International Business Machines Corporationv.Elec. & Telecomm'n Research Inst.

Patent Trial and Appeal Board

Dec 9, 2015

09753245 (P.T.A.B. Dec. 9, 2015)

Trials@uspto.gov
571-272-7822
Paper 28
Entered: December 9, 2015
UNITED STATES PATENT AND TRADEMARK OFFICE
____________
BEFORE THE PATENT TRIAL AND APPEAL BOARD
____________
VMWARE, INC., INTERNATIONAL BUSINESS MACHINES
CORPORATION and ORACLE AMERICA, INC.,
Petitioner,
v.
ELECTRONICS AND TELECOMMUNICATIONS RESEARCH
INSTITUTE,
Patent Owner.
____________
Case IPR2014-00901
Case IPR2014-00949
1

Patent 6,978,346 B2
Before BRIAN J. McNAMARA, MIRIAM L. QUINN, and
GREGG I. ANDERSON, Administrative Patent Judges.
ANDERSON, Administrative Patent Judge.
FINAL WRITTEN DECISION
35 U.S.C. § 318(a) and 37 C.F.R. § 42.73
1
This case was joined with IPR2014-00901 on Jan. 28, 2015 by Order in
IPR2014-00949, Paper 25.
IPR2014-00901
Patent 6,978,346 B2

2
I. INTRODUCTION
A. Background
On July 21, 2014, VMWare, Inc. (“Petitioner”)
2
filed a Second
Corrected Petition requesting an inter partes review of claims 1–9 of U.S.
Patent No. 6,978,346 B2 (Ex. 1001, “the ’346 patent”). Paper 8 (“Pet.”).
Electronics and Telecommunications Research Institute (“Patent Owner”)
filed a Preliminary Response. Paper 11. Based on these submissions, on
December 11, 2014, we granted the Petition and instituted trial for claims 1–
9 of the ’346 patent on one of the grounds of unpatentability alleged in the
Petition. Paper 14 (“Institution Decision” or “Dec. Inst.”).
After institution of trial, Patent Owner filed a Patent Owner Response.
Paper 19 (“PO Resp.”). Petitioner filed a Reply. Paper 28 (“Pet. Reply”).
In addition, the parties rely upon expert testimony. Petitioner proffered the
Declaration of Dr. Robert Horst (“Horst Declaration,” Ex. 1003). Patent
Owner proffered two declarations of Dr. Thomas Conte. First is a
declaration from Dell Inc. v. Electronics and Telecommunications Research
Institute, IPR2013-00635 (“’635 IPR”), a trial directed to the same ’346
patent (“Conte ’635 Declaration,” Ex. 2003). Second is a declaration of Dr.
Conte filed with Patent Owner’s Response (“Conte ’901 Declaration,” Ex.
2301). Patent Owner also filed a declaration of Dr. Randy Katz (“Katz
Declaration,” Ex. 2202), from another inter partes review directed at the
’346 patent, International Business Machines Corp. v. Electronics and
Telecommunications Research Institute, Case IPR2014-00976. A transcript
of Dr. Conte’s deposition (“Conte Dep.,” Ex. 1017) was submitted by

2
International Business Machines Corporation (“IBM”) and Oracle
America, Inc. (“Oracle”) are petitioners in IPR2014-00949, which is joined
to this case, and are included in any reference to “Petitioner.”
IPR2014-00901
Patent 6,978,346 B2

3
Petitioner. A transcript of Dr. Horst’s deposition (“Horst Dep.,” Ex. 2302)
was submitted by Patent Owner.
An oral hearing was held on August 28, 2015. The transcript of the
consolidated hearing has been entered into the record. Paper 34 (“Tr.”).
We have jurisdiction under 35 U.S.C. § 6(c). This Final Written
Decision is issued pursuant to 35 U.S.C. § 318(a). We conclude for the
reasons that follow that Petitioner has not shown by a preponderance of the
evidence that claims 1–9 of the ʼ346 patent are unpatentable.
B. Related Proceedings
Petitioner advises that the ’346 patent is involved in the following co-
pending district court cases: Safe Storage LLC v. StoneFly, Inc., 1:13-cv-
01152;
3
Safe Storage LLC v. Int’l Business Machines Corp., 1:13-cv-01151;
Safe Storage LLC v. Emulex Corporation et al, 1:13-cv-01150; Safe Storage
LLC v. 3PAR Inc., 1:13-cv-01088; Safe Storage LLC v. Oracle America Inc.
et al, 1:13-cv-01089; Safe Storage LLC v. ATTO Technology Inc. et al.,
1:13-cv-01090; Safe Storage LLC v. VMware Inc., 1:13-cv-00928; Safe
Storage LLC v. Promise Technology Inc., 1:13-cv-00927; Safe Storage LLC
v. Nexsan Corporation, 1:13-cv-00931 ; Safe Storage LLC v. Overland
Storage Inc., 1:13-cv-00932; Safe Storage LLC v. IQSS LLC, 1:13-cv-
00930; Safe Storage LLC v. Infortrend Corporation, 1:13-cv-00929; Safe
Storage LLC v. Cisco Systems, Inc., 1:13-cv-00926; Safe Storage LLC v.
Silicon Graphics Int’l Corp., 1:12-cv-01629; Safe Storage LLC v. Dot Hill
Systems Corp., 1:12-cv-01625 ; Safe Storage LLC v. Hitachi Data Systems
Corp., 1:12-cv-01627; Safe Storage LLC v. Dell Inc., 1:12-cv-01624; Safe
Storage LLC v. NetApp Inc., 1:12-cv-01628; Safe Storage LLC v. Hewlett-

3
Litigation concluded. Paper 3, 3.
IPR2014-00901
Patent 6,978,346 B2

4
Packard Company, 1:12-cv-01626, all pending in the United States District
Court for the District of Delaware. Pet. 1, Paper 3, 2–3.
In a final written decision in the ’635 IPR we determined that claims
1–3 and 5–8 of the ’346 patent had not been shown to be unpatentable. ’635
IPR, Paper 39. We declined to institute inter partes review of the ’346
patent in the following cases: Dell Inc., Hewlett-Packard Co., & NetApp,
Inc. v. Electronics & Telecommunications Research Institute, IPR2014-
00152, Paper 12 (PTAB May 16, 2014); Dell Inc., Hewlett-Packard Co., &
NetApp, Inc. v. Electronics & Telecommunications Research Institute,
IPR2014-00549, Paper 10 (PTAB Mar. 26, 2015); and International
Business Machines Corp. v. Electronics & Telecommunications Research
Institute, Case IPR2014-00976, Paper 14 (PTAB Dec. 11, 2014)(“’976
IPR”).
C. The ’346 Patent
The ’346 Patent describes an apparatus with “redundant
interconnection between multiple hosts and a redundant array of inexpensive
disks (hereinafter referred to as ‘RAID’).” Ex. 1001, Abstract. As a result
of the redundant interconnection, the apparatus allows increased bandwidth
in the event one of the two RAID controllers 460 and 461 has a failure. Id.
at 3:1–9.
IPR2014-00901
Patent 6,978,346 B2

5
Figure 4 of the ’346 patent is reproduced below:

Figure 4 is a block diagram of a host matching system including RAID 490
and its interconnection to host computers 400–405. Ex. 1001, 2:643:6.
RAID 490 includes two RAID controllers 460, 461 and hubs 440, 441. Id.
at 3:10–18. Each RAID controller includes a pair of network interface
controllers. For example, RAID controller 460 includes network interface
controllers 470, 471, and RAID controller 461 includes network interface
controllers 480, 481. Id. at 3:11–13. Each host computer has its own
network interface controller (410 to 415), which connects the host computer
through the hubs and to network interface controllers (470, 471, 480, 481) of
RAID controllers 460, 461. Id. at 3:31–35.
The ’346 patent describes that the result is two independent networks
with twice the bandwidth of a single network and a “communication
passage” between the two RAID controllers. Id. at 3:62–64. The
communication passage creates a “fault tolerant function” should one of
RAID controllers 460 or 461 fail. Id. at 3:64–66. According to Figure 4,
communications line 450 interconnects network interface controller 480 of
RAID controller 461 and network interface controller 470 of RAID
IPR2014-00901
Patent 6,978,346 B2

