Ex Parte Hammarlund et al

Board of Patent Appeals and Interferences

Jun 15, 2010

10749271 (B.P.A.I. Jun. 15, 2010)

UNITED STATES PATENT AND TRADEMARK OFFICE
____________________

BEFORE THE BOARD OF PATENT APPEALS
AND INTERFERENCES
____________________

Ex parte PER H. HAMMARLUND and STEPHAN J. JOURDAN
____________________

Appeal 2009-005596
Application 10/749,2711
Technology Center 2100
____________________

Decided: June 16, 2010
____________________

Before JOHN A. JEFFERY, JEAN R. HOMERE, and JAMES R. HUGHES,
Administrative Patent Judges.

HUGHES, Administrative Patent Judge.

DECISION ON APPEAL

1 Application filed December 30, 2003. The real party in interest is Intel
Corporation. (Br. 2.)
Appeal 2009-005596
Application 10/749,271

2
STATEMENT OF THE CASE
Appellants appeal from the Examiner’s rejection of claims 1-30 under
authority of 35 U.S.C. § 134(a). The Board of Patent Appeals and
Interferences (BPAI) has jurisdiction under 35 U.S.C. § 6(b).
We affirm.

Appellants’ Invention
Appellants invented an adaptive replay system that replays computer
processing instructions that are not properly executed when initially
dispatched to an execution unit. (Spec. ¶ [0018].) The adaptive replay
system or loop efficiently allows the instructions to be executed again by
allowing instructions to change position in the replay loop, decreasing
instruction execution. (Spec. ¶ [0022].)
The system includes an execution unit, a staging unit that forwards an
instruction through a replay loop operating in parallel to the execution unit,
a selector (MUX), coupled to the staging unit, that places the instruction in
an optimal position within the replay loop, and a scoreboard, coupled to the
selector, that stores status information for the instruction. (Spec. ¶¶ [0029],
[0030], [0037]-[0039].)2

2 We refer to Appellants’ Specification (“Spec.”); Appeal Brief (Response
to Non-Compliant Appeal Brief) (“App. Br.”) filed May 28, 2008; and
Reply Brief (“Reply Br.”) filed October 6, 2008. We also refer to the
Examiner’s Answer (“Ans.”) mailed August 4, 2008.
Appeal 2009-005596
Application 10/749,271

3
Representative Claim
Independent claim 1 further illustrates the invention. It reads as
follows:
1. An adaptive replay system comprising:
a staging unit to forward an instruction in a replay loop
parallel to an execution unit;
a selector device coupled to said staging area to place
said instruction in an optimal position within said replay loop;
and
a scoreboard coupled to said selector device to store
status information for said instruction.

References
The Examiner relies on the following references as evidence of
unpatentability:
Merchant US 6,163,838 Dec. 19, 2000

Merchant US 6,385,715 B1 May 7, 2002

Topham US 6,944,853 B2 Sep. 13, 2005
(filed May 22, 2001)

Rejections on Appeal
The Examiner rejects claims 1, 2, 4, and 7 under 35 U.S.C. § 102(b)
as anticipated by Merchant (US 6,385,715 B1 (“Merchant”)).
The Examiner rejects claims 3, 9-15, 18-23, 25-27, 29, and 30 under
35 U.S.C. § 103(a) as being unpatentable over Merchant.
The Examiner rejects claims 5, 6, 16, and 17 under 35 U.S.C. § 103(a)
as being unpatentable over Merchant and Merchant (US 6,163,838
(“Merchant ‘838”)).
Appeal 2009-005596
Application 10/749,271

