Ex Parte Cases et al

Patent Trial and Appeal Board

Jun 27, 2016

12133480 (P.T.A.B. Jun. 27, 2016)

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.
12/133,480 06/05/2008 Moises Cases 5120.005/
AUS920071048US1
9748
108176 7590 06/28/2016
Heslin Rothenberg Farley & Mesiti/ GlobalFoundries
5 Columbia Circle
Albany, NY 12203
EXAMINER
DINH, PAUL
ART UNIT PAPER NUMBER
2851
MAIL DATE DELIVERY MODE
06/28/2016 PAPER
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.
PTOL-90A (Rev. 04/07)

UNITED STATES PATENT AND TRADEMARK OFFICE
________________

BEFORE THE PATENT TRIAL AND APPEAL BOARD
________________

Ex parte MOISES CASES, JINWOO CHOI, BHYRAV M.
MUTNURY, and CALEB J. WESLEY
________________

Appeal 2014-007656
Application 12/133,480
Technology Center 2800
________________

Before JEAN R. HOMERE, JEREMY J. CURCURI, and
JOHN R. KENNY, Administrative Patent Judges.

CURCURI, Administrative Patent Judge.

DECISION ON APPEAL
Appellants appeal under 35 U.S.C. § 134(a) from the Examiner’s
rejection of claims 1–20. Notice of Appeal. We have jurisdiction under 35
U.S.C. § 6(b).

Appeal 2014-007656
Application 12/133,480

2
Claims 1, 3–6, 8–11, 13–16, and 18–20 are rejected under 35 U.S.C. §
102(e) as anticipated by Usadi (US 2010/0082142 A1; Apr. 1, 2010). Non-
Final Act. 4–7.
Claims 1, 3–6, 8–11, 13–16, and 18–20 are rejected under 35 U.S.C. §
102(b) as anticipated by Daems (US 2005/0257178 A1; Nov. 17, 2005).
Non-Final Act. 7–10.
Claims 1, 3–6, 8–11, 13–16, and 18–20 are rejected under 35 U.S.C. §
102(b) as anticipated by Wang (US 6,578,176 B1; June 10, 2003). Non-Final
Act. 10–11.
Claims 1, 3–6, 8–11, 13–16, and 18–20 are rejected under 35 U.S.C. §
102(e) as anticipated by Solomon (US 2009/0066366 A1; Mar. 12, 2009).
Non-Final Act. 11–12.
Claims 2, 7, 12, and 17 are rejected under 35 U.S.C. § 103(a) as
obvious over one or more of: Usadi, Daems, Wang, and Solomon in view of
one or more of: Patel (US 2009/0317470 A1; Dec. 24, 2009), Lavallee (US
2002/096478 A1; July 25, 2002), Van Endert (US 2008/0205210 A1; Aug.
28, 2008), Seefeldt (US 2009/0076247 A1; Mar. 19, 2009), and Schuppert
(US 2005/0197986 A1; Sep. 8, 2005). Non-Final Act. 13–14.
We affirm.

STATEMENT OF THE CASE
Appellants’ invention relates to “a method of optimizing the electrical
operation of an integrated circuit design.” Spec. 1:7–8. Claim 1 is illustrative
and reproduced below with the key disputed limitation emphasized:
1. An automated method of optimizing a system design,
comprising:
Appeal 2014-007656
Application 12/133,480

3
receiving a description for the system design which
includes a plurality of related system components and
characteristic variables assigned to the system components, by
executing first program instructions in a computer system;
computing at least a first solution for the characteristic
variables using a statistical analysis, by executing second
program instructions in the computer system; and
computing at least a second solution for the characteristic
variables using an evolutionary analysis seeded by the first
solution, by executing third program instructions in the computer
system.