6
controller 460. Id. at 4:2–6; Fig. 4. Then, RAID controller 461 may send
information to RAID controller 460. Id. In like manner, network interface
controller 471 of RAID controller 460 may be connected over
communications lines to network interface controller 481 of RAID controller
461, allowing RAID controller 460 to send information to RAID controller
461. Id. at 3:66–4:2.
By the arrangement described, the apparatus continues to operate in
the event either RAID controller 460 or 461 has an “occurrence of an error.”
Ex. 1001, 4:19–25. The interconnected network interface controller of the
operational RAID controller assumes the functions of the network interface
controller of the failed RAID controller. Id.
D. Illustrative Claims
Claims 1 and 9, the two independent claims of the challenged claims,
are reproduced below:
1. An apparatus for a redundant interconnection between
multiple hosts and a RAID, comprising:

a first RAID controlling units and a second RAID controlling
unit for processing a requirement of numerous host computers,
the first RAID controlling unit including a first network
controlling unit and a second network controlling unit, and the
second RAID controlling unit including a third network
controlling unit and a fourth network controlling unit; and

a plurality of connection units for connecting the first RAID
controlling units and the second RAID controlling unit to the
numerous host computers, wherein the first RAID controlling
unit and the second RAID controlling unit directly exchange
information with the numerous host computers through the
plurality of connecting units, and the first network controlling
unit exchanges information with the fourth network controlling
IPR2014-00901
Patent 6,978,346 B2

7
unit, and the second network controlling unit exchanges
information with the third network controlling unit.
Ex. 1001, 5:7–26.

9. An apparatus for a redundant interconnection between
multiple host computers and a RAID, the apparatus comprising:

a plurality of connection units for connecting the host
computers and the RAID;

a first and a second RAID controllers, included in the RAID,
each of which having a first network interface controller and a
second network interface controller for processing requests
from the plurality of the host computers connected through the
plurality of the connection units,

wherein the first network interface controller in the first RAID
controller supplies data to the host computers connected
through the plurality of connection units and processes
information transmitted from the second network interface
controller in the second RAID controller,

wherein the first network interface controller in the second
RAID controller supplies data to the host computers connected
through the plurality of connection units and processes
information transmitted from the second network interface
controller in the first RAID controller,

wherein the second network interface controller in the first
RAID controller is used for fault tolerance by performing
functions of the first network interface controller in the second
RAID controller when the second RAID controller is faulty,
and

wherein the second network interface controller in the second
RAID controller is used for fault tolerance by performing
functions of the first network interface controller in the first
RAID controller when the first RAID controller is faulty, and
IPR2014-00901
Patent 6,978,346 B2

8
wherein the first network controlling unit in the first RAID
controlling unit exchanges information with the second network
controlling unit in the second RAID controlling unit, and the
second network controlling unit in the first RAID controlling
unit exchanges information with the first network controlling
unit in the second RAID controlling unit.
Ex. 1001, 6:20–59.

E. Grounds Upon Which Trial was Instituted
Inter partes review was instituted on the following grounds: 1–9 as
obvious over Mylex
4
and Hathorn.
5
Dec. Inst. 22.
F. Claim construction
In an inter partes review, “[a] claim in an unexpired patent shall be
given its broadest reasonable construction in light of the specification of the
patent in which it appears.” 37 C.F.R. § 42.100(b); see In re Cuozzo Speed
Techs., LLC, 793 F.3d 1268, 1275–76 (Fed. Cir. 2015); see also Office
Patent Trial Practice Guide, 77 Fed. Reg. 48,756, 48,766 (Aug. 14, 2012)
(Claim Construction). Under the broadest reasonable construction standard,
claim terms are given their ordinary and customary meaning, as would be
understood by one of ordinary skill in the art in the context of the entire
disclosure. In re Translogic Tech., Inc., 504 F.3d 1249, 1257 (Fed. Cir.
2007). Any special definition for a claim term must be set forth in the
specification with reasonable clarity, deliberateness, and precision. In re
Paulsen, 30 F.3d 1475, 1480 (Fed. Cir. 1994). In the absence of such a
special definition or other consideration, “limitations are not to be read into

4
Kevin J. Smith, Mylex Corp., STORAGE AREA NETWORKS; UNCLOGGING
LANS AND IMPROVING DATA ACCESSIBILITY, published May 29, 1998
(“Mylex,” Ex. 1006).
5
US 5,574,950 to R. G. Hathorn et al., issued Nov. 12, 1996 (“Hathorn,”
Ex. 1006).
IPR2014-00901
Patent 6,978,346 B2

9
the claims from the specification.” In re Van Geuns, 988 F.2d 1181, 1184
(Fed. Cir. 1993).
1. Terms Disputed After the Decision on Institution
Patent Owner identifies three terms for construction in its Response.
PO Resp. 12–22. Petitioner contests two of the terms, “RAID” and “hub and
switch.” Pet. Reply 2–7.
a. “RAID” (Claims 1 and 9)
In the Final Written Decision in the ’635 IPR (“’635 Final Decision”
or “’635 Final Dec.”), we construed “RAID,” as the term is used in the ’346
patent, to mean “a single logical unit for mass storage using multiple
physical disk drives.” ’635 Final Dec. 9. In the Petition, prior to the ’635
Final Decision, Petitioner asserted that “RAID” should be construed as
“redundant array of inexpensive disks.” Pet. 6 (citing Ex. 1001, Abstract,
Ex. 1003 ¶¶ 14–16).
Both Patent Owner’s Response and Petitioner’s Reply were filed after
the ’635 Final Decision. For purposes of this case, Patent Owner accepts the
construction from the ’635 Final Decision, providing facts and arguments
supporting that construction. PO Resp. 12–16. Petitioner’s Reply continues
to urge adoption of its proposed construction from the originally filed
Petition. Pet. Reply 2–6.
The written description of the ’346 patent restates the acronym for
RAID, but otherwise lacks additional description of RAID or its
functionality. Consistently throughout the written description, RAID is
referred to in the singular, i.e., “the apparatus for a redundant
interconnection between multiple hosts and a RAID comprises a plurality of
RAID controllers.” Ex. 1001, 2:16–18 (emphasis added). The claims also
IPR2014-00901
Patent 6,978,346 B2

10
recite “a RAID.” Id. at 5:7–8. Figure 4 of the ’346 patent shows RAID 490
as a single component within a box which includes two RAID controllers
460 and 461. Similarly, the ’346 patent represents the prior art RAID as a
single component. Ex. 1001, Fig. 1, element 130; Fig. 2, element 240; Fig.
3, element 340. Neither party relies on the prosecution history (Exhibits
1002, 2001), and our independent review of that history failed to reveal any
additional insight as to the term’s meaning.
Patent Owner’s expert, Dr. Conte, testified that a RAID is a mass
storage device built from multiple, physical disk drives. Ex. 2003 ¶ 18, Ex.
2301 ¶ 30. Patent Owner and Dr. Conte both point to dictionary evidence to
respectively argue and opine that a RAID is a “single logical drive.” PO
Resp. 14, Ex. 2301 ¶ 30 (citing Exs. 2004,
6
2005
7
). That a RAID is built
from multiple, physical disk drives is uncontroverted. That a RAID is a
single logical unit is supported by Petitioner’s expert, Dr. Horst, who
testified that “[a] RAID array needs to be able to present a set of drives as a
single logical unit, but it does not always have to present a single logical
unit.” Ex. 2302, 16:14–21.
Although the two dictionary definitions cited by Patent Owner and Dr.
Conte are found in dictionaries published after the foreign priority date of
the ’346 patent, they further substantiate the proposition that those of
ordinary skill generally refer to RAID as a single logical unit. Webster’s, for
example, generally defines RAID as proposed by Petitioner. Ex. 2004, 308.
However, Webster’s proceeds to define various levels of RAID, e.g., RAID
0, 1, 2, 5, and 10, all including as part of the definition a “single logical

6
Webster’s Computer Dictionary (9th ed. 2001)(“Webster’s”).
7
Microsoft Computer Dictionary (5th ed. 2002).
IPR2014-00901
Patent 6,978,346 B2