4
The Examiner rejects claims 8, 24, and 28 under 35 U.S.C. § 103(a) as
being unpatentable over Merchant and Topham.

Grouping of the Claims
Appellants argue independent claim 1 as representative of claims 1, 2,
4, and 7 (rejected under § 102(b)). (App. Br. 7.) Appellants do not
separately argue the remaining claim groups: Group (1) – claims 3, 9-15, 18-
23, 25-27, 29, and 30 rejected under § 103(a) over Merchant (App. Br. 8);
Group (2) – claims 5, 6, 16, and 17 under § 103(a) over Merchant and
Merchant ‘838 (App. Br. 8); and Group (3) – claims 8, 24, and 28 under
§ 103(a) over Merchant and Topham (App. Br. 9). Rather, Appellants
reiterate their arguments made with respect to independent claim 1 and the
§ 102(b) rejection. (Br. 8-9.) We accept Appellants’ grouping of the claims,
and choose independent claim 1 as representative of Appellants’ arguments
and groupings. Accordingly, we treat Appellants’ claims 2-30 as standing or
falling with representative claim 1. See 37 C.F.R. § 41.37(c)(1)(vii) (2007).

ISSUE
Based on our review of the administrative record, Appellants’
contentions, and the findings and conclusions of the Examiner, the pivotal
issue before us is as follows:
Did the Examiner err in finding the Merchant reference discloses, and
would have taught or suggested to a skilled artisan, an adaptive replay
system including a selector device coupled to a staging unit to place an
instruction in an optimal position within a replay loop?
Appeal 2009-005596
Application 10/749,271

5
FINDINGS OF FACT (FF)
Appellants’ Specification
1. Appellants do not explicitly define in their Specification a
“selector device,” or an “optimal position” of an instruction in a replay loop.
Appellants explain that:
When an instruction needs to be replayed, adaptive replay
selector MUX 425 can utilize the information stored in
scoreboard 430 to check if the instruction is at the optimal
position in the replay loop. Based on the information, adaptive
replay MUX 425 can change the instruction’s relative position
in the replay loop.
(Spec. ¶ [0039].) Appellants provide examples of such an “optimal
position,” for example, where “a Uop that is scheduled late as compared to
the Uop it depends on is allowed to move ahead within replay stages 420
closer to the Uop it depends on” (Spec. ¶ [0039]) – see Appellants’
Specification ¶ [0042], and Figures 5a & 5b.

Merchant Reference
2. Merchant describes a multi-threading processor and a multi-
stage replay system including a checker and a multiple replay queues for
“replaying” (re-executing) instructions that did not execute properly in the
first instance. (Abstract; col. 1, ll. 19-21; col. 1, l. 65 to col. 2, l. 6.)
3. Merchant’s replay system includes two staging sections
(staging queues A, B, C, & D, and staging queues E & F). Merchant’s
processor stages instructions in staging queues A, B, C, & D in parallel with
an execution unit (element 118), and supplies the instructions to a checker
(element 150) that determines if the instruction properly executed. If the
Appeal 2009-005596
Application 10/749,271

6
instruction does not execute properly, it is replayed. The instruction is
routed through a replay loop (element 156) including a controller (element
154) that may send the instruction through staging queues E & F to a
multiplexer (MUX, element 116) and then (again) through staging queues A,
B, C, & D. Alternatively, the controller may send the instruction to a replay
queue (element 170) that temporarily stores the instruction before sending it
to the MUX. (Col. 5, ll. 16-30; col. 6, ll. 7-25, 39-49; Fig. 1.)

ANALYSIS
Rejection Under § 102(b)
Appellants contend the Merchant reference does not disclose the
features of independent claim 1, specifically, “‘an adaptive replay system
comprising:… a selector device coupled to said staging area to place said
instruction in an optimal position within said replay loop...’ (e.g., as
described in claim 1).” (App. Br. 5.) The Examiner finds that the Merchant
reference discloses each feature of Appellants’ independent claim 1, and
maintains that the claim is properly rejected as anticipated by Merchant.
(Ans. 3-4, 14-16.) Specifically, the Examiner finds that Merchant’s checker
(150) and controller (154) meet the limitation of a selector device that places
an instruction in an optimal position in the replay loop. (Ans. 3-4, 15-16.)
Accordingly, we decide the question of whether the Examiner erred in
finding the Merchant reference discloses an adaptive replay system
including a selector device coupled to a staging unit to place an instruction
in an optimal position within a replay loop.
After reviewing the record on appeal, we agree with the Examiner’s
findings with respect to the Merchant reference, and we find that Merchant
Appeal 2009-005596
Application 10/749,271

