00403
Patent 8,370,583 B2

31
accessible portion “together with” the non-locally accessible portion, as
recited in proposed substitute claims 29 and 31. Id. This argument is not
persuasive, as Patent Owner again fails to account for the parent proposed
substitute claims, which recite retrieval of one previously-stored copy of the
data, as prior art. Further, we are not persuaded that McCanne ’700 teaches
away from the proposed substitute claims. A reference does not teach away
if it expresses merely a general preference for an alternative invention from
amongst options available to the ordinarily skilled artisan, and the reference
does not “criticize, discredit, or otherwise discourage the solution claimed.”
In re Fulton, 391 F.3d 1195, 1201 (Fed. Cir. 2004). Patent Owner does not
point to anything in McCanne ’700 criticizing, discrediting, or otherwise
discouraging the storage of an additional copy of the locally accessible
portion of the data together with the non-locally accessible portion. The
mere fact that McCanne ’700 groups related segment references does not
mean that it teaches away from storing an additional copy.
Patent Owner has not shown a patentable distinction between
proposed substitute claims 29 and 31 and their parent proposed substitute
claims 28 and 30, and has not shown any other special circumstance for
adding the “another copy” limitation to proposed substitute claims 29 and
31. The “another copy” limitation, therefore, is not responsive to a ground
of unpatentability involved in the trial. For the same reasons, proposed
substitute claims 29 and 31 amount to a second proposed substitute claim (in
addition to the parent proposed substitute claims 28 and 30) for claims 2 and
13, and Patent Owner has not demonstrated a sufficient need for exceeding
the presumption that only one substitute claim is needed to replace a
challenged claim. Patent Owner’s Motion to Amend, therefore, is denied as
IPR2013-00403
Patent 8,370,583 B2

32
to proposed substitute claims 29 and 31 under 37 C.F.R. §§ 42.121(a)(2)(i)
and (a)(3).

6. Claims 32 and 33
As explained above, the patent owner in an inter partes review bears
the burden in a motion to amend to show a patentable distinction of each
proposed substitute claim over the prior art, including “construction of new
claim terms, sufficient to persuade the Board that the proposed substitute
claim is patentable.” Idle Free at 7; see supra Section II.B.3. “If there is
any new term used in a proposed substitute claim, the meaning of which
reasonably can be anticipated as subject to dispute, the patent owner should
provide a proposed claim construction in the motion to amend. If a
proposed substitute claim adds a means-plus-function element, the
corresponding structure, material, or acts described in the specification
should be identified.” Toyota at 5.
Patent Owner has not met its burden with respect to proposed
substitute claims 32 and 33 for two reasons. First, Patent Owner has not
provided a proposed claim interpretation for certain limitations in the claims
expressed in means-plus-function format under 35 U.S.C. § 112, sixth
paragraph.
7
Proposed substitute claim 32 depends from independent claim
23, and proposed substitute claim 33 depends from proposed substitute claim
32. Mot. 4–6. Independent claim 23 recites eight means-plus-function
limitations (emphases added):

7
Section 4(c) of the AIA re-designated 35 U.S.C. § 112, sixth paragraph, as
35 U.S.C. § 112(f). Because the ’583 patent has an effective filing date
before September 16, 2012, we refer to the pre-AIA version of 35 U.S.C.
§ 112.
IPR2013-00403
Patent 8,370,583 B2