11
drive.” Id. Similarly, the Microsoft Computer Dictionary’s definition of
RAID, which does not define levels of RAID, states that the “data is
distributed across a group of computer disk drives that function as a single
storage unit.” Ex. 2005, 437.
The prior art also supports a construction of RAID as being a “single
logical drive.” Weygant discloses that a RAID is a single logical unit, but
also in “various combinations of striped and mirrored configurations.” Ex.
2101, 153.
8
Chen’s discussion of RAID technology states that the problem
of obtaining high performance is addressed by “arrays, which organize
multiple independent disks into a large, high-performance logical disk.” Ex.
2102, 2 (emphasis added).
In further support of its position that our construction from the ’635
Final Decision is correct, Patent Owner presents evidence not of record in
the ’635 IPR, including DeKoning (U.S. Patent No. 6,073,218, Ex. 1010).
Patent Owner quotes from DeKoning as explaining that a RAID controller
“makes the subsystem [RAID] appear to the host computer as a single,
highly reliable, high capacity disk drive.” PO Resp. 15–16 (citing Ex.
1010, 1:65–2:14).
Patent Owner quotes from the petition in the ’976 IPR (Ex. 2201),
where Petitioners IBM and Oracle argued that “the Chong reference (U.S.
Patent No. 6,070,251; Ex. 2303) discloses a RAID ‘because the two data
storage devices appear to the hosts as a single, reliable drive.’” PO Resp. 16
(citing Ex. 2201, 13). Patent Owner also points to testimony from the Katz
Declaration in the ’976 IPR that Chong’s alleged “combination of data

8
Page reference is to the actual page number of Weygant and Chen and not
Patent Owner’s exhibit number.
IPR2014-00901
Patent 6,978,346 B2

12
mirroring and fault tolerance makes the two data storage devices appear as
a single, reliable drive to the hosts, or in other words, a RAID.” Id. at 16
(citing Ex. 2202 ¶ 36)(emphasis added).
Petitioner argues for its proposed construction, Redundant Array of
Inexpensive Disks, is based on the use of quotation marks around “RAID” as
used in the ’346 patent. Pet. Reply 2–3 (citing Ex. 1001, 1:9–10). We are
not persuaded that the use of the acronym in the Specification dictates that
the broadest reasonable interpretation of RAID is the words of the acronym.
For example, the Specification describes RAID as a single item, i.e., as a
“single logical entity.”
Petitioner argues that the construction of RAID from the ’635 Final
Decision is incorrect for a number of reasons. Petitioner points out that the
Webster’s definition does not support the “single logical drive” construction
of RAID. Pet. Reply 3 (citing Ex. 2004, 308). As discussed above,
Webster’s describes RAID generally as including levels, then defines the
levels separately, each as including a “single logical drive.” However, that
the general RAID definition and the definition of RAID 10 from Webster’s,
or any other dictionary, do not include the “single logical drive,” is not
dispositive. Rather, we must look to how the person of ordinary skill in the
art would understand a RAID. See Translogic Tech., 504 F.3d at 1257. The
evidence of what would be understood by one of ordinary skill, as
represented by the testimony of Drs. Horst, Conte, and Katz, and authors
Weygant, Chen, and Chong, is compelling. Petitioner’s argument based on
the Webster’s definition of RAID is, at best, equivocal from an evidentiary
standpoint. Indeed, Petitioner acknowledges that RAID can present as a
“single logical unit [drive].” Pet. Reply 3. We, therefore, are not persuaded
IPR2014-00901
Patent 6,978,346 B2

13
that Petitioner has met its burden of proving that the proffered dictionaries
support a broader construction of RAID, i.e., to not be defined as a “single
logical drive.”
Petitioner also submits various arguments concerning Hathorn’s
teachings of the storage devices to inform our analysis of the scope of the
term “RAID.” Pet. Reply 36. These arguments do not persuade us to
adopt Petitioner’s proposed construction. First, Petitioner’s arguments beg
for a comparison of the definition of RAID to the storage devices disclosed
in Hathorn. We are not persuaded by Petitioner’s arguments for three
reasons.
First, we are tasked in this Decision with determining whether
Hathorn, the asserted prior art, discloses a “RAID.” Therefore, relying on
Hathorn to guide the definition of the term is putting the cart before the
horse. Second, Hathorn’s teaching of a RAID is inconclusive. Hathorn
statement of a RAID is limited to a discussion in the Background of the
Invention where a “redundant array of inexpensive devices (RAID)” is
discussed as a data back-up alternative. Ex. 1005, 2:47. Hathorn provides
no further instruction as to whether the disclosed disks operate in a RAID
configuration.
Third, Hathorn confirms our analysis of the scope of the term RAID
including a single logical unit. Hathorn appears to dismiss the use of a
RAID configuration and single volume of mirrored data for use in disaster
recovery of an entire system or site. Id. at 2:1924. For example, Hathorn
states that the reason the aforementioned back-up solutions (including a
RAID configuration) are useful only for device failures is that “the
secondary data has the same volume serial numbers (VOLSERs) and DASD
IPR2014-00901
Patent 6,978,346 B2

14
addresses as the primary data.” Id. at 2:1618. That is, a RAID-based back-
up had the feature of appearing as a single unit. Hence, Hathorn focuses its
disclosure of the invention as multiple DASDs with emphasis on remote
dual copies. See, e.g., Ex. 1005, 5:2536, 6:2139 (describing “remote dual
copy systems” and referring to Figure 2 depicting multiple primary DASDs).

9
We also note that despite RAID being known at the time, Hathorn does
not mention a RAID in describing the Hathorn remote dual copy system.
Petitioner concludes with an argument that our construction of RAID
from the ’635 Final Decision “excludes a fundamental aspect of a
RAID―redundant storage―yet includes terms such as ‘single logical unit,’
‘mass storage,’ and ‘physical disk drives’ that are found nowhere in the ’346
patent.” Pet. Reply 6. Petitioner’s argument does not cite any supporting
authority and, accordingly, is unpersuasive. Further, while the Specification
is important in our construction, we still need to consider how RAID would
be understood by the person of ordinary skill. See Translogic, 504 F.3d at
1257.
Accordingly, having considered the claim construction arguments
made by Petitioner in its Reply regarding Hathorn, and weighing that against
the evidence presented by Patent Owner, including the testimony of Dr.
Conte, which we credit, we conclude that a RAID would have been
understood by a person of ordinary skill in the art to have a customary
meaning of “a single logical unit for mass storage using multiple physical
disk drives.”

9
Petitioner’s arguments regarding Hathorn’s allege disclosure of a RAID are
considered in more detail in the analysis section of this Decision.
IPR2014-00901
Patent 6,978,346 B2

15
b. “connection unit/hub/switch” (Claim 5)
In the ’635 Final Decision, we found that, consistent with the
definition provided in the Specification, “connection unit” is “a hub or
switch.” ’635 Final Dec. 14 (citing Ex. 1001, 3:13–18). We noted that
Figure 4 of the ’346 patent shows components 440 and 441 labeled as a
“HUB OR SWITCH.” Id. Petitioner proposed this construction in the
Petition and in the Reply. Pet. 7, Pet. Reply 6–7. Thus, we determined that
the Specification treats “hub” and “switch” as equivalents. ’635 Final Dec.
14.
Relying on the Conte Declaration, Patent Owner argues the person of
skill in the art would view “hub” as different from “switch.” PO Resp. 20
(citing Ex. 2301 ¶¶ 58–59). Patent Owner acknowledges that the ’346 patent
equates hub and switch, but argues that they are not the same “for all
purposes.” Id. at 21. If “hub” were intended to mean “hub or switch,”
Patent Owner argues that instead of “HUB OR SWITCH” as the label for
connection units 440 and 441 in Figure 4, the drawings would have just used
the word “hub.” Id.
Patent Owner argues our construction based on the Specification
“does not so clearly redefine the term ‘hub’ so as to conflate terms that have
different meanings.” PO Resp. 21–22 (citing Renishaw PLC v. Marposs
Societa’ per Azioni, 158 F.3d 1243, 1249 (Fed. Cir. 1998)). Patent Owner
also argues against our construction based on claim differentiation, arguing
claim 5 refers to “hub equipment” whereas claims 6 and 7 refer to “network
switch equipment.” Claims 5 and 6 are otherwise identical, claim 7 differing
with respect to another limitation not at issue.
IPR2014-00901
Patent 6,978,346 B2