ANALYSIS
THE ANTICIPATION REJECTION OF CLAIMS 1, 3–6, 8–11, 13–16, AND 18–20 BY
USADI
The Examiner finds Usadi discloses all limitations of claim 1. Non-
Final Act. 4–6.
Appellants present the following principal argument: “Usadi mentions
statistical algorithms in paragraph 53 but does not relate them to any
evolutionary analysis. Usadi also discusses design of experiments (DOE), a
form of statistical analysis, but it is presented as an exclusive alternative to
evolutionary analysis.” App. Br. 19; see Usadi ¶¶ 53, 70, 87.
In response, the Examiner explains Usadi discloses using a
combination of statistical analysis and evolutionary analysis. Ans. 4 (citing
Usadi ¶¶ 49, 70, 82, 87).
In the Reply Brief, Appellants reply:
i. “Paragraph 49 states that the intelligent performance assistant
(IPA) may incorporate different techniques, but these techniques are used in
the alternative, not in combination.” Reply. Br. 3.
Appeal 2014-007656
Application 12/133,480

4
ii. “What paragraph 70 says is, if DOE/RSM is used then the
number n of presets available in the template is determined by the factory,
but if instead a genetic algorithm is used then this number n is
indeterminate.” Reply Br. 4.
iii. “Paragraph 82 indicates that the IPA may be provided with an
indication on when to restart a genetic search, but there is nothing in
paragraph 82 that suggest[s] the restarted genetic search would be seeded
with a solution from DOE or any other [] statistical analysis.” Reply Br. 4.
iv. “Paragraph 87 only serves to further make clear that the DOE
and genetic approaches of Usadi are mutually exclusive.” Reply. Br. 4.
For the reasons given by Appellants in the Appeal Brief and Reply
Brief, we are persuaded that the Examiner erred in finding that Usadi
discloses the recited (claim 1) (emphasis added) “computing at least a
second solution for the characteristic variables using an evolutionary
analysis seeded by the first solution.”
In the rejection, regarding seeding an evolutionary algorithm, the
Examiner refers to Usadi’s simulator using previously discovered
parameters. Non-Final Act. 5. However, we do not agree with the
Examiner’s position. Rather, we are persuaded by Appellants’ arguments in
the Appeal Brief. To the extent Usadi uses an evolutionary algorithm, the
evolutionary algorithm determines the parameters, and we see no
explanation of how or why the evolutionary algorithm would have been
seeded with results from a statistical analysis. See Usadi ¶¶ 53 (discussing
statistical algorithms without relating them to evolutionary analysis), 70 and
87 (discussing genetic algorithm methods as an alternative to statistical
algorithms).
Appeal 2014-007656
Application 12/133,480

5
Regarding the Examiner’s further explanation in the Examiner’s
Answer (Ans. 4), we still do not agree with the Examiner’s position. Rather,
we are persuaded by Appellants’ further arguments in the Reply Brief.
Usadi (¶ 49) discloses Usadi’s Intelligent Performance Assistant (IPA)
may use a genetic search technique to determine optimal parameters;
however, genetic search is presented as an alternative to statistical analysis.
We see no explanation of how or why the genetic algorithm would have
been seeded with results from a statistical analysis.
Usadi (¶ 70) discloses a genetic algorithm method as an alternative to
statistical analysis. Usadi (¶ 82) discloses genetic algorithms as an
alternative. Usadi (¶ 87) discloses a genetic algorithm framework as an
alternative. We see no explanation of how or why the genetic algorithm
would have been seeded with results from a statistical analysis.
We, therefore, do not sustain the Examiner’s rejection based on Usadi
of claim 1, or of claims 3–5, which depend from claim 1.
We also do not sustain the Examiner’s rejection based on Usadi of
independent claim 6, which recites “computing at least a second solution for
the characteristic variables using an evolutionary analysis seeded by the first
solution,” or of claims 8–10, which depend from claim 6.
We also do not sustain the Examiner’s rejection based on Usadi of
independent claim 11, which recites “computing at least a second solution
for the characteristic variables using an evolutionary analysis seeded by the
first solution,” or of claims 13–15, which depend from claim 11.
We also do not sustain the Examiner’s rejection based on Usadi of
independent claim 16, which recites “computing at least a second solution
Appeal 2014-007656
Application 12/133,480

6
for the characteristic variables using an evolutionary analysis seeded by the
first solution,” or of claims 18–20, which depend from claim 16.