33
23. A network memory system comprising:
a source-site appliance comprising: means for identifying
locally accessible data of at least a destination-site appliance,
means for intercepting transmitted data sent from a source-site
computer and directed to a destination-site computer, means for
performing a determination of whether a portion of the
transmitted data corresponds to the locally accessible data of
the destination-site appliance, means for generating an
instruction based on the determination, and means for sending
the instruction to the destination-site appliance over a wide area
network; and
the destination-site appliance comprising: means for
receiving the instruction from the source-site appliance over the
wide area network, means for processing the instruction to
obtain the transmitted data, and means for transferring the
transmitted data to the destination-site computer.
In the Decision on Institution, we explained that “[b]ecause Patent
Owner did not file a Patent Owner Preliminary Response, we [did] not have
the benefit of ascertaining Patent Owner’s position on the claim construction
of the means-plus-function limitations,” and, therefore, interpreted the
means-plus-function limitations “based on the record before us to the extent
necessary to determine whether to institute an inter partes review.” Dec. on
Inst. 10–14. We specifically informed Patent Owner that it would have the
opportunity in its Patent Owner Response to explain how it believes the
limitations should be interpreted, and instructed the parties that any claim
interpretation should identify specifically the alleged corresponding structure
in the Specification, “including any computer or microprocessor, computer
program, and algorithm.” Id. at 10–11. In addition, prior to Patent Owner
filing its Motion to Amend, we instructed Patent Owner that if its “proposed
substitute claims contain any means-plus-function limitations, Patent Owner
must identify in the respective motion (1) the written description support for
IPR2013-00403
Patent 8,370,583 B2

34
each claim, as a whole, including such a limitation, and (2) the specific
portions of the specification that describe the structure, material, or acts
corresponding to the claimed function(s).” Paper 22, 4 (emphasis added).
In its Motion to Amend, Patent Owner does not provide a proposed
interpretation for any of the means-plus-function limitations in independent
claim 23 (which are part of proposed substitute claims 32 and 33 by virtue of
their dependency), or even indicate whether it agrees or disagrees with the
preliminary interpretations in the Decision on Institution. See Mot. 6–7.
Interpreting the means-plus-function limitations in the proposed substitute
claims is critical to determining whether the proposed substitute claims are
patentable (e.g., determining whether the prior art disclosed identical or
equivalent structure to the corresponding structure for each limitation),
particularly here where the limitations are computer-related. See WMS
Gaming Inc. v. International Game Tech., 184 F.3d 1339, 1349 (Fed. Cir.
1999) (“In a means-plus-function claim in which the disclosed structure is a
computer, or microprocessor, programmed to carry out an algorithm, the
disclosed structure is not the general purpose computer, but rather the
special purpose computer programmed to perform the disclosed
algorithm.”). By not specifying the function and corresponding structure for
each means-plus-function limitation in proposed substitute claims 32 and 33,
Patent Owner has not met its burden to demonstrate patentability.
Second, Patent Owner’s analysis of the two additional means-plus-
function limitations in proposed substitute claim 32 is deficient. Proposed
substitute claim 32 recites:
means for identifying sync points in the transmitted data
having matches in the locally accessible data by (i) determining
hash values corresponding to different byte locations of the
IPR2013-00403
Patent 8,370,583 B2

35
transmitted data, (ii) finely filtering the hash values using a fine
filter to determine a finely filtered set of the hash values
corresponding to fine sync points, and coarsely filtering the
hash values using a coarse filter to determine a coarsely filtered
set of the hash values corresponding to coarse sync points, and
(iii) determining from the finely filtered set of the hash values
and the coarsely filtered set of the hash values, a plurality of
hash values matching hash values of the locally accessible data;
and
means for performing for byte locations with matching
hash values, a forward and backward memory comparison
between the transmitted data and data representing the locally
accessible data, the forward and backward memory comparison
to identify a size of a matching region of the transmitted data
with the data representing the locally accessible data, and
performing based on results of the forward and backward
memory comparison, a determination of a locally accessible
portion of the transmitted data that is locally accessible at the
destination-site appliance and a non-locally accessible portion
of the transmitted data that is not locally accessible at the
destination-site appliance.
Mot. 4–5 (emphases added). Patent Owner identifies the following as the
corresponding structure for both limitations:
at least the “hardware and/or software elements” of the
branch appliance 350 or central appliance 380 as described in
FIGs. 8–9 and [0073]–[0075] [of the ’697 application]
including use of the data structure of FIG. 5 as described in
[0052]–[0059] (including fine filter 560, coarse filter 565, hash
tables 505, 525, and flow history pages 545), and process step
730 of FIG. 7B as described in [0067]–[0068].
Id. at 6–7.
Even assuming (without deciding) that Patent Owner’s identification
of a common structure for two distinct limitations is appropriate, we are not
persuaded that the structure identified is sufficient. The corresponding
structure of a means-plus-function limitation must be “more than simply a
IPR2013-00403
Patent 8,370,583 B2

