Ex Parte Carlson et al

Patent Trial and Appeal Board

Aug 29, 2016

13087558 (P.T.A.B. Aug. 29, 2016)

UNITED STA TES p A TENT AND TRADEMARK OFFICE
APPLICATION NO. FILING DATE
13/087,558 04/15/2011
130683 7590 08/31/2016
Middleton Reutlinger (IBMP)
401 S. 4th Street, Suite 2600
Louisville, KY 40202
FIRST NAMED INVENTOR
David G. Carlson
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.
ROC920100026US 1 6494
EXAMINER
MA, WING
ART UNIT PAPER NUMBER
2456
NOTIFICATION DATE DELIVERY MODE
08/31/2016 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):
usptomail@middletonlaw.com
PTOL-90A (Rev. 04/07)
UNITED STATES PATENT AND TRADEMARK OFFICE
BEFORE THE PATENT TRIAL AND APPEAL BOARD
Ex parte DAVID G. CARLSON,
JEFFREY S. MCALLISTER, and
NELSON RAMIREZ1
Appeal2015-000940
Application 13/087,558
Technology Center 2400
Before MICHAEL J. STRAUSS, MICHAEL M. BARRY, and
AARON W. MOORE, Administrative Patent Judges.
BARRY, Administrative Patent Judge.
DECISION ON APPEAL
Appellants appeal under 35 U.S.C. § 134(a) from a Final Rejection of
claims 1-23. We have jurisdiction under 35 U.S.C. § 6(b).
We REVERSE.
1 Appellants identify the real party in interest as International Business
Machines Corp. App. Br. 1.
Appeal2015-000940
Application 13/087,558
Introduction
Appellants state their invention relates "particularly to executing tasks
in a parallel computing system." Spec. i-f 1. Claim 1 is representative:
1. A method of streaming data between a host element and
a target element in a parallel computing system, the method
comprising, in the host element:
partitioning data used to perform a task into a plurality of
data streams based at least in part upon data associated with the
task;
partitioning a data stream from the plurality of data
streams into a plurality of data stream views based at least in
part on a memory size associated with the target element
configured to process the data stream, wherein partitioning the
data stream based at least in part on the memory size associated
with the target element includes partitioning the data stream
based on an amount of memory available in the target
element for processing data stream views; and
sequentially communicating at least a portion of the
plurality of data stream views to the target element via a data
streaming infrastn1cture implemented on at least the host
element and the target element, wherein a first portion of the
data streaming infrastructure configured on the host element
transfers the portion of the plurality of data stream views to a
second portion of the data streaming infrastructure configured
on the target element.
App. Br., Claims App'x 1 ( dispositive limitation highlighted).2
Rejections
Claims 1-12, 14--20, 22, and 23 stand rejected under 35 U.S.C.
§ 103(a) as obvious over Liddell (US 2011/0208872 Al; Aug. 25, 2011),
2 Appellants' Claims Appendix at the end of the Brief does not include page
numbers; we refer to its pages as if numbered consecutively starting with 1.
2
Appeal2015-000940
Application 13/087,558
Lakshmanan (US 2011/0047555 Al; Feb. 24, 2011), and Amini (US 2008/
0005392 Al; Jan. 3, 2008). Final Act. 5-13.
Claims 13 and 21 stand rejected under 35 U.S.C. § 103(a) as obvious
over Liddell, Lakshmanan, Amini, and Zeis (US 8,07 4,043 B 1; Dec 6,
2011). Final Act. 13-14.
ISSUE3
The dispositive issue is whether the Examiner errs in finding that
Amini would have taught or suggested to one of ordinary skill the
requirement of "partitioning the data stream based on an amount of memory
available in the target element for processing data stream views," as recited
in claims 1 and 23 and as commensurately recited in claim 1 7.
ANALYSIS
"The Examiner relies upon Amini to teach [the disputed] limitation."
Ans. 15-16. In particular "[t]he Examiner finds Amini teaches the disputed
limitation as follows:
Amini teaches partitioning the data stream (i.e.,
adjusting the amount of SDOs [stream data objects]
communicated per second) based at least in part on a
memory size associated with the target element (as shown by
if0083 adjusting the rate/amount of SDOs communicated is
based on the buffer occupancy levels associated with a target
element) includes an amount of memory available in the
target element for processing data stream views. ([A]s
shown by if0083 adjusting the rate/amount of SDOs
communicated is based on the buff er occupancy levels
associated with a target element. Buff er occupancy levels
disclose an amount of memory available for processing SDOs.)
3 Because the issue we address below is dispositive, we do not address other
arguments raised by Appellants.
3
Appeal2015-000940
Application 13/087,558
Ans. 19; see also Final Act. 7 (citing Amini iii! 5, 62, 81, 91, Fig. 8).
The Examiner also finds that Amini's "output rate (e.g., outputting 30
Stream Data Objects [SDOs] per second) is a size of data being
communicated between processing elements." Ans. 17 (citing Amini
iii! 16, 62) (brackets in original).
Appellants reply by arguing, inter alia,4 that "rate" and "size" are not
the same (Reply Br. 3), and that Amini "focuses on adjusting input, output
and processing rates to 'keep input buffers near target levels"' (Reply Br. 5
(quoting Amini if 83)). Appellants contend "controlling a rate is not
analogous to controlling a size" and that the Examiner errs in "find[ing]
correspondence between changing the rate at which SDO' s/packets are
communicated or processed and how a data stream is 'partitioned' (i.e.,
divided into parts or portions). Reply Br. 5.
We agree with Appellants that Amini' s disclosure of adjusting the rate
of communicating stream packets or objects neither teaches nor suggests
"partitioning the data stream based on an amount of memory available in the
target element for processing data stream views" as required by the
independent claims. Amini does not teach partitioning a data stream based
on the amount of memory available in buffers; rather, it teaches changing the
rate of the data stream based on multiple considerations, including buffer
status. See Amini iii! 44, 58, 83. While the buffers in Amini's processing
elements can constitute "an amount of memory available in the target
element" and such buffers can be used "for processing data stream views,"
Appellants persuade us that controlling the rate of distribution of a data
4 See supra note 3.
4
Appeal2015-000940
Application 13/087,558
stream is not analogous to partitioning the data stream based on available
memory, as required by the dispositive limitation.
We accordingly do not sustain the rejection of the independent claims,
1, 17, and 23. We thus also do not sustain the rejection of the dependent
claims, 2-16 and 18-22.
DECISION
For the above reasons, the Examiner's rejection of claims 1-23 is
reversed.
REVERSED
5