THE ANTICIPATION REJECTION OF CLAIMS 1, 3–6, 8–11, 13–16, AND 18–20 BY
DAEMS
The Examiner finds Daems discloses all limitations of claim 1. Non-
Final Act. 7–9.
Appellants present the following principal argument: “While Daems
separately discusses statistical analysis and evolutionary analysis, it never
links these two analyses as recited in Appellants’ claims, not only the
combination, but also the sequence.” App. Br. 18; see Daems ¶¶ 18, 183.
In response, the Examiner explains Daems discloses:
The claims do not recite any specific link and/or specific
sequence between statistics and evolution except using
statistic[s] and evolution as a first solution to
compute/explore/search for a second solution based on the first
solution.
On page 14, left column, table 3 and notes [1]-[4], the prior
art clearly disclose[s] a link/combination/sequence between
statistical analysis and evolution/genetic analysis.
Ans. 5.
In the Reply Brief, Appellants reply:
Table 3 clearly describes a genetic algorithm—section 3 of the
pseudo-code describes how an “offspring” set (Si,offspring) is
created from a “parents” set (Si,parents) using genetic operators
(see step 3.4). The initial solution for this algorithm is seeded in
step 2.1, but there is no explanation of how the seeding occurs
other than a reference to the Deb paper. In other words, the
genetic algorithm of table 3 is not seeded by an earlier solution
from a statistical analysis, as required by Appellants’ claims.
Reply Br. 5.
Appeal 2014-007656
Application 12/133,480

7
For the reasons given by Appellants in the Appeal Brief and Reply
Brief, we are persuaded that the Examiner erred in finding that Daems
discloses the recited (claim 1) (emphasis added) “computing at least a
second solution for the characteristic variables using an evolutionary
analysis seeded by the first solution.”
To the extent Daems discloses a genetic algorithm, we see no
explanation of how or why the genetic algorithm would have been seeded
with results from a statistical analysis. See Daems ¶¶ 18 (discussing
statistical techniques, but not discussing genetic techniques), 183 (discussing
evolutionary algorithms, but not discussing seeding thereof); page 14, table
3 and notes [1]–[4] (discussing a genetic algorithm, but not discussing
seeding thereof). In short, we see no explanation of how or why the genetic
algorithm would have been seeded with results from a statistical analysis.
We, therefore, do not sustain the Examiner’s rejection based on
Daems of claim 1, or of claims 3–5, which depend from claim 1.
We also do not sustain the Examiner’s rejection based on Daems of
independent claim 6, which recites “computing at least a second solution for
the characteristic variables using an evolutionary analysis seeded by the first
solution,” or of claims 8–10, which depend from claim 6.
We also do not sustain the Examiner’s rejection based on Daems of
independent claim 11, which recites “computing at least a second solution
for the characteristic variables using an evolutionary analysis seeded by the
first solution,” or of claims 13–15, which depend from claim 11.
We also do not sustain the Examiner’s rejection based on Daems of
independent claim 16, which recites “computing at least a second solution
Appeal 2014-007656
Application 12/133,480

8
for the characteristic variables using an evolutionary analysis seeded by the
first solution,” or of claims 18–20, which depend from claim 16.

THE ANTICIPATION REJECTION OF CLAIMS 1, 3–6, 8–11, 13–16, AND 18–20 BY
WANG
The Examiner finds Wang discloses all limitations of claim 1. Non-
Final Act. 10–11.
Appellants present the following principal argument: Wang does not
seed an evolutionary analysis with a statistical solution. See App. Br. 15–17.
In response, the Examiner explains Wang discloses:
Compared with the conventional prior art random simulation
based techniques, the GA-based (genetic algorithm based)
power dissipation optimization process can generate tighter
lower bounds within shorter time. In so doing, the GA[-]based
optimization process of [the] present invention is capable of
vigorously optimizing a complex integrated circuit design for
power dissipation within the reasonable time and resource
constraints of modern computer implemented EDA design
synthesis processes.
Ans. 6.
The Examiner further explains:
Regarding the statistics and evolution genetic in the
claims, col 16 lines 10-60 in Wang discloses combination of
statistical analysis (i.e., stochastic processes and/or probability)
and Genetic algorithm (GA) processing evolution process as first
solution to find/compute/explore/search for a second solution
([i].e., optimized power) from the first solution[.]
Ans. 7.
In the Reply Brief, Appellants reply:
There is no identification of any specific element of Wang which
ostensibly corresponds to the “first solution” from a statistical
Appeal 2014-007656
Application 12/133,480

