Ex Parte Goss et al

Patent Trial and Appeal Board

Oct 30, 2018

13777810 (P.T.A.B. Oct. 30, 2018)

UNITED STA TES p A TENT AND TRADEMARK OFFICE
APPLICATION NO. FILING DATE
13/777,810 02/26/2013
73462 7590 11/01/2018
Hall Estill Attorneys at Law
(Seagate Technology LLC)
100 North Broadway, Suite 2900
Oklahoma City, OK 73102-8820
FIRST NAMED INVENTOR
Ryan James Goss
UNITED STATES DEPARTMENT OF COMMERCE
United States Patent and Trademark Office
Address: COMMISSIONER FOR PATENTS
P.O. Box 1450
Alexandria, Virginia 22313-1450
www .uspto.gov
ATTORNEY DOCKET NO. CONFIRMATION NO.
STL17576 5534
EXAMINER
LOONAN, ERIC T
ART UNIT PAPER NUMBER
2131
NOTIFICATION DATE DELIVERY MODE
11/01/2018 ELECTRONIC
Please find below and/or attached an Office communication concerning this application or proceeding.
The time period for reply, if any, is set in the attached communication.
Notice of the Office communication was sent electronically on above-indicated "Notification Date" to the
following e-mail address(es):
danderson@hallestill.com
okcipdocketing@hallestill.com
USPTO@dockettrak.com
PTOL-90A (Rev. 04/07)
UNITED STATES PATENT AND TRADEMARK OFFICE
BEFORE THE PATENT TRIAL AND APPEAL BOARD
Ex parte RYAN JAMES GOSS, DAVID SCOTT EBSEN,
MARK ALLEN GAERTNER, MICHAEL JOSEPH STEINER, and
ANTOINE KHOUEIR
Appeal2018-002875
Application 13/777 ,810
Technology Center 2100
Before MAHSHID D. SAADAT, ALLEN R. MacDONALD, and
JOHN P. PINKERTON, Administrative Patent Judges.
MacDONALD, Administrative Patent Judge.
DECISION ON APPEAL
STATEMENT OF THE CASE
Appellants 1 appeal under 35 U.S.C. § 134(a) from a final rejection of
claims 1-12, 14--16, and 20-24. We have jurisdiction under
35 U.S.C. § 6(b).
We affirm-in-part.
1 According to Appellants, the real party in interest is Seagate Technology
LLC. See App. Br. 1.
Appeal2018-002875
Application 13/777,810
Representative Claims2
Claims 1, 4, 7, 8, and 21 under appeal read as follows ( emphases
added):
1. A method comprising:
receiving data updates to a multi-tier memory structure of
a local user device across a cloud network from a remote server
during a cloud computing session and storing the received data
updates as working data in a first location of an upper memory
tier of the memory structure;
predicting degradation in the first location of the upper
memory tier with a data log circuit;
migrating a current version of the working data to a
different, second location in the upper memory tier responsive
to predict degradation in the first location;
logging the current version of the working data to a lower
memory tier in the memory structure of the local user device as
at least one snapshot responsive to the migration of the current
version of the working data to the different, second location in
the upper memory tier, the upper and lower memory tiers each
comprising rewritable non-volatile memory cells having
different constructions and storage attributes;
generating a request for a selected data update from the
remote server; and
transferring the selected data update from the lower
memory tier to the upper memory tier in lieu of issuing the
request to the remote server.
4. The method of claim 1, wherein the cloud computing
session is a first cloud computing session, and the method further
comprises generating a session log comprising a sequential list
of access operations that were executed upon the working data
during the first cloud computing session, storing the session log
2 Although claim 7 is not representative, claim 8 depends upon claim 7, and
thus, claim 7 is also reproduced for context.
2
Appeal2018-002875
Application 13/777,810
in the lower memory tier of the memory structure, and using the
session log to detect a subsequent, second cloud computing
session matching a portion of the first cloud computing session
and to prefetch data updates from the remote server to the upper
memory tier of the memory structure identified by the session
log.
7. The method of claim 1, further comprising analyzing a
sequence of access operations upon the working data and
initiating a prefetch command script to speculatively request
selected data in anticipation of a future request for said selected
data.
8. The method of claim 7, in which the request for the
selected data is issued to the lower memory tier of the multi-tier
memory structure and the selected data are transferred from
the lower memory tier to the upper memory tier of the multi-
tier memory structure.