16
We are not persuaded by Patent Owner’s arguments. It is always
necessary to review the Specification to determine whether the inventor has
used any terms in a manner inconsistent with their ordinary meaning. See
Vitronics Corp. v. Conceptronic, Inc., 90 F.3d 1576, 1582 (Fed. Cir. 1996).
10

The claims must be read in view of the Specification. Id. Patent Owner’s
evidence of how a person of ordinary skill would understand the terms is
that the terms, though different, are both “network interconnection devices.”
Ex. 2301 ¶¶ 58–59. Additionally, Patent Owner’s expert acknowledged
“hub” and “switch” may be used interchangeably. See Pet. Reply 7 (citing
Ex. 2302, 30:16–18). The Specification recognizes this as well, stating that
“hubs 440, 441 are provided to connect a system connected to these hubs by
one network and maintain the network …, and it can be as a hub or a switch.
Hereinafter, they are named a ‘hub’ altogether.” Id. at 6–7 (citing Ex.
1001, 3:13–18).
Thus, consistent with the Specification and our construction from the
’635 Final Decision, “connection unit” is “a hub or switch.”
c. “exchange/exchanges information” (Claims 1 and 9)
In the ’635 Final Decision, we interpreted “exchange” and “exchanges
information” according to their ordinary sense: to transmit and receive
information reciprocally.
11
’635 Final Dec. 12. The Petitioner proposed this
construction in the Petition. Pet. 7 (citing Ex. 1003 ¶¶ 14–16).

10
While district courts do not apply our broadest reasonable interpretation
standard in infringement cases, the Federal Circuit is our reviewing court
and this principle is universal.
11
Definition exchange (vb) (3), WEBSTER’S THIRD NEW INTERNATIONAL
DICTIONARY, UNABRIDGED (1993), available at
http://lionreference.chadwyck.com (Dictionaries/Webster’s Dictionary)
(Exhibit 3001).
IPR2014-00901
Patent 6,978,346 B2

17
Patent Owner neither agrees or disagrees with the construction from
the ’635 Final Decision. PO Resp. 17–19. Instead, based on the language
differences in claims 1, 4, and 9, Patent Owner contends there are
differences in how “network interface controllers” exchange information.
Id. at 17. Patent Owner does not raise any issue as to how our construction
of “exchanges information” should change as a result.
Consistent with our construction from the ’635 Final Decision, we
construe “exchange/exchanges information” to mean “to transmit and
receive information reciprocally.”
2. Terms not Disputed After the Decision on Institution
The following two terms were construed in the Institution Decision
but not addressed in Patent Owner’s Response nor Petitioner’s Reply.
a. “RAID controller/RAID controlling unit” (Claims 1 and 9)
In the’635 Final Decision, we construed “RAID controlling unit,” as
the term is used in the ’346 patent, to mean “a component that controls
operation of the RAID.” ’635 Final Dec. 10–11. Petitioner proposed this
construction in the Petition. Pet. 7 (citing Ex. 1003 ¶¶ 14–16).
Neither the Response nor the Petitioner’s Reply propose any
alternative construction. Accordingly, applying the broadest reasonable
interpretation to the term RAID controlling unit, we construe “RAID
controlling unit” to mean “a component that controls operation of the
RAID.”
b. “network interface controller”/”network controlling
unit”/”network interface controlling unit” (Claims 1 and 9)
In the ’635 Final Decision we interpreted “network controlling unit.”
as “a component, connected to a network, for providing communication over
the network.” ’635 Final Dec. 12–14. Petitioner treats the terms “network
IPR2014-00901
Patent 6,978,346 B2

18
interface controller,” “network controlling unit,” and “network interface
controlling unit” as equivalents and urges they should be construed
identically. Pet. 7, see also Prelim. Resp. 20 (“synonymous phrases
‘network controlling unit’ and ‘network interface controller’”).
The Specification describes each “network interface controller” as
having a counterpart “network interface controller,” each “network interface
controller” associated with a separate RAID controller. Ex. 1001, 2:2630.
Furthermore,
information from a second network interface controller 622 of a
first RAID controller 620 is sent to a first network interface
controller 632 of a second RAID controller 630, and
information from a second network interface controller 632 of
the second RAID controller 630 is transmitted to a first network
interface controller 621 of the first RAID controller 620.
Id. at 4:4046. Thus, we construe “network controlling unit” and its
equivalents to mean “a component, connected to a network, for providing
communication over the network.”
II. ANALYSIS
Petitioner contends that claims 19 of the ’346 patent are obvious
under 35 U.S.C. § 103 over Mylex and Hathorn. Pet. 1742. To support
this position, Petitioner relies on the testimony of Dr. Horst. Ex. 1003
¶¶ 56138.
IPR2014-00901
Patent 6,978,346 B2

19
A. Overview of Prior Art
1. Mylex
Mylex generally describes Storage Area Networks (“SAN”) and
associated architecture of such networks. Ex. 1006, 4.
12
Storage Area
Networks can be configured with switched fabrics or hubs and switches to
exchange data between nodes of the network. Id. at 8.
Figure 6 of Mylex is reproduced below.

Figure 6 shows that switches are used to create fibre channel fabric. Id.
Mylex describes that hub-connected Storage Area Networks bandwidth per
node decreases as more nodes are added while bandwidth of fabric
connected nodes increases as nodes are added. Id.
Figure 7 of Mylex is reproduced below.

12
Page references are to Mylex page numbers at the bottom right corner of
each page. Petitioner uses the same numbers and not the exhibit page
numbers centered at the bottom of the page.
IPR2014-00901
Patent 6,978,346 B2

20
Figure 7 shows a four node cluster with shared access to RAID arrays. Id.
at 9. Mylex describes that external RAID controllers can be used in a write-
back caching scheme to protect data. Id. at 12.
Mylex controllers are linked by a private network used to transmit
“I’m alive” heartbeat messages. Ex. 1006, 17. The absence of heartbeat
messages signals that one of the controllers is off-line and the remaining
controller immediately initiates a failover operation and then begins
servicing I/O requests directed to itself and its off-line partner to provide
non-stop access to data. Id. Figure 17 of Mylex is reproduced below.

As shown in Figure 17, for fibre controllers, Mylex describes that if a
controller fails, i.e., a failover, the surviving controller senses the absence of
heartbeats and fails over the ID of the active port (Port 1 and Port 2) on the
failed controller to its reserved port (Reserved), and updates its data
structures with configuration information stored on the disk. Id. at 19.
2. Hathorn
Hathorn discloses a “remote dual copy system” with dynamically
modifiable ports on the storage controller that are alternatively configurable.
Ex. 1005, Abstract, 7:57–8:16, Fig.3. A primary storage controller can
appear as a host processor to a secondary storage controller. Id. Hathorn
describes a method for communicating between host processors and storage
IPR2014-00901
Patent 6,978,346 B2

21
controllers, or between storage controllers. Hathorn incorporates direct
access storage device (“DASD”) for storage of data. Id. at 1:28–33.
Figure 3 of Hathorn is reproduced below.

