null

Patent Trials and Appeals Board

Nov 18, 2019

14165326 - (D) (P.T.A.B. Nov. 18, 2019)

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.
14/165,326 01/27/2014 Charlie Jing 2013EM027 5780
34477 7590 11/18/2019
ExxonMobil Upstream Research Company
22777 Springwoods Village Parkway
(EMHC-N1.4A.607)
Spring, TX 77389
EXAMINER
GEBRESILASSIE, KIBROM K
ART UNIT PAPER NUMBER
2129
NOTIFICATION DATE DELIVERY MODE
11/18/2019 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):
urc-mail-formalities@exxonmobil.com
PTOL-90A (Rev. 04/07)
UNITED STATES PATENT AND TRADEMARK OFFICE
BEFORE THE PATENT TRIAL AND APPEAL BOARD
Ex parte CHARLIE JING and DENNIS E. WILLEN
Appeal 2019-000934
Application 14/165,326
Technology Center 2100
Before MIRIAM L. QUINN, DANIEL N. FISHMAN, and
JON M. JURGOVAN, Administrative Patent Judges.
QUINN, Administrative Patent Judge.
DECISION ON APPEAL
STATEMENT OF THE CASE
Pursuant to 35 U.S.C. § 134(a), Appellant1 appeals from the
Examiner’s decision to reject claims 1–20. See Final Act. 1. We have
jurisdiction under 35 U.S.C. § 6(b).
We REVERSE.

1 We use the word Appellant to refer to “applicant” as defined in 37 C.F.R.
§ 1.42. Appellant identifies the real party in interest as ExxonMobil
Upstream Research Company. Appeal Br. 2.
Appeal 2019-000934
Application 14/165,326

2
CLAIMED SUBJECT MATTER
The claims are directed to an inversion of geophysical data on
computer system having parallel processors. Claim 1, reproduced below, is
illustrative of the claimed subject matter:
l. A method for efficient use of a computing system of
parallel processors to update a geophysical model, the method
comprising:
assigning at least one control processor to control
sequence of operations and to reduce load imbalance between
the processors, and
assigning a first dedicated processor group of one or
more processors to update one or more model parameters in the
geophysical model;
assigning a second dedicated processor group of one or
more processors to simulate forward modeling simulated data;
and
performing an iterative geophysical model optimization,
wherein each iteration comprises:
generating simulated measurement data from the
geophysical model with the second dedicated processor group;
comparing the simulated measurement data to measured
data; and
updating, with the first dedicated processor group, one or
more model parameters of the geophysical model to lessen
misfit based on the comparing,
wherein the at least one control processor dynamically
balances processing load resulting from the iterative
geophysical model optimization across the first dedicated
processor group and the second dedicated processor group.
Appeal Br. 12 (Claim Appendix).
Appeal 2019-000934
Application 14/165,326

3
REFERENCES
The prior art relied upon by the Examiner is:
Name Reference Date
Wang US 5,991,695 Nov. 23, 1999
Carazzone US 7,808,420 B2 Oct. 5, 2010
REJECTION
Claims 1–20 stand rejected under 35 U.S.C. § 103(a) as being
unpatentable over Wang and Carazzone.
OPINION
This appeal presents one dispositive issue: whether the Examiner
erred in determining that the combination of Wang and Carazzone teaches “a
first dedicated processor group” and “a second dedicated processor group,”
as recited in claim 1, and similarly recited in claims 13 and 14 (argued
together with claim 1). We therefore decide the appeal for this rejection
based on claim 1. See 37 C.F.R. § 41.37(c)(1)(iv).
Having reviewed the Examiner’s rejections in light of Appellant’s
arguments, as presented in the Appeal Brief and Reply Brief, we are
persuaded that the Examiner has erred as Appellant alleges.
Claim 1 requires two groups of processors, each group dedicated to
performing a particular function. For instance, the “first dedicated processor
group” is assigned “to update one or more model parameters in the
geophysical model” (hereinafter referred to as the “updating task”) and the
“second dedicated processor group” is assigned “to simulate forward
modeling [of] simulated data” (hereinafter referred to as the “simulation
task”). These groups of processors are “dedicated” to performing the recited
functions during an interactive process. For example, the claim requires that
Appeal 2019-000934
Application 14/165,326

4
the second dedicated group of processors group perform the simulation
tasks, and then the first dedicated group of processors perform the updating
task based on a comparison of the output of the simulation with measured
data. Appellant argues that the combination of Wang and Carazzone does
not teach these groups of processors dedicated as recited to perform the
required tasks. We are persuaded by Appellant’s argument.
The Examiner relies on Wang, either alone or in combination with
Carazzone, as teaching the two groups of processor. We discuss each of
these positions below.
Analysis of Wang, Alone
The Examiner finds that Wang alone teaches the two processor
groups. For instance, the Examiner states that naming the groups as first or
second “show[s] that the two groups [of processors] are performing different
tasks independently.” Ans. 3. The Examiner therefore finds that Wang
discloses “multiple groups [of processors] . . . performing at least an initial
simulation assigned to a second group of processors . . . as claimed.” Id.
(internal citations omitted). According to the Examiner, Wang teaches the
second dedicated group of processors, which is assigned to perform the
simulation tasks, and also teaches a first dedicated group of processors to
update a model. Id. We do not agree.
Wang describes an initial simulation process at column 9, lines 2−4
and 18−20 (cited by the Examiner). The initial simulation process, as
Appellant correctly notes, is performed in one of Wang’s processor groups
and for the purpose of determining how many processors in each group are
needed to carry out the actual traveltime calculations. Reply Br. 3 (citing
Wang, 9:10−18, 25−32). This description of the initial simulation in column
Appeal 2019-000934
Application 14/165,326