9
analysis which is used to seed a subsequent evolutionary
analysis. The Examiner continues to refer generally to the
“stochastic processes” and “probability” mentioned at column
16, lines 10-60, but does not explain how those supposedly relate
to the genetic algorithm. Appellants pointed out in their Appeal
Brief that this portion of Wang is comparing a prior art random
simulation technique to the genetic algorithm which
distinguishes the two techniques as alternatives, not using them
in combination, but the Examiner has provided no rebuttal to this
point.
Reply Br. 6.
For the reasons given by Appellants in the Appeal Brief and Reply
Brief, we are persuaded that the Examiner erred in finding that Wang
discloses the recited (claim 1) (emphasis added) “computing at least a
second solution for the characteristic variables using an evolutionary
analysis seeded by the first solution.”
To the extent Wang discloses a genetic algorithm, we see no
explanation of how or why the genetic algorithm would have been seeded
with results from a statistical analysis. See Wang col. 7 (discussing a genetic
algorithm for power optimization and circuit optimization generally, but not
discussing seeding of the genetic algorithm), 10–11 (discussing genetic
algorithms generally, and discussing random seeding, but not discussing
seeding using a statistical analysis), 15–16 (discussing the genetic algorithm,
and discussing a comparison of genetic algorithm techniques and random
simulation techniques, but not discussing the seeding of the genetic
algorithm). In short, we see no explanation of how or why the genetic
algorithm would have been seeded with results from a statistical analysis.
We, therefore, do not sustain the Examiner’s rejection based on Wang
of claim 1, or of claims 3–5, which depend from claim 1.
Appeal 2014-007656
Application 12/133,480

10
We also do not sustain the Examiner’s rejection based on Wang of
independent claim 6, which recites “computing at least a second solution for
the characteristic variables using an evolutionary analysis seeded by the first
solution,” or of claims 8–10, which depend from claim 6.
We also do not sustain the Examiner’s rejection based on Wang of
independent claim 11, which recites “computing at least a second solution
for the characteristic variables using an evolutionary analysis seeded by the
first solution,” or of claims 13–15, which depend from claim 11.
We also do not sustain the Examiner’s rejection based on Wang of
independent claim 16, which recites “computing at least a second solution
for the characteristic variables using an evolutionary analysis seeded by the
first solution,” or of claims 18–20, which depend from claim 16.

THE ANTICIPATION REJECTION OF CLAIMS 1, 3–6, 8–11, 13–16, AND 18–20 BY
SOLOMON
The Examiner finds Solomon discloses all limitations of claim 1.
Non-Final Act. 11–12.
Appellants present the following principal argument:
Solomon describes the use of evolutionary computing
techniques (such as hybrid GA in paragraph 131) but never
mentions the use of a statistical analysis to create a first solution.
In this regard, the Office Action suggests that the “stochastic
probabilities” mentioned only in paragraph 48 of Solomon
correspond to the claimed statistical analysis. However,
stochastic probabilities are not a statistical analysis as that term
is used in the art.
App. Br. 21. “Solomon never describes how any evolutionary analysis is
seeded. Solomon mentions genetic algorithms (and other evolutionary
Appeal 2014-007656
Application 12/133,480

11
analyses) only in paragraphs 44 and 115, but never says how those
algorithms are initialized.” App. Br. 21.
In response, the Examiner explains “Solomon clearly teach[es]
combination of evolutionary and statistic (stochastic probabilities) as a first
solution to evolve to solve specific application problems (solving specific
application problems is a second solution)[.]” Ans. 7 (citing Soloman ¶¶ 48,
131).
In the Reply Brief, Appellants reply: “Nothing in paragraph 48 or any
other part of Solomon ever says that the stochastic probabilities seed or
otherwise initialize a genetic algorithm. Solomon mentions genetic
algorithms (and other evolutionary analyses) only in paragraphs 44 and 115,
but never says how those algorithms are initialized.” Reply Br. 7.
We are persuaded by the Examiner’s finding that Solomon discloses
the recited (claim 1) (emphasis added) “computing at least a second solution
for the characteristic variables using an evolutionary analysis seeded by the
first solution.”
Solomon (¶¶ 44, 115, 131) discloses metaheuristics including genetic
algorithms and evolving optimization problems.
Solomon (¶ 48) discloses “[s]ince the 3D SoC is used in uncertain
environments, it uses evolutionary mechanisms described herein to change
its hardware configurations.”
Solomon (¶ 48) further discloses:
in indeterministic situations, hybrid metaheuristic models are
applied to solve complex eMOOPs ([multi-objective
optimization problems]) in real time. The system creates
modeling simulations of the indeterministic environment by
developing alternative scenarios based on stochastic
probabilities. These anticipations of experience develop into
Appeal 2014-007656
Application 12/133,480