Figure 3 is a block diagram of the preferred embodiment of the remote
dual copy system of the invention described in Hathorn. Ex. 1005, 5:27–30.
The storage components 323, 326, 333, and 336 are labelled as DASDs.
Primary storage controller 322 communicates through port A 321 with
secondary storage controller 332. Id. at 8:1115. As shown in Figure 3,
port A 321 acts as a channel link-level facility through communication links
350, dynamic switch 305, communication links 351, dynamic switch 315,
and communication links 346 to communicate with secondary storage
controllers 332 and/or 335. Id.
IPR2014-00901
Patent 6,978,346 B2

22
B. Obviousness Analysis
1. Introduction
Graham v. John Deere Co., 383 U.S. 1, 17–18 (1966) establishes an
objective analysis for determining obviousness under 35 U.S.C. § 103(a) as
follows:
[T]he scope and content of the prior art are ... determined;
differences between the prior art and the claims at issue are ...
ascertained; and the level of ordinary skill in the pertinent art
resolved. Against this background, the obviousness or
nonobviousness of the subject matter is determined. Such
secondary considerations as commercial success, long felt but
unsolved needs, failure of others, etc., might be utilized to give
light to the circumstances surrounding the origin of the subject
matter sought to be patented.
See 35 U.S.C. § 103(a).
Under the Graham analysis, Petitioner has identified the difference
between Mylex and the claims at issue: that Mylex does not show “a direct
exchange of information between network interface controlling units.” Pet.
18. Petitioner asserts that the difference is found in Hathorn. Id. Patent
Owner argues that the person of ordinary skill in the art would not have
combined Hathorn with Mylex, making several arguments in support. PO
Resp. 26–56.
Dr. Horst testifies that, as of the year 2000, the person of ordinary
skill would have had a B.S. in Electrical Engineering or Computer Science
and at least two years of experience in designing storage systems. Ex. 1003
¶ 12. Dr. Conte does not provide any specific opinion in this regard. See,
e.g., Ex. 2301 ¶ 25. We determine Dr. Horst’s opinion is both not disputed
and reasonable. Accordingly, we adopt Dr. Horst’s description of the level
of ordinary skill for this case.
IPR2014-00901
Patent 6,978,346 B2

23
KSR Int’l Co. v. Teleflex Inc., 550 U.S. 398 (2007) provides a
framework for analyzing obviousness of claimed subject matter. KSR
addresses specifically combining elements found in the prior art to establish
that the differences between the claim and the prior art combination would
have been obvious to the person of ordinary skill at the time the invention
was made. KSR summarizes factors considered previously, including
whether the combination yields no more than predictable results. KSR, 550
U.S. at 416.
Often, it will be necessary to look to interrelated teachings of
multiple patents; the effects of demands known to the design
community or present in the marketplace; and the background
knowledge possessed by a person having ordinary skill in the
art, all in order to determine whether there was an apparent
reason to combine the known elements in the fashion claimed
by the patent at issue.
Id. at 418. “[I]t can be important to identify a reason that would have
prompted a person of ordinary skill in the relevant field to combine the
elements in the way the claimed new invention does.” Id. At its core, the
conclusion of obviousness under § 103 “cannot be sustained by mere
conclusory statements; instead, there must be some articulated reasoning
with some rational underpinning to support the legal conclusion of
obviousness.” Id.
2. Summary of Parties’ Arguments
Petitioner argues that claims 1–9 of the ’346 patent are obvious over
Mylex and Hathorn. Pet. 17. Petitioner argues specifically that Mylex
teaches all the limitations of claims 1–9 with the exception of a “direct
exchange of information between network interface controlling units.” Id. at
18. Petitioner acknowledges that Mylex discloses only “heartbeat”
IPR2014-00901
Patent 6,978,346 B2

24
communication, but alleges that Hathorn teaches use of “the existing switch
network for communication between RAID controllers.” Id.
Petitioner argues that one of ordinary skill would have been motivated
to combine Hathorn with Mylex. Pet. 14–15 (citing Ex. 1003 ¶¶ 33–34).
First, Petitioner alleges that there is a “close relationship” between the
references. Pet. 14–15. Petitioner next argues that both Mylex and Hathorn
are directed to the same field of endeavor. Id. at 15 (citing Ex. 1003 ¶¶ 33–
34). Petitioner further argues, regarding motivation, that both Mylex and
Hathorn disclose redundancy in “terms of sending communications between
two or more RAID controllers and/or network interface controller ports.” Id.
Petitioner argues that Hathorn teaches that communication paths are
expensive and discloses using a switch network for communication, and is
less costly than the direct heartbeat path of Mylex. Id. at 18 (citing Ex. 1003
¶¶ 32, 48–55). Petitioner also argues more generally that, because of their
similarity, the modification of Mylex’s heartbeat communications with the
information exchange taught by Hathorn would lead to predictable results.
Id. at 15 (citing Ex. 1003 ¶¶ 3334).
Patent Owner’s main argument is that Hathorn is not a RAID and its
teachings are not readily combinable with the RAID of Mylex. PO Resp.
26–27. Patent Owner lists additional reasons why the proposed modification
of Mylex with Hathorn would not have been obvious. Id. at 27. The
additional reasons are as follows: (1) it would not have been possible to
modify Mylex to transmit those messages via the fibre loop as claimed
because the “Reserved” ports are inactive/offline until a failure occurs on the
primary ports (PO Resp. 28–30); (2) the only feasible way to transport the
heartbeats over the host-side network in Mylex would not satisfy the claim
IPR2014-00901
Patent 6,978,346 B2

25
language (PO Resp. 30–33); (3) modifying Mylex to send the heartbeats via
the host-side network is not supported by a sufficient motivation (PO Resp.
33–37); (4) the proposed modification is not a combination of known
elements according to known methods to yield predictable results (PO Resp.
37–46); and (5) if one skilled in the art did not want to send heartbeats over
the dedicated path provided for that purpose, then there is another alternative
that would have been more apparent to one skilled in the art than the
modification proposed by the Petitioner (PO Resp. 46–49).
As set forth in Petitioner’s Reply, Petitioner responds to Patent
Owner’s main argument by stating Mylex and Hathorn are “readily
combinable.” Pet. Reply 7. Specifically, Petitioner contends Mylex and
Hathorn both disclose a RAID. Id. at 8. Even if Hathorn does not disclose a
RAID, Petitioner asserts Hathorn is “properly combinable” with Mylex. Id.
at 8–13.
3. Petitioner’s Arguments
Petitioner has the burden of proof, and, thus, we address each of the
arguments made by Petitioner in support of combining Hathorn with Mylex.
a. Whether Hathorn is Analogous Art to Mylex and a RAID
While Patent Owner concedes that Hathorn and Mylex are in the same
field of endeavor, it challenges Petitioner’s contention that Hathorn is a
RAID. Tr. 44:9–45:3.
As discussed above in the claim construction section, Hathorn
discloses a remote dual copy system based on the use of DASDs. Ex. 1005,
6:22–25, Fig. 2. Petitioner argues that Hathorn is a RAID based on our
construction here and in the ’635 Final Decision because Hathorn’s “two
controllers are logically paired for data shadowing and treated as a single
IPR2014-00901
Patent 6,978,346 B2

26
logical unit by the hosts.” Pet. Reply 8 (citing Ex. 1005, 9:45–49; Ex. 2302,
20:21–21:1, 107:9–25). We find that Petitioner has failed to meet its burden
to show that Hathorn’s DASDs teach or suggest the use of a RAID, as we
have defined the term.
First, we are not persuaded by Petitioner’s argument that Hathorn is a
RAID that operates in two modes, one where storage is presented as a single
logical unit to the host. Pet. Reply 3–4. The argument is premised on
Hathorn’s disclosure that explains that the secondary device or volume, i.e.,
storage device, is “fenced from” normal data access but still receive control
commands. Id. at 4 (citing Ex. 1005, 9:45–49). Petitioner then notes that
RAIDs, according to Dr. Horst’s deposition testimony, are “fenced off” so
that the host cannot access individual drives of the array. Id. (citing Ex.
2302, 20:21–21:1, 107:9–25). The cited Hathorn disclosure occurs in the
discussion of “data shadowing,” where primary and secondary DASDs are
paired into duplex pairs. Ex. 1005, 9:29–51. Hathorn describes “data
shadowing” as the storage of data in a “secondary or remote location.” Id. at
2:40–44. The use of “fenced from” in Hathorn, however, only means that
for each pair of DASDs, the “secondary device or volume is fenced from
normal data access.” Id. at 9:45–49. Although there is some similarity
between the function of Hathorn’s DASDs and a RAID, with regard to a
“fenced from” volume of data, the fact remains that the multiple duplex pairs
of Hathorn have not been shown to teach a single logical unit or device.
We also find unpersuasive the argument that Hathorn discloses a
RAID because a RAID can have different configurations, such as
individually addressable disks or multiple logical units. Pet. Reply, 4–5
(citing Ex. 2302, 161:25–162:17, 16:14–21). The testimony in this regard
IPR2014-00901
Patent 6,978,346 B2