5
9 yields two factual determinations regarding Wang’s assignment of groups
of processors. First, Wang’s groups of processors, which are not assigned to
the initial simulation, perform no other task. Second, when Wang completes
the initial simulation, all of Wang’s groups of processors perform traveltime
calculations, i.e., simulation tasks. See Ans. 4 (stating that Wang teaches at
one point assigning the second group of processors for simulation tasks,
pointing to Wang’s disclosure of the initial simulation process which
performs the traveltime calculations). Notwithstanding these facts, the
Examiner states that “it appears Wang . . . also discloses updated one or
more model parameters” because Wang describes performing a Schneider
process. Id. at 3−4 (citing Wang, 11:14−16). The Examiner, however, does
not explain which or how any processor group, separate and distinct from
the second processor group performing the simulation tasks, would perform
in Wang the updating task alleged to be performed by executing the
Schneider process. That is, even if Wang suggests performing an updating
task, which is not entirely clear from the Examiner’s explanation of Wang,
all the Wang processors are actually programmed to perform the simulation,
i.e., traveltime calculation (using the Schneider process). See Wang,
11:14−16 (“The process of calculating traveltimes in each [processing
element] is discussed with reference to FIG. 3, which follows the steps given
by Schneider.”).
The Examiner has not shown that Wang assigns to any other
processor group the updating tasks. And, as Appellant correctly notes,
Wang does not teach or suggest dividing different tasks among different
processor groups. Appeal Br. 7; Reply Br. 3. Therefore, we are persuaded
by Appellant’s argument that Wang alone does not teach or suggest the “first
Appeal 2019-000934
Application 14/165,326

6
dedicated processor group” and the “second dedicated processor group” as
claimed.
Analysis of Wang in Combination with Carazzone
The Examiner states that “for clarity,” he relies on Carazzone in
combination with Wang. Ans. 4. In particular, the Examiner states that
Carazzone discloses assigning a dedicated group of processors, i.e., M
processors shown in Figure 1, item 13 of Carazzone, as performing the
updating task because Carazzone describes adjusting parameters to reduce
misfit between data. Id. (citing Carazzone, Fig. 1, 7:24−31, 8:22−24). The
Examiner further states that Wang and Carazzone are “analogous arts that
teach performing computations and organizing operations on a system of
parallel processors.” Id. Therefore, the Examiner finds that the first and
second groups of processors are “general purpose processors as no details of
specific processors are claimed,” that Wang “teaches at one point assigning
[a] second group of processors for simulation tasks,” and that “it would have
been obvious to use the other (first) group[] of processors for updating the
parameters as shown by the second reference (Carazzone et al) . . . .” Id.
(citations omitted). Thus, according to these findings, Carazzone contributes
the teaching of using a group of processors for updating tasks, which would
have been obvious to assign to a group of processors in Wang. This
combination of teachings, however, is erroneous.
As Appellant argues, and we agree, “Wang makes no such reservation
of dedicated processor groups to tasks other than traveltime calculations.”
Reply Br. 4. Both Wang and Carazzone describe groups of processors
performing certain tasks, but both Wang and Carazzone assign the same task
to the groups of processors. As stated above, Wang assigns the traveltime
Appeal 2019-000934
Application 14/165,326

7
calculations to all of its groups of processors, notwithstanding that it
performs an initial simulation of the traveltime calculation for the purpose of
determining how many processors to assign to each group. Wang, 9:1−32.
Carazzone is no different. Each of Carazzone’s groups of processors,
referred to as a “bank” of processors, performs Maxwell’s equations for a
particular data slice. Carazzone, 7:24−27 (“This requires N banks with M
processors per bank, which must be arrayed physically and provided with
communication, each bank being programmed to solve Maxwell’s
equations.”). Thus, we agree with Appellant that there is no dividing of
tasks among Carazzone’s groups of processors because Carazzone teaches
that each bank performs the same simulation task, albeit on a particular data
slice, and computes the total error misfit contribution for the data slice.
Reply Br. 4; Carazzone, 7:51−54.
With neither Wang nor Carazzone teaching or suggesting two groups
of processors each performing a different task, a person of ordinary skill in
the art would not have reason to either modify Wang or Carazzone to
perform the assignment of tasks to different processor groups or to combine
the teachings to result in the claimed groups of processors. It is not enough
to state that the processors are general purpose processors. Ans. 4. Such a
statement fails to show why, having two references where all the processor
groups are assigned the same calculations, a person of ordinary skill in the
art would dedicate one group of processors to an assigned task of performing
simulations, while another group of processors is dedicated to updating the
model after the simulation reveals a misfit. Indeed, Carazzone’s teachings
support the contrary, because each of Carazzone’s groups of processors
Appeal 2019-000934
Application 14/165,326

8
performs the simulation and also the misfit calculation. See Carazzone,
7:51−54.
Accordingly, we are persuaded that the Examiner erred in finding that
Wang, alone, or in combination with Carazzone, teaches or suggests the
recited “first dedicated processor group” and “second dedicated processor
group,” as recited.
CONCLUSION
For the above-stated reasons, we do not sustain the Examiner’s
rejection of claims 1−20 over Wang and Carazzone.
DECISION SUMMARY
Claims
Rejected
35 U.S.C. § Reference(s)/Basis Affirmed Reversed
1−20 103(a) Wang and Carazzone 1−20

REVERSED