21. A storage device comprising:
a multi-tiered memory structure comprising a first non-
volatile memory tier and a second non-volatile memory tier;
a data log circuit adapted to predict degradation in a first
location of the first nonvolatile memory tier; and
a programmable processor having associated
programming in memory which, when executed by the
programmable processor, operates to establish a first cloud
computing session between the storage device and a remote
server across a cloud network, to generate a session log of access
operations carried out during the first cloud computing session
to transfer data objects between the remote server and the first
non-volatile memory tier, to store a duplicate copy of at least one
data object from the first non-volatile memory tier to the second
non-volatile memory tier, to perform post-session processing at
a conclusion of the first cloud computing session by/ orming at
least one data pre/etch command script that describes the
3
Appeal2018-002875
Application 13/777,810
session log at a conclusion of the first cloud computing session,
and to store the at least one data prefetch command script in the
second non-volatile memory tier while maintaining the duplicate
copy of the at least one data object therein, the programmable
processor adapted to migrate a current version of the working
data to a different, second location in the first non-volatile
memory tier responsive to predicted degradation in the first
location.
App. Br. 18-20, 23-24 (Claims Appendix).
Rejections on Appeal3
1. The Examiner rejected claims 4, 5, 10-12, 14--16, and 20 under
35 U.S.C. § 112, second paragraph, as being indefinite for failing to
particularly point out and distinctly claim the subject matter which the
Appellants regard as the invention. 4 See Final Act. 3--4.
2. The Examiner rejected claims 1-12, 14--16, and 20-24 under
35 U.S.C. § I03(a) as being unpatentable over various combinations of Chen
et al. (US 2007/0288589 Al; published Dec. 13, 2007) ("Chen"), Ong
(US 2013/0212207 Al; published Aug. 15, 2013) ("Ong"), Li
(US 2014/0050026 Al; published Feb. 20, 2014) ("Li"), Park
(US 2012/0311237 Al; published Dec. 6, 2012) ("Park"), Ranade et al.
(US 2012/0089781 Al; published Apr. 12, 2012) ("Ranade"), Woolford
(US 2012/0323597 Al; published Dec. 20, 2012) ("Woolford"), and/or
3 The Examiner withdrew the rejection of claims 4, 5, 10-12, 14--16, and 20
under 35 U.S.C. § 112, first paragraph, for failing to comply with the
enablement requirement. See Ans. 2. Thus, this rejection is not before us.
4 The patentability of claims 5, 10-12, 14--16, and 20 is not separately
argued from that of claim 4. See App. Br. 10. Accordingly, except for our
ultimate decision, the rejection of claims 5, 10-12, 14--16, and 20 under
35 U.S.C. § 112, second paragraph, is not discussed further herein.
4
Appeal2018-002875
Application 13/777,810
Prahlad et al. (US 2010/0332401 Al; published Dec. 30, 2010) ("Prahlad"). 5
See Final Act. 5-29.
Issues on Appeal
Did the Examiner err in rejecting claim 4 as being indefinite?
Did the Examiner err in rejecting claims 1, 8, and 21 as being obvious?
ANALYSIS
A. Section 112 Argument
Appellants raise the following argument in contending that the
Examiner erred in rejecting claim 4 under 35 U.S.C. § 112, second
paragraph:
[T]he ordinary artisan would find both visual and textual
support for the claimed matching of portions of first and second
cloud computing devices. In sum, the ordinary artisan would
point to FIGS. 8 and 13 as well as pages 12 and 15 of the
specification as clearly pointing out how a computing session
consists of a sequence of I/0 requests that are evaluated against
other computing session to be the same ("match").
App. Br. 10 (emphases added).
The Examiner finds one of ordinary skill in the art would be "unable
to ascertain the metes and bounds which cover 'matching a portion of an
active session to a session log." Final Act. 3. In response to Appellants'
arguments, the Examiner further finds:
5 The patentability of claims 2-7, 9-12, 14--16, 20, and 22-24 is not
separately argued from that of claims 1, 8, and 21. See App. Br. 10-17.
Accordingly, except for our ultimate decision, the rejection of claims 2-7,
9-12, 14--16, 20, and 22-24 under 35 U.S.C. § I03(a) is not discussed
further herein.
5
Appeal2018-002875
Application 13/777,810
Page 12 of the specification notes "a number of read and/or write
accesses that closely match a previously captured session history
sequence"; However, such a teaching does not define the metes
and bounds which cover 'matching a portion of' and active
session to a session log. In the rejection of record, the Office
presents a scenario ("if a session log states a 'read' was
performed to address 'A' and an active session log performs a
'read' to address 'B ', would that result in 'matching a portion of
as both perform a read?") in which the Office is unable to
determine whether or not such a scenario would be within the
metes and bounds which the applicant defines as 'matching a
portion of as recited in the claims; as such, the Office maintains
the indefiniteness rejection.