27
does not support Petitioner’s argument. We find that at best, Dr. Horst’s
testimony alludes to the ability of a system to include both a RAID volume
and individual disks. See Ex. 2302, 161:79 (“the system at 3Ware could be
configured either with RAID volumes or what’s called JABOD. Just a
bunch of disks is what that stands for.”) (emphasis added). Further, the
proffered testimony is inconsistent with the context of the claim language.
See id. at 162:1617 (“As long as it supports some level of RAID, it’s still a
RAID.”). The claims do not refer to “supporting a RAID.” The claims
affirmatively require “a RAID.” That is, the claims are not concerned with
whether the disks could be configured as a RAID. Therefore, we find
unpersuasive the testimony bearing on whether the capability of a RAID
configuration is present in Hathorn.
Petitioner argues that because a RAID can perform “mirroring,” like
Hathorn, then Hathorn is a RAID. Petitioner cites as support Patent Owner’s
expert, Dr. Conte, who testified that the ’346 patent encompasses RAID
mirroring. Id. at 5 (citing Ex. 1017, 52:15–17). Petitioner acknowledges the
’346 patent uses RAID in the singular, but contends that does not preclude
reference to mirroring, i.e., RAID level 1. Pet. Reply 5 (citing Ex. 1010,
1:57–60, Ex. 2302, 16:24–17:15). Again, we find that Hathorn’s reference
to mirroring does not make Hathorn’s DASDs a RAID. We agree with
Patent Owner that Hathorn is directed to non-RAID mirroring. See Tr.
45:19–46:13. That a RAID may also perform the mirroring function does
not tend to prove the DASDs of Hathorn are a RAID, as claimed.
IPR2014-00901
Patent 6,978,346 B2

28
The only use of the word “RAID” in Hathorn is as follows:
Another data back-up alternative that overcomes the need
to double the storage devices involves writing data to a
redundant array of inexpensive devices (RAID) configuration.
In this instance, the data is written such that the data is
apportioned amongst many DASDs. If a single DASD fails,
then the lost data can be recovered by using the remaining data
and error correction procedures. Currently there are several
different RAID configurations available.

Ex. 1005, 2:4–11. Hathorn describes that a mirroring RAID is useful for
device failures, not system failures, because that type of back-up method
results in the primary and secondary DASD having the same serial number.
Id. at 2:1417. For this reason, Hathorn turns to a “remote dual copy
system.” Id. at 7:57–8:16, Fig.3. For example, Figure 3 of Hathorn labels
the storage components 323, 326, 333, and 336 as DASDs, which are direct
access storage devices, each having its own storage controller. Id.
Therefore, although we agree that Hathorn discusses mirroring, Hathorn
does not employ RAID mirroring. We find, instead, that Hathorn describes
mirroring as a known feature, in the Background of the Invention, and
instead implements a new copy technique, using dynamic port assignments
for either channel or control unit link-level facility. Id. at 8:315. The
evidence is that Hathorn teaches DASDs not implemented as a RAID.
Petitioner argues “[b]oth Mylex and Hathorn disclose RAID 1-type
systems (disk mirroring/shadowing).” Pet. 15 (citing Ex. 1006, 12; Ex.
1005, 1:9–12). As was discussed above, that a RAID may also perform the
mirroring function does not tend to prove the DASDs of Hathorn are a RAID
as claimed. Based on our construction of RAID, we agree with Patent
Owner that Hathorn’s disks are not a RAID.
IPR2014-00901
Patent 6,978,346 B2

29
Notwithstanding the lack of a RAID in Hathorn, we separately
determine whether Petitioner has provided a sufficient rationale for a person
of ordinary skill in the art to combine Hathorn and Mylex. One of the
arguments presented involves the allegation that Mylex and Hathorn may be
combined based on both being analogous art. Pet. 14–15, Pet. Reply 8–13.
That Mylex and Hathorn are in one field is not, by itself, dispositive. See
KSR, 550 U.S. at 418. We must look to other factors, like design incentives
and other market forces, to determine whether the person of ordinary skill
would be prompted to make the modification. Id.
In reviewing the record here, Petitioner’s arguments in its briefs that
Hathorn is analogous art, i.e., in the same field, do not reach, let alone
explain, why the person of ordinary skill would be prompted to combine the
two. Pet. 14–15, Pet. Reply 8–13, 20. At the oral hearing, Petitioner did not
point to any additional argument or evidence on its field of endeavor
argument. Tr. 6:14–23, 7:13–14. Conversely, Patent Owner asserted
correctly that, even if the field of endeavor is the same, “[y]ou still need
some motivation or other rationale to make a combination and that is what is
missing in this case.” Id. at 47:5–7. Accordingly, Petitioner has not
persuaded us that the fact that Mylex and Hathorn are in the same field of
endeavor establishes a rationale for combining the two.
b. Close Ownership Relationship of Mylex and Hathorn
Petitioner alleges there is a motivation to combine based on a close
ownership relationship because of the common ownership by IBM of Mylex
(Mylex Corporation was acquired by IBM) and Hathorn (assigned to IBM).
Pet. 14–15 (citing Ex. 1003 ¶ 34). Petitioner argues common sense, i.e.,
“one skilled in the art reviewing either reference would look to teachings of
IPR2014-00901
Patent 6,978,346 B2

30
related references within this close business relationship.” Pet. Reply 19.
As legal authority, Petitioner cites to KSR’s statement that: “[r]igid
preventative rules that deny factfinders recourse to common sense, however,
are neither necessary under our case law nor consistent with it.” Id. at 19–20
(citing KSR, 550 U.S. at 426).
We are not persuaded that the close ownership relationship would
prompt the person of ordinary skill to make the proposed combination.
Petitioner does not explain how the person of ordinary skill would be
motivated to combine the teachings of the references because of the “close
ownership relationship” between the two references. Neither are we
persuaded that the use of common sense endorsed in KSR would form a
rationale for making the combination. Other than a common ownership by
IBM, there is no factual support for the allegation that a person of ordinary
skill would have had the common sense to combine the teachings of these
references, at the time of the invention. At best, Petitioner alleges, without
factual support, that “later IBM products were partly based on the
technology IBM acquired by Mylex.” Pet. 1415. Even if we regarded the
allegation as substantiated with facts, the inference we draw from this
allegation is that IBM, at some unknown date, made use of acquired
technology in products branded by IBM. Nevertheless, there is no evidence
of the alleged activity by IBM being relevant to the Mylex teachings at issue
in this proceeding or relevant to the Hathorn system. There is also no
evidence of the time frame of these events or to what extent there were
modifications. In short, we are not persuaded that later IBM products
alluded to by Petitioner is evidence of the motivation to combine the
particular teachings at issue in this proceeding. We find the close
IPR2014-00901
Patent 6,978,346 B2

31
relationship allegation proffered by Petitioner in this case an unsatisfactory
rationale or motivation to combine Hathorn with Mylex.
c. Whether Both Mylex and Hathorn Disclose Redundant
Communication and Mylex Would be Modified According to
Hathorn’s Teachings

Petitioner asserts that both Mylex and Hathorn describe redundancy
and fault tolerance in terms of sending communications between two or
more RAID controllers and both disclose RAID 1-type systems (disk
mirroring/shadowing). Pet. 15 (citing Ex. 1006, 12; Ex. 1005, 1:9–12); Pet.
Reply 20. Petitioner concludes that these similarities would motivate a
person of ordinary skill to modify Mylex’s heartbeat communication system
with the information communication taught by Hathorn, leading to
predictable results. Pet. 15. (citing Ex. 1003 ¶¶ 33–34).
Citing to the Horst Declaration, Petitioner contends that “[o]ne of
ordinary skill in the art would have been motivated to use [the Hathorn]
configuration to achieve the redundancy goals set forth in the Mylex paper
…”. Pet. Reply 20 (citing Ex. 1003 ¶ 118). At the oral hearing, however,
Petitioner acknowledged that the heartbeat path of Mylex is just a single
path and not a redundant path.
13
Tr. 17:23–18:12. Also in argument at the
oral hearing, Petitioner summarized its position on why one of ordinary skill
would combine Hathorn with Mylex: “[s]o for redundancy and cost
purposes, one would be inclined to look at Hathorn, realize it teaches that
you can change the communication paths and can achieve both of those

13
Petitioner cites to the Conte Deposition where Dr. Conte states that
Hathorn and Mylex both “disclose redundancy.” Pet. Reply 9 (quoting Ex.
1017, 51:25–53:9). However, Dr. Conte clarifies that both disclose
“redundant storage,” as opposed to redundant pathways. Ex. 1017, 44:12–
14, 52:5–9.
IPR2014-00901
Patent 6,978,346 B2