12
scenario trajectories, which are then used to model
reconfigurable transformations of the FPGA layer positions.
With this approach, the chip will experiment with providing and
testing solutions to complex optimization problems. IP cores are
evolved and applied to solve specific application problems.
First, we agree with the Examiner that Solomon’s developing
alternative scenarios based on stochastic probabilities constitutes a first
solution computed using a statistical analysis because developing alternative
scenarios based on stochastic probabilities involves the interpretation of
data. See Ans. 7, Solomon ¶ 48.
This finding is consistent with the plain meaning of “statistics,” which
in the pertinent sense is “1 : a branch of mathematics dealing with the
collection, analysis, interpretation, and presentation of masses of numerical
data.” MERRIAM-WEBSTER’S COLLEGIATE DICTIONARY 1149 (10th ed. 1997).
Further, Appellants’ Specification (Spec. 3:16–17) discloses: “A
variety of statistical approaches have also been employed for circuit
simulation, including orthogonal arrays, design of experiments, and central
composite design”; and Appellants’ Specification (Spec. 14:26–15:1) further
discloses: “For example, other statistical analysis techniques may be used to
seed the evolution computation besides CCD, such as orthogonal arrays or
design of experiment.” However, such permissive language does not
preclude the Examiner’s finding.
Second, we agree with the Examiner that Solomon’s stochastic
probabilities are used to seed an evolutionary analysis because Solomon’s
alternative scenarios seed the scenarios trajectories, which model
transformations. The scenario trajectories are reasonably described as an
evolutionary analysis because solutions are tested and IP cores are evolved
to solve problems. See Ans. 7, Solomon ¶ 48; see also Solomon Fig. 32.
Appeal 2014-007656
Application 12/133,480

13
We, therefore, sustain the Examiner’s rejection based on Solomon of
claim 1, as well as claims 4–6, 9–11, 13–16, 19, and 20, which are not
separately argued with particularity.
Regarding claims 3, 8, 13, and 18, we find Appellants’ arguments (see
App. Br. 21–22, Reply Br. 8) not persuasive. Solomon discloses
metaheuristics including genetic algorithms. Thus, Solomon discloses the
evolutionary analysis being a genetic algorithm. See Non-Final Act. 12
(citing Solomon ¶¶ 115, 131); see also Ans. 7.
We, therefore, also sustain the Examiner’s rejection based on
Solomon of claims 3, 8, 13, and 18.

THE OBVIOUSNESS REJECTION OF CLAIMS 2, 7, 12, AND 17 OVER USADI,
DAEMS, WANG, AND SOLOMON IN VIEW OF ONE OR MORE OF: PATEL,
LAVALLEE, VAN ENDERT, SEEFELDT, AND SCHUPPERT
Appellants argue:
The prior art is vacant of any teaching or suggestion that a central
composite design could be used to seed any other type of
simulation analysis, so the proposed combinations accordingly
cannot render the present invention unpatentable, and the Office
Action fails to make a prima facie case of obviousness.
App. Br. 22; see also Reply Br. 8.
Regarding claims 2, 7, 12, and 17, we find Appellants’ arguments (see
App. Br. 22, Reply Br. 8) not persuasive. The Examiner explained multiple
reasons why one of ordinary skill in the art would have combined the
references. See Non-Final Act. 13–14; see also Ans. 8–9. Further, the
various secondary references are relied on for teachings of Central
Composite Design (CCD). See Non Final Act. 13–14. As discussed above,
Appeal 2014-007656
Application 12/133,480

14
Solomon already discloses using a statistical analysis to seed an evolutionary
analysis. See Solomon ¶ 48.
We, therefore, sustain the Examiner’s rejections of claims 2, 7, 12,
and 17.

DECISION
Because we have affirmed at least one ground of rejection with
respect to each claim on appeal, the Examiner’s decision rejecting claims 1–
20 is affirmed. See 37 C.F.R. § 41.50(a)(1).
No time period for taking any subsequent action in connection with
this appeal may be extended under 37 C.F.R. § 1.136(a)(1).

AFFIRMED