Ex Parte Kallio et al

Patent Trial and Appeal Board

Mar 27, 2018

13298051 (P.T.A.B. Mar. 27, 2018)

UNITED STA TES p A TENT AND TRADEMARK OFFICE
APPLICATION NO. FILING DATE
13/298,051 11/16/2011
15150 7590 03/29/2018
Shumaker & Sieffert, P.A.
1625 Radio Drive, Suite 100
Woodbury, MN 55125
FIRST NAMED INVENTOR
Kiia Kaappoo Kallio
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.
1010-828US01/l 11841 5968
EXAMINER
GE, JIN
ART UNIT PAPER NUMBER
2616
NOTIFICATION DATE DELIVERY MODE
03/29/2018 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):
pairdocketing@ssiplaw.com
ocpat_uspto@qualcomm.com
PTOL-90A (Rev. 04/07)
UNITED STATES PATENT AND TRADEMARK OFFICE
BEFORE THE PATENT TRIAL AND APPEAL BOARD
Ex parte KHA KAAPPOO KALLIO and WKKA-PEKKA ARVO
Appeal2017-005248
Application 13/298,051 1
Technology Center 2600
Before ERIC B. CHEN, MONICA S. ULLAGADDI, and SCOTT E. BAIN,
Administrative Patent Judges.
BAIN, Administrative Patent Judge.
DECISION ON APPEAL
Appellants appeal under 35 U.S.C. § 134(a) from the Examiner's
Final Rejection of claims 1, 2, 4---6, 8-17, 19-22, 24--26, 28-30, 32-34, 36-
41, and 43--48, which constitute all the claims pending in the
application. We have jurisdiction under 35 U.S.C. § 6(b ).
We REVERSE.
1 Appellants identify the real party