32
benefits.” Id. at 13:3–5. Petitioner also conceded that, other than Figures 2
and 3 of Hathorn (see Pet. Reply 12–13), Hathorn does not specifically
address redundancy. Id. at 18:23–19:14.
The Horst Declaration (Ex. 1003 ¶¶ 33–34), cited by Petitioner at
page 15 of its Reply, repeats the field of endeavor and ownership arguments
discussed above without any additional factual support for the conclusion
that the person of ordinary skill would be motivated to combine Hathorn
with Mylex. As such, we give it little weight. See 37 C.F.R. § 42.65(a).
Regardless, Patent Owner asserts that the alleged combination is not
feasible because Mylex’s reserved ports are inactive until a failover
condition is detected. PO Resp. 28–30; see Ex. 1006, 18, Fig. 15.
According to Patent Owner, it is impossible for both alleged first and fourth
network controllers to exchange information, as required by claims 1 and 9,
because they are never active at the same time. Id. at 29–30 (citing Ex. 2301
¶¶ 92–96). Patent Owner explains that after a failover in Mylex, the
“Reserved” port assumes the ID address for the failed Port 1 or Port 2. Id. at
30–31. Therefore, at all times, the Mylex system has only two network
controllers exchanging information with each other, even though the two
others are on reserve.
Petitioner’s assertion that the Reserved port is a “network controlling
unit,” is shown in Petitioner’s annotation of Figure 17 of Mylex.
Petitioner’s annotation of Figure 17 is reproduced below.
IPR2014-00901
Patent 6,978,346 B2

33
Petitioner’s Annotated Figure 17 of Mylex

Pet. 18. Petitioner argues the above annotation illustrates the alleged
exchange of information of claim 1’s first and and fourth (blue path) and
second and third (green path) “network controlling units.” Id. ( citing Ex.
1003¶¶ 32, 48–55). Patent Owner contends that this does not meet the
limitation because:
In other words, the two ports at the ends of this blue path share the
same address/ID. As such, they can never communicate with each
other on the network.
Likewise, the green path [referring to drawing figure] is also an
impossible communication path, as the “Reserved” port of Controller
0 is inactive and has no network address/ID before failover.

PO Resp. 30–31.
Patent Owner argues this deficiency in Mylex is not cured by Hathorn
because it does not teach activating an inactive port. Id. at 30. Petitioner’s
response to this argument is that there is a motivation to make the
modification “so that the expense of the heartbeat path is avoided by
modifying the network interface controlling units to exchange information
with each other, as claimed by the ’346 patent.” Pet. Reply 13 (citing Ex.
1003 ¶¶ 80, 157; Ex. 2302, 95:12–25). Petitioner alleges that Patent
Owner’s argument has no substance because the reserved ports can be made
addressable. Pet. Reply 17.
IPR2014-00901
Patent 6,978,346 B2

34
Patent Owner takes the next step of showing what would be a feasible
modification to Mylex, based on its own annotation of Figure 17, as shown
below:
Patent Owner’s Annotated Figure 17 of Mylex

PO Resp. 32. Patent Owner argues the path shown in purple in its
annotation of Figure 17 of Mylex (communication between Port 1 of
Controller 0 and Port 2 of Controller 1) could be used instead of the
“Heartbeats” path shown (direct connection between Controller 0 and
Controller 1). Id. (citing Ex. 2301 ¶ 99). However, as Patent Owner argues,
and we agree, this proposed path does not satisfy the language of claim 1
reciting “the first network controlling unit exchanges information with the
fourth network controlling unit, and the second network controlling unit
exchanges information with the third network controlling unit.” Id. “In
other words, claim 1 requires the blue path and the green path, not the purple
path. Claim 9 uses different language but requires the same.” Id. In
addition, Patent Owner points out correctly that two communication paths
are required by the claim language. Id. As previously discussed,
Petitioner’s position requires use of the inactive “Reserved” ports to meet
the claim language, making the proposed modification not feasible.
IPR2014-00901
Patent 6,978,346 B2

35
Petitioner argues the operability of Mylex’s reserved ports being
addressable because fibre channel chips can have multiple addresses at
different times. Pet. Reply 15 (citing Ex. 1017, 80:22–81:2), 21 (citing Ex.
1006, 14). Further, Petitioner argues Dr. Conte agrees that the address of
ports in a fibre channel system could be changed in “multiple ways.” Id. at
16. Although Petitioner argues that Hathorn teaches establishing peer-to-
peer paths, Petitioner acknowledges that Hathorn does not “specifically
disclose the modification of the ports.” See Pet. Reply 1617; Tr. 25:12–13.
We disagree with how Petitioner has characterized the testimony of
Dr. Conte to support Petitioner’s contention of a motivation to combine
Hathorn with Mylex. The line of questioning of Dr. Conte referred to
whether a person of skill in the art would know from the disclosure of the
’346 patent how the data is rerouted in the event of a failover. Ex. 1007,
72:975:24. Dr. Conte testifies that the way to achieve that in the 2000 time
frame was to reconfigure the switch. Id. And he clarified that whether a
port address in an active network interface controller could be changed was
an option “if there was no other facility to change the routing in an active
switch.” Id. at 79:17822. That is, we are not persuaded that Dr. Conte
admitted that the port addressability was achievable under the failover
conditions disclosed in Mylex and Hathorn. Neither are we persuaded that
Dr. Conte’s testimony regarding how a person of ordinary skill would regard
the disclosure of the switch operation of the ’346 patent switch compels us
to find that Mylex and Hathorn would be equally regarded.
Furthermore, we are not persuaded that Hathorn’s teaching of peer-to-
peer paths, as alleged in the Reply, rebuts the inference of unfeasible
communication path drawn from the disclosure of Mylex’s “Reserved”
IPR2014-00901
Patent 6,978,346 B2

36
ports. First, the ground of unpatentability at issue relies on modification of
Mylex’s reserved ports to establish the heartbeat path through a network
switch, motivated by Hathorn’s communication paths through network
switches. See Ex. 1003 ¶¶ 78–80. Petitioner does not allege, nor do we find,
that Hathorn teaches or suggests modifying an offline port in a controller to
establish a peer-to-peer connection through the network switch. The address
of the Hathorn secondary device, although set to zero, is an address
nonetheless. See Ex. 1005 at 11:6212:2. We credit Dr. Conte’s testimony
that “the storage pertinent controller ports in Hathorn are always active.”
Ex. 2301 ¶ 116. Petitioner’s offer of Hathorn’s peer-to-peer connection as
some evidence of addressability of ports is incompatible with Mylex’s port
operation.
Moreover, even if Mylex’s “Reserved” ports were to be modified to
operate like Hathorn’s ports, Petitioner offers no credible explanation of how
or why the person of ordinary skill would make modification. The Petition
states that the expense of communication paths would be reduced “by
modifying network interface controlling unit ports to use the existing switch
network for communications between RAID controllers (instead of using a
direct “hearbeat” path). Pet. 18 (emphasis added). The Declaration of Dr.
Horst repeats this general statement verbatim with no added factual support
or further explanation. See Ex. 1003 ¶ 32.
Furthermore, to the extent Petitioner’s declarant testifies that Mylex’s
ports are replaced with or substituted by Hathorn’s “dynamically modifiable
ports,” we find the testimony not persuasive. See Ex. 1003 ¶¶ 4855 (cited
in Petition at page 18). Instead, we agree with Patent Owner that Hathorn’s
ports are not interchangeable with Mylex’s ports because of the fundamental
IPR2014-00901
Patent 6,978,346 B2