36
general purpose computer or microprocessor” to avoid impermissible
functional claiming. Aristocrat Techs. Austl. Pty Ltd. v. Int’l Game Tech.,
521 F.3d 1328, 1333 (Fed. Cir. 2008). If a special purpose computer is
required to carry out the recited function, the specification must disclose
“enough of an algorithm to provide the necessary structure” for performing
the function, and the algorithm may be expressed “in any understandable
terms including as a mathematical formula, in prose, or as a flow chart, or in
any other manner that provides sufficient structure.” Finisar Corp. v.
DirecTV Group, Inc., 523 F.3d 1323, 1340 (Fed. Cir. 2008) (citation
omitted). Patent Owner’s proposed interpretation is impermissibly vague, as
Patent Owner identifies “at least” hardware and/or software elements
“including” (but presumably not limited to) certain things. It also does not
identify specifically the steps of any algorithm, instead stating generally the
“use” of a data structure, described in Figure 5 and eight paragraphs of the
Specification, and a single step in a different figure, Figure 7B.
Moreover, Patent Owner’s analysis of the prior art in its Motion to
Amend is not directed to the particular means-plus-function limitations of
proposed substitute claim 32 as Patent Owner proposes they should be
interpreted. Patent Owner instead analyzes the proposed substitute claims as
a group, arguing proposed substitute claims 28, 30, and 32 together, and
proposed substitute claims 29, 31, and 33 together. See Mot. 9–15. Because
Patent Owner has not identified proper corresponding structure for the
means-plus-function limitations in proposed substitute claim 32 and
proposed substitute claim 33, which depends from proposed substitute claim
32, and has not provided a sufficient analysis of the prior art in view of the
IPR2013-00403
Patent 8,370,583 B2

37
proposed substitute claims as interpreted, we determine that Patent Owner
has not met its burden.
Patent Owner has not shown, by a preponderance of the evidence, that
proposed substitute claims 32 and 33 are patentable over the prior art.
Patent Owner’s Motion to Amend, therefore, is denied as to proposed
substitute claims 32 and 33.

III. ORDER
Based on the record presented in this proceeding, Patent Owner has
shown, by a preponderance of the evidence, that proposed substitute claims
28 and 30 are patentable, but has not met its burden with respect to proposed
substitute claims 29 and 31–33.
In consideration of the foregoing, it is hereby:
ORDERED that Patent Owner’s Motion to Amend is granted to the
extent it requests the cancellation of claims 1–27 of the ’583 patent;
FURTHER ORDERED that Patent Owner’s Motion to Amend is
granted as to proposed substitute claims 28 and 30, and denied as to
proposed substitute claims 29 and 31–33;
FURTHER ORDERED that the copy of Exhibit 1002 filed on
June 28, 2013, and labeled as “Part 2 of 3,” is renumbered as Exhibit 1008;
and
FURTHER ORDERED that the copy of Exhibit 1002 filed on
June 28, 2013, and labeled as “Part 3 of 3,” is renumbered as Exhibit 1009.
This is a final decision. Parties to the proceeding seeking judicial
review of the decision must comply with the notice and service requirements
of 37 C.F.R. § 90.2.
IPR2013-00403
Patent 8,370,583 B2

38
PETITIONER:

David M. O’Dell
Andrew S. Ehmke
John Russell Emerson
Kyle Howard
HAYNES AND BOONE, LLP
david.odell.ipr@haynesboone.com
andy.ehmke.ipr@haynesboone.com
russell.emerson.ipr@haynesboone.com
kyle.howard.ipr@haynesboone.com

PATENT OWNER:

Darren E. Donnelly
Jason E. Amsel
FENWICK & WEST LLP
ddonnelly-ptab@fenwick.com
jamsel-ptab@fenwick.com