Ans. 4 (emphases added).
In response, Appellants further argue:
The ordinary artisan would clearly identify from page 12
of the specification that "accesses that closely match a previously
captured session history" initiate prefetching and from routine
280 of FIG. 13 and page 15 of the specification that a session
detection involves detecting "I/0 requests as well as matching a
sequence of such requests to one or more stored command
history table structures for other, previously encountered
sessions." Hence, the ordinary artisan would recognize that the
claimed "matching a portion" of first and second computing
sessions involves detecting at least one sequence of requests
that are practically the same as in the computing sessions.
As such, the ordinary artisan would respond to the
Examiner's posited scenario by asking if the sequence of
requests in the two session logs are the same. If the additional
requests proximal to the read to address 'A' in the session log
are the same as the requests proximal to the read to address 'B,'
then the ordinary artisan would characterize the session log as
matching a portion of the active session log.
Reply Br. 1-2 ( emphases added).
We are persuaded the Examiner erred. Breadth of a claim is not to be
equated with indefiniteness. In re Miller, 441 F.2d 689,693 (CCPA 1971).
6
Appeal2018-002875
Application 13/777,810
Although the claimed "using ... [a] session log to detect a subsequent,
second cloud computing session matching a portion of ... [a] first cloud
computing session" is broad enough to encompass matching of any portion
of a second cloud computing session with any portion of a first cloud
computing session, the breadth of claim 4, by itself, does not render the
claim indefinite. Further, Appellants provide examples in Appellants'
Specification that allow for a match of any sequence of requests between the
first cloud computing session and the second cloud computing session to
constitute a match of a portion of the first cloud computing session and the
second cloud computing session. See, e.g., Spec. 12:25-28; 15:20-22.
Thus, we conclude claim 4 is not indefinite.
Accordingly, Appellants have shown the Examiner erred in rejecting
claim 4 under 35 U.S.C. § 112, second paragraph. Therefore, we do not
sustain the rejection.
B. Section 103 (a) Arguments
Appellants raise the following argument in contending that the
Examiner erred in rejecting claim 1 under 35 U.S.C. § 103(a):
The Examiner acknowledges that Chen and Ong are each
silent with respect to the claimed migration of data in response
to degradation of data.
From the specification, the data log circuit would be
understood as evaluating "a variety of inputs, such as
read/write counts, parametric drift measurements, temperature
measurements, observed bit error rate (BER) data" that are
"combined into a weighted computation (bloom filter)
empirically determined to correspond to a point in time when
the stored data may begin to exhibit degradation." An ordinary
artisan would appreciate the difference between the claimed
data log circuit that predicts future degradation and the
7
Appeal2018-002875
Application 13/777,810
exclusively reactive wear leveling scheme of Li that reacts to
real-time measured temperature to determine the risk of data
reading errors.
The Examiner posits that all wear leveling inherently is
predictive in nature "in order to extend the memory life" and that
the actual measurement of temperature in Li predicts data
degradation by evaluating "how temperature differences result in
read failures." Any ordinary artisan would recognize the logical
fallacy of the Examiner's assertions in that wear leveling is not
predictive in nature, but instead reactive to a previously
determined number of read/write cycles.
The ordinary artisan would further recognize that Li
teaches no "data log circuit" predicting data degradation and
moving data proactively, as claimed. Instead Li exclusively
determines if a high risk of read error is currently present due
to temperature. Thus, Li would be understood as moving data in
response to actual write temperatures that are known to
correspond with read errors, which is fundamentally different
than speculatively predicting data degradation based on a
variety of factors, as set out in the specification.
The Examiner appears to assert that an ordinary artisan
would combine the very broad web page content management of
Chen with the product test system of Ong and the wear leveling
scheme of Li to arrive at the present claims because they involve
"memory storage systems." The ordinary artisan would disagree
and point to the contradicting purposes of the three references
as evidence of no reasonable motivation to combine the
references.
The ordinary artisan would identify that Chen exclusively
and explicitly handles archiving webpage content while Ong is
directed only to testing write errors in a network and Li
exclusively involves temperature-based wear leveling. The
archiving of webpage content would be understood by the
ordinary artisan as teaching away from a combination with the
testing system of Ong as well as the data wear leveling in Li.