7
discloses the disputed feature of Appellants’ claim 1. We begin our analysis
by construing Appellants’ claim.
We determine the scope of the claims in patent applications not solely
based on the claim language, but upon giving claims “their broadest
reasonable interpretation consistent with the [S]pecification” and “in light of
the [S]pecification as it would be interpreted by one of ordinary skill in the
art.” In re Am. Acad. of Sci. Tech Ctr., 367 F.3d 1359, 1364 (Fed. Cir. 2004)
(citations omitted).
Appellants do not explicitly define a “selector device” or an “optimal
position” in their Specification. (FF 1.) Appellants also do not provide any
limitations to the structure of the “selector device” in their claim. Rather,
Appellants define their “selector device” by the claim language in terms of
its function – placing (“to place”) “said instruction in an optimal position
within said replay loop.” (Br. 11, Claim App’x, claim 1.) Thus, we broadly
but reasonably construe Appellants’ “selector device” to simply be a module
capable of selecting instructions.
Appellants explain in their Specification that the selector device
repositions the instruction within the replay loop based on information input
to the selector device so that the instruction executes more efficiently
(quickly). (FF 1.) Appellants provide examples of “optimal” repositioning,
but do not explicitly describe the criteria utilized to determine the “optimal
position,” or the process (or algorithm) utilized by the selector device to
reposition the instruction. (FF 1.) Thus, we broadly but reasonably construe
placing the instruction in an “optimal position” in the replay loop – utilizing
undisclosed and unclaimed criteria, and undisclosed and unclaimed
processes or mechanisms – to simply mean repositioning an instruction to a
Appeal 2009-005596
Application 10/749,271

8
position within the replay loop that is subjectively determined to be more
efficient (optimal). Accordingly, we construe Appellants’ disputed
limitation to mean a module for selecting an instruction and repositioning
the instruction to a position within the replay loop that is more efficient.
We find that Merchant discloses a replay system including two
separate staging sections. The first staging section (staging queues A, B, C,
& D) supplies instructions to a checker (150) that determines if the
instructions have properly executed. If not, the checker routes the
instruction(s) through a replay loop including a controller (154), so that the
instruction is re-executed. The controller may route the instruction through
a second staging section (staging queues E & F) or through a replay queue
(170) that temporarily stores the instruction. In either case, the instruction is
then routed to MUX (116) and then sent through the first staging section
again. (FF 2, 3.)
Thus, we find Merchant discloses a replay system including a checker,
a controller, two separate staging sections, and a replay queue. We construe
Appellants’ recited “replay loop” to (broadly but reasonably) include
Merchant’s second staging section (staging queues E & F) and replay queue.
We further find, as did the Examiner (see Ans. 15-16), that Merchant’s
checker and controller route instructions through the second staging section
or through the replay queue, and therefore reposition the instructions within
the replay loop. Thus, we find Merchant discloses a module for selecting
instructions and repositioning them within the replay loop so that the
instructions execute more efficiently – i.e., repositioning the instructions to
an optimal position within the replay loop.
Appeal 2009-005596
Application 10/749,271

