Ex Parte Lee et al

Patent Trial and Appeal Board

Mar 14, 2017

12813635 (P.T.A.B. Mar. 14, 2017)

United States Patent and Trademark Office
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
APPLICATION NO. FILING DATE FIRST NAMED INVENTOR ATTORNEY DOCKET NO. CONFIRMATION NO.
12/813,635 06/11/2010 Gunho Lee 82261088 7383
56436 7590 03/16/2017
Hewlett Packard Enterprise
3404 E. Harmony Road
Mail Stop 79
Fort Collins, CO 80528
EXAMINER
MUDRICK, TIMOTHY A
ART UNIT PAPER NUMBER
2194
NOTIFICATION DATE DELIVERY MODE
03/16/2017 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):
hpe.ip.mail@hpe.com
chris. mania @ hpe. com
PTOL-90A (Rev. 04/07)
UNITED STATES PATENT AND TRADEMARK OFFICE
BEFORE THE PATENT TRIAL AND APPEAL BOARD
Ex parte GUNHO LEE, NIRAJ TOLIA, and
PARTHASARATHY RANGANATHAN
Appeal 2014-003448
Application 12/813,635
Technology Center 2100
Before KRISTEN L. DROESCH, MICHAEL J. STRAUSS, and
LINZY T. McCARTNEY, Administrative Patent Judges.
STRAUSS, Administrative Patent Judge.
DECISION ON APPEAL
Appeal 2014-003448
Application 12/813,635
STATEMENT OF THE CASE
Appellants appeal under 35 U.S.C. § 134(a) from a rejection of claims
1, 3—14, and 16—20. Claims 2 and 15 are canceled. We have jurisdiction
over the pending claims under 35 U.S.C. § 6(b).
We reverse.
THE INVENTION
The claims are directed to a computing cluster performance
simulation using a genetic algorithm solution. Spec., Title. Claim 1,
reproduced below with a disputed limitation emphasized in italics, is
illustrative of the claimed subject matter:
1. A computer implemented method comprising:
identifying, using a system having a hardware processor,
a search space comprising a plurality of interconnected
computers, the search space to be used to satisfy a resource
request;
selecting, using the system, from the search space an initial
candidate set of clusters, each cluster of the initial candidate set
of clusters comprising a unique combination of computers to
satisfy the resource request',
assigning, using the system, a fitness score, based upon a
predicted performance, to each cluster of the initial candidate set
of clusters;
transforming, using the system, the initial candidate set of
clusters into a fittest candidate set of clusters, wherein each
cluster of the fittest candidate set is an offspring from at least one
cluster of the initial candidate set; and
selecting, using the system, at least one cluster of the fittest
candidate set of clusters having a best predicted performance to
satisfy the resource request.
2
Appeal 2014-003448
Application 12/813,635
REFERENCES
The prior art relied upon by the Examiner in rejecting the claims on
appeal is:
Di Martino, Vincenzo, "Sub Optimal Scheduling In A Grid
Using Genetic Algorithms." Parallel Computing 30.5 (2004): 553—
565. (Hereinafter “Di Martino”.)
Hartmann, Sonke. Self-Adapting Genetic Algorithms with an
Application to Project Scheduling. No. 506. Manuskripte aus den
Instituten fur Betriebswirtschaftslehre der Universitat Kiel, 1999.
(Hereinafter “Hartmann”.)
REJECTION
The Examiner rejected claims 1, 3—14, and 16—20 under 35 U.S.C.
§ 103(a) as being unpatentable over Di Martino in view of Hartmann. Final
Act. 2-13.
APPELLANTS’ CONTENTION1
Appellants contend Di Martino discloses simulating an allocation of a
sequence of jobs to nodes of a computing grid, “not a cluster including a
unique combination of computers [as required by independent claims 1,10,
and 14].” App. Br. 7.
ANALYSIS
We have reviewed the Examiner’s rejections in light of Appellants’
arguments that the Examiner has erred in rejecting independent claims 1,10,
and 14 under 35 U.S.C. § 103(a) as being unpatentable over Di Martino in
1 We note Appellants raise additional contentions of error but we do not
reach them as our resolution of this contention is dispositive of the appealed
rejection under 35 U.S.C. § 103(a).
3
Appeal 2014-003448
Application 12/813,635
view of Hartmann. We agree with Appellants’ conclusions as to this
rejection of the claims.
The Examiner finds Di Martino’s disclosure that “[t]he goal of the
superscheduler is to find the allocation sequence on each node of a
computational Grid that minimize the release time of jobs with respect to a
set of constraints” teaches or suggests the disputed limitation that each
cluster of the initial candidate set of clusters comprises a unique combination
of computers to satisfy a resource request. Final Act. 3. Appellants contend,
rather than disclosing each node has a unique combination of computers, the
cited portion of Di Martino discloses a single grid configuration used in
simulating difference sequences of job submissions, Appellants emphasizing
De Martino’s stated goal of minimizing job release time by optimizing “the
allocation sequence on each node of a computation Grid.” App. Br. 8.
The Examiner responds by finding “the superscheduler must take into
account the different types of computer resources available in the entire
computer grid” which is used as a constraint in defining hardware
architecture and alternatives. Ans. 4. The Examiner further finds
De Martino’s grid environment of interconnected computing nodes, each
with a set of heterogeneous and possibly different computing recourses,
teaches or suggests clusters having unique combinations of computers. Ans.
6. The Examiner finds De Martin’s simulation of different job allocation
sequences, which identifies those having the highest fitness function, teaches
selecting an initial candidate set of clusters as required by claim 1. Id.
We are not persuaded De Martino teaches or suggests the disputed
limitation. The Examiner fails to identify any portion of De Martino
explicitly disclosing clusters including unique combinations of computers.
4
Appeal 2014-003448
Application 12/813,635
De Martino’s disclosure is simply too vague about what constitutes a node
and how resources are selected during each simulation run, if at all.
Although the Examiner’s explanation of how De Martino’s simulation
probably operates is reasonable, the explanation goes beyond what is
actually disclosed. For example, although the nodes may have different or
unique combinations of computers (see Ans. 6), we cannot sustain a
rejection based on speculation alone. Therefore, constrained by the record
before us, we do not sustain the rejection of independent claims 1, 10, or 14
under 35 U.S.C. 103(a) over Di Martino and Hartmann or the rejection of
dependent claims 3—9, 11—13, and 16—20 which stand with their respective
based independent claims.
DECISION
We reverse the Examiner’s decision to reject claims 1, 3—14, and 16—
20.
REVERSED
5