Additionally, the network testing structure and function of Ong
8
Appeal2018-002875
Application 13/777,810
would be viewed by the ordinary artisan as teaching away from
the self-diagnosing wear leveling scheme of Li.
App. Br. 11-13 (Appellants' emphases and citations omitted; panel
emphases added); see also Reply Br. 2-3.
We are not persuaded because this argument is not commensurate
with the scope of claim 1, as the claim merely requires the prediction of
memory degradation and does not require the prediction to be based on
inputs including read/write counts, parametric drift measurements,
temperature measurements, and observed bit error rate ("BER") data. As Li
discloses predicting whether the performance of wear leveling on a flash
memory device will cause memory degradation based on a comparison of an
ambient temperature associated with the writing of data in the memory
device and the current ambient temperature (see Final Act. 7 (citing Li ,r,r 5,
15-18, 27, 29); see also Ans. 5), Appellants have not persuasively
distinguished the claimed prediction of memory degradation from Li's
disclosure. Regarding motivation to combine the cited references, we
conclude the Examiner has articulated a sufficient rationale to combine Ong
and Li with Chen (i.e., to customize Chen's storage entity given benefits and
drawbacks of memory types as taught by Ong, and to prevent data loss in a
flash memory when memory is used outside a threshold temperature as
taught by Chen). See Final Act. 6-7. Accordingly, Appellants have not
shown the Examiner erred in rejecting claim 1 under 35 U.S.C. § 103(a).
Appellants also raise the following argument in contending that the
Examiner erred in rejecting claim 21 under 35 U.S.C. § 103(a):
The Examiner notes that Chen, Ong, and Li are silent with
respect to the claimed session log of access operations and
storage of prefetch command script in a second memory tier of a
memory structure.
9
Appeal2018-002875
Application 13/777,810
The Examiner posits that timestamping data accesses
reads on the claimed session log. The ordinary artisan would
disagree and point at least to FIGS. 5, 8, & 11-13, along with
pages 12-13 of the specification, that describe a session log as
recording the sequence of read and write accesses to working
data along with "suitable metrics such as LBA, time/date,
whether the write operation is an update to a previously stored
LBA or is a newly presented LBA, the source of the written
LBA (if known)." Hence, the simple timestamping of data in
Woolford does not logically teach the claimed session log of
access operations from a remote server to a tier of a memory
structure.
While the deficiencies of Ranade with respect to
prefetching data have been discussed, the Examiner further cites
Ranade as teaching the claimed generation of prefetch command
script that describes the session log. The ordinary artisan would
appreciate at least from FIG. 9 and pages 12-13 of the
specification that a pre/etch command script is generated by a
pre/etch analysis engine as a result of multiple inputs, such as
data accesses session sequence, and session data objects, that
allow for speculative pre/etch commands to be subsequently
executed at will. Such appreciation of what a prefetch command
script entails would be understood as fundamentally different
than the simple pre/etching of data in Ranade based on an
ambiguous "transaction cost."
Review of Woolford and Ranade by the ordinary artisan
would further yield that the references teach away from one
another by respectively involving storage of healthcare data
and network pre/etching of compressed data. That is, the
ordinary artisan would never logically find motivation to
combine the healthcare data organization of Woolford with the
retrieval of compressed network data of Ranade to arrive at the
multitier memory structure with a session log and pre/etch
command script.
App. Br. 16-17 (Appellants' emphasis omitted; panel emphases added); see
also Reply Br. 3--4.