9
We are not persuaded by Appellants’ argument that the reference does
not disclose the disputed limitation because “making a conditional
determination upon which one of two steps may be followed is not the same
as placing an instruction in an optimal position within a replay queue (as
described in embodiments of the present application) at all.” (App. Br. 6.)
As we explained supra, Merchant discloses a selector (checker and
controller) that selects instructions and routes them through staging queues
or a replay queue. Routing an instruction through the staging queues, or
alternatively the replay queue, results in repositioning of the instruction in
the replay loop – the instruction will, depending on its path, reach the MUX
at different times. These different routes result in the instructions executing
more efficiently (see Ans. 15-16). Appellants’ recited claim limitation does
not preclude utilizing alternate paths/loops within a replay loop, and
Appellants’ do not explain how Merchant fails to disclose the recited
functionality – i.e., how “making a conditional determination” to route an
instruction along alternate paths “is not the same as placing an instruction in
an optimal position.”
We note that a prior art reference that achieves the same result in the
same manner anticipates a disputed limitation. See Continental Can Co. v.
Monsanto Co., 948 F.2d 1264, 1268-1269 (Fed. Cir. 1991) (“If, however,
the disclosure is sufficient to show that the natural result flowing from the
operation as taught would result in the performance of the questioned
function, it seems to be well settled that the disclosure should be regarded as
sufficient.”); Tate Eng’g, Inc. v. United States, 477 F.2d 1336, 1342, (Ct. Cl.
1973) (“To anticipate, a prior art reference must disclose each and every
element of a claimed invention, or its equivalents, and the element must
Appeal 2009-005596
Application 10/749,271

10
function in substantially the same way to produce substantially the same
result.”). With regard to Appellants’ use of the terms “place,” “optimal
position,” and “replay loop,” we are guided by In re Bond, 910 F.2d 831,
832-33 (Fed. Cir. 1990); and In re May, 574 F.2d 1082, 1090 (CCPA 1978)
(citing In re Schaumann, 572 F.2d 312, 317 (CCPA 1978) (“A reference
does not fail as an anticipation merely because it does not contain a
description of the subject matter of the appealed claim in ipsissimis
verbis.”). Merchant may anticipate Appellants’ claim 1 even if it does not
use identical terminology. Thus, Appellants have not persuaded us to find
error in the Examiner’s anticipation rejection of claims 1, 2, 4, and 7.
Therefore, we affirm the Examiner’s rejection of these claims.

Rejections Under § 103(a)
As we explained supra, Appellants simply reiterate the same
arguments with respect to the Examiner’s obviousness rejections, as set forth
in response to the anticipation rejection of independent claim 1.
Specifically, Appellants argue that Merchant does not teach placing an
instruction in an optimal position within a replay loop, and that the Merchant
‘838 and Topham references do not cure the deficiencies of Merchant – see
(1) rejection of claims 3, 9-15, 18-23, 25-27, 29, and 30 (in view of
Merchant) (App. Br. 8); (2) rejection of claims 5, 6, 16, and 17 (in view of
Merchant and Merchant ‘838) (App. Br. 8); and (3) rejection of claims 8, 24,
and 28 (in view of Merchant and Topham) (App. Br. 9). We have already
addressed these arguments and found them to be unpersuasive in our
discussion of independent claim 1. Thus, Appellants have not persuaded us
Appeal 2009-005596
Application 10/749,271

11
to find error in the Examiner’s obviousness rejection of claims 3, 5, 6, and 8-
30.

CONCLUSIONS OF LAW
On the record before us, we find the Examiner did not err in finding
the Merchant reference discloses, and would have taught or suggested to a
skilled artisan, “[a]n adaptive replay system” including “a selector device
coupled to [a staging unit] to place [an] instruction in an optimal position
within [a] replay loop” (as recited in Appellants’ claim 1). Thus, on the
record before us, we find the Examiner did not err in rejecting claims 1, 2, 4,
and 7 under 35 U.S.C. § 102(b), and claims 3, 5, 6, and 8-30 under
35 U.S.C. § 103(a).

DECISION
We affirm the Examiner’s rejection of claims 1, 2, 4, and 7 under
35 U.S.C. § 102(b).
We affirm the Examiner’s rejections of claims 3, 5, 6, and 8-30 under
35 U.S.C. § 103(a).

Appeal 2009-005596
Application 10/749,271

12
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)(iv).

AFFIRMED

rwk

Gerbera/BSTZ
Blakely Sokoloff Taylor & Zafman LLP
1279 OAKMEAD PARKWAY
SUNNYVALE, CA 94085