37
operational differences of these devices. See PO Resp. 5254. For example,
we credit Dr. Conte’s testimony that the Hathorn ports operate as endpoints
of established links in a circuit-switched network, whereas Mylex’s ports are
nodes in a fibre-channel network that use packet-switched communications.
Ex. 2301 ¶ 117. Further, we credit Dr. Conte’s testimony that the nature of
the heartbeats and disk drive traffic data are different such that the links
dedicated to transferring disk drive traffic data would not motivate the
elimination of the heartbeat path. Id. at ¶¶ 118–119. Lastly, we credit Dr.
Conte’s testimony that the transmission protocols in Hathorn and Mylex are
different standards, Hathorn’s is IBM’s proprietary technology, and
Mylex’s is an open, industry standard. Id. at ¶¶ 48, 85. As such, we find
that although Hathorn’s controller ports are dynamically modifiable, the
modifications are the result of a protocol and link architecture that is unique
to Hathorn and not interchangeable with Mylex’s system.
In summary, we are not persuaded that both Mylex and Hathorn
disclose redundant communications, as opposed to redundant storage.
Therefore, we find unpersuasive Petitioner’s proffered obviousness rationale
that relies on redundancy. Notwithstanding the redundancy issue, we are
persuaded by Patent Owner’s argument and evidence of record that the
proffered rationale of Mylex’s modification, as proposed by Petitioner, is not
reasonable. Consequently, we find that a person of ordinary skill would not
be prompted to make the proposed modification given the unfeasibility of
the proposed redundant communication and the differences in the taught
ports of Mylex and Hathorn.
IPR2014-00901
Patent 6,978,346 B2

38
d. Cost as a Motive to Modify Mylex With Hathorn
Petitioner contends that a person of ordinary skill would modify the
Mylex direct path for sending the heartbeat messages by substituting a
switch network, such as taught by Hathorn, for communication between
RAID controllers. Pet. 18 (citing Ex. 1003 ¶¶ 32). This approach, according
to Petitioner is less costly than the direct heartbeat path of Mylex. Id. (citing
Ex. 1003 ¶¶ 32, 48–55). Dr. Horst states that Hathorn teaches that the need
for expensive direct controller-to-controller communication, as disclosed by
Mylex, is avoided by using the switch network of Hathorn. Id.; see Ex. 1003
¶ 54. Petitioner points to its annotated Figure 17 of Mylex (reproduced
above) as illustrative of the proposed combination. Id.
As Patent Owner notes, Petitioner’s proposed combination requires
modification of Mylex to send the heartbeat message over the fibre loop
(football shaped connection between HBAs of Figure 17 above) rather than
through Mylex’s private network specifically provided for the heartbeats.
PO Resp. 33 (citing Pet. 18, 25–27, 42; Ex. 1003 ¶¶ 32, 76–80, 138). Patent
Owner does not dispute that controller-to-controller ESCON (enterprise
system connections) described in Hathorn (see Ex. 2301 ¶¶ 83–86) are
expensive, but notes that is because the sites are far apart. PO Resp. 34
(citing Ex. 2301 ¶ 103). Mylex, according to Patent Owner, has two co-
located RAID controllers, and, thus, the distance expense is avoided. Id.
(citing Ex. 2301 ¶ 104). Petitioner argues the lower cost motivation in its
Reply without pointing to persuasive evidence that we may disregard the
distance differences between Hathorn and Mylex. See Pet. Reply 19.
Patent Owner hypothetically proposes that, if one skilled in the art
“were motivated to eliminate the direct heartbeat connection and to send the
IPR2014-00901
Patent 6,978,346 B2

39
heartbeats by some other path, then there is another, more apparent path to
do so.” PO Resp. 46. That path is via the disk bus illustrated in Figures 23
and 24 of Mylex. Id. To the extent that the direct heartbeat path is too
costly, the disk channel already exists and is utilized already for cache
mirroring. Id. (citing Ex. 2301 ¶¶ 108–110).
We agree with Patent Owner that there is no evidence that the
proposed cost savings motivation would be applicable to Mylex. PO Resp.
35. We also agree that one of ordinary skill would look to other alternatives,
i.e., Mylex alone discloses less expensive ways to communicate between the
RAID controllers. Thus, we are not persuaded that cost consideration is a
reasonable rationale to combine Hathorn with Mylex.
e. The Combination Would Lead to Predictable Results
Petitioner argues that both Mylex and Hathorn use known techniques
and components to achieve predictable results. Pet. 15 (citing Ex. 1006, 15);
Ex. 1005, 6:25–34). Petitioner concludes that as a result of their similarity,
“one of ordinary skill would have been able to apply the fault tolerance
teachings of Mylex to the system disclosed by Hathorn, or the modifying
NICs to communicate teachings of Hathorn to the system disclosed by
Mylex with predictable results.” Id. (citing Ex. 1003 ¶¶ 33–34).
Petitioner argues “predictable results” based in part on KSR. See KSR
550 U.S. at 416 (“The combination of familiar elements according to known
methods is likely to be obvious when it does no more than yield predictable
results.”). Patent Owner acknowledges both references would use
unspecified computer equipment. PO Resp. 38. Patent Owner argues that
the Petition fails to set forth evidence with specificity. Id. For example,
Patent Owner contends there is a lack of evidence that “the particular
IPR2014-00901
Patent 6,978,346 B2

40
components [as between Mylex and Hathorn] disclosed were compatible or
interchangeable.” Id. (citing Ex. 2301 ¶ 111).
In its Reply, Petitioner argues there are “functional and structural
similarities” associated with redundant storage systems. Pet. Reply 21.
Absent some specificity, we find that Petitioner’s argument is conclusory
and lacks factual support sufficient to weigh in favor of a finding of
obviousness. KSR, 550 U.S. at 418 (“[R]ejections on obviousness grounds
cannot be sustained by mere conclusory statements; instead, there must be
some articulated reasoning with some rational underpinning to support the
legal conclusion of obviousness”)(internal citation omitted).
Patent Owner argues that the combination of Mylex and Hathorn does
not teach “that sending heartbeats over the host-side network, such as a fibre
channel network, was known.” PO Resp. 40. Although this argument may
go to motivation, it does not relate to a claim element, as Patent Owner
proposes. Id. at 39 (citing MPEP § 2143(I)(1)).
Nonetheless, we agree with Patent Owner that what was known to
Mylex was a direct path transport for heartbeats, and nothing in Hathorn
pertains to transporting heartbeats at all. Id. at 40. Patent Owner contends
that Hathorn’s system has no need for heartbeats, and we agree. Id. (citing
Ex. 2301 ¶ 112). This fact persuades us that a person of ordinary skill faced
with Mylex would not need to consult Hathorn for teachings on improving
the heartbeat communication exchange or that there is a problem that would
be solved by the combination.
Petitioner’s predictable result argument is also unpersuasive because
there is no evidence or argument that a person of ordinary skill would be
motivated to make the combination in the first instance. As KSR points out:
IPR2014-00901
Patent 6,978,346 B2

41
When there is a design need or market pressure to solve a
problem and there are a finite number of identified, predictable
solutions, a person of ordinary skill has good reason to pursue
the known options within his or her technical grasp. If this leads
to the anticipated success, it is likely the product not of
innovation but of ordinary skill and common sense. In that
instance the fact that a combination was obvious to try might
show that it was obvious under § 103.
KSR, 550 U.S. at 421 (emphasis supplied).
The Petition lacks specific evidence that the claimed invention could
be constructed by interconnecting off-the-shelf components in order to lead
to a predictable result. See Ex. 2301 ¶ 111. We find instructive Dr. Conte’s
testimony that given the differences between the heartbeats of Mylex and the
Hathorn system, customization would have been required. Id. The
customization would raise additional problems to the person of ordinary skill
making the motivation to combine less likely. Id. We credit Dr. Conte’s
testimony in this regard. Therefore, we are not persuaded that Petitioner has
met its burden in showing that the combination of Mylex and Hathorn are
merely the use of known techniques to achieve a predictable result.
f. Conclusion
Petitioner has failed to show by a preponderance of the evidence that
the person of ordinary skill would have been motivated to combine Hathorn
with Mylex. Thus, Petitioner has not shown that claims 1–9 would have
been unpatentable over Mylex and Hathorn.
IPR2014-00901
Patent 6,978,346 B2

42
III. ORDER
For the reasons given, it is
ORDERED that claims 1–9 of U.S. Patent No. 6,978,346 have not
been shown by a preponderance of the evidence to be unpatentable; and
FURTHER ORDERED that, because this is a final written 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.

IPR2014-00901
Patent 6,978,346 B2

43

For PETITIONER:

Katherine Lutton
lutton@fr.com

Timothy Riffe
IPR27450-0011IP1@fr.com

For PATENT OWNER:

Matthew Phillips
Matthew.phillips@renaissanceiplaw.com

Derek Meeker
Derek.meeker@renaissanceiplaw.com