10
Appeal2018-002875
Application 13/777,810
We are not persuaded by this argument either, as this argument is also
not commensurate with the scope of claim 21. More specifically, claim 21
merely requires the generation of a session log that includes access
operations, and does not require the inclusion of metrics, such as a logical
block address ("LBA"), time/ data, whether a write operation is an update to
a previously stored LBA or a newly presented LBA, or a source of a written
LBA. As Woolford discloses a session log including access operations
to/from a network storage device logged by date/timestamp (see Final
Act. 18 ( citing Woolford ,r 111 ); see also Ans. 7), Appellants have not
persuasively distinguished the claimed generation of a session log including
access operations from Woolford' s disclosure. Further, claim 21 also
merely requires the forming of a data prefetch command script associated
with a cloud computing session, and does not require multiple inputs, such
as a data access session sequence, and session data objects. As Ranade
discloses an agent prefetching files from a storage cloud (see Final Act. 19
( citing Ranade ,r 102); see also Ans. 7), Appellants also have not
persuasively distinguished the claimed forming of a data prefetch command
script from Ranade's disclosure. Regarding motivation to combine, we
conclude the Examiner has articulated a sufficient rationale to combine
Woolford and Ranade with Chen, Ong, and Li (i.e., to track time-sensitive
data entries as taught by Woolford, and to reduce access time to data stored
in Chen's memory by speculatively fetching data likely to be requested in
the future as taught by Ranade). See Final Act. 18-19. Accordingly,
Appellants have not shown the Examiner erred in rejecting claim 21 under
35 U.S.C. § 103(a).
11
Appeal2018-002875
Application 13/777,810
Appellants also raise the following argument in contending that the
Examiner erred in rejecting claim 8 under 35 U.S.C. § 103(a):
Upon review of Ranade, the ordinary artisan would identify that
the reference exclusively receives prefetch data from the cloud,
which explicitly contradicts the pre/etching of data from a
lower tier of memory in the same memory structure.
The ordinary artisan would point to FIGS. 5 & 9-13 and
pages 12-15 of the specification as describing how pre/etching
data from a lower memory tier is advantageous over
pre/etching from a cloud location. Therefore, the ordinary
artisan would understand the juxtaposition of the exclusive
cloud pre/etching of Ranade with the lower memory tier
pre/etching presently claimed.
App. Br. 14 (emphases added). 6
In response, the Examiner finds:
The appellant traverses the rejection to the claims 7-9 as a
group with claim 8 being a representative claim. While the
claims are traversed as a group ( thus the claims should stand or
fall as a group), the broadest claim should be the representative
claim of the group. Claims 8 and 9 are dependent on claim 7; as
such, claim 7 should be the representative claim of the group as
arguments directed at claim 8 may not be pertinent to claim 7.
The appellant alleges the specification describes "how
prefetching from a lower memory tier is advantageous over
prefetching from a cloud location". While the Office maintains
a "lower memory tier" may be analogous to "a cloud location",
the Office respectfully notes applicant's arguments are
exclusive to claim 8 and not pertinent to claim 7. As such, the
applicant presents no arguments beyond a motivational argument
to the rejection of claim 7; the Office maintains the motivation to
6 Appellants also argue that Chen, Ong, and Li do not provide any
motivation for a person of ordinary skill in the art to modify and/or combine
Ranade to arrive at the claimed prefetching, and Ranade teaches away from
Chen, Ong, and Li (see App. Br. 14), but it is not necessary to reach these
arguments.
12
Appeal2018-002875
Application 13/777,810
combine cited prior art references as noted in the rejection of
record.
Ans. 6 (emphases added).
As a threshold matter, we use our discretion to treat claim 8 as
representative only of itself, and claim 7 as the representative claim for
claims 7 and 9. As Appellants have not provided any arguments for claim 7,
we sustain the Examiner's rejection of claims 7 and 9. Further, Appellants'
argument is persuasive as to claim 8. We agree with Appellants that Ranade
teaches a speculative fetching module that fetches data from a storage cloud,
but fails to teach or suggest fetching data from a lower memory tier of a
multi-tier memory structure and transferring the fetched data into an upper
memory tier of the multi-tier memory structure.
Accordingly, Appellants have shown the Examiner erred in rejecting
claim 8 under 35 U.S.C. § 103(a). Therefore, we do not sustain the
rejection.
CONCLUSIONS
(1) Appellants have established that the Examiner erred in rejecting
claims 4, 5, 10-12, 14--16, and 20 under 35 U.S.C. § 112, second paragraph.
(2) The Examiner has not erred in rejecting claims 1-7, 9-12, 14--16,
and 20-24 under 35 U.S.C. § 103(a).
(3) Appellants have established that the Examiner erred in rejecting
claim 8 under 35 U.S.C. § 103(a).
(4) Claims 1-7, 9-12, 14--16, and 20-24 are not patentable.
(5) On this record, claim 8 has not been shown to be unpatentable.
13
Appeal2018-002875
Application 13/777,810
DECISION
We reverse the Examiner's rejection of claims 4, 5, 10-12, 14--16, and
20 under 35 U.S.C. § 112, second paragraph.
We affirm the Examiner's rejections of claims 1-7, 9-12, 14--16, and
20-24 under 35 U.S.C. § 103(a).
We reverse the Examiner's rejection of claim 8 under
35 U.S.C. § 103(a).
No time period for taking any subsequent action in connection with
this appeal may be extended under 37 C.F.R. § 1.136(a)(l )(iv).
AFFIRMED-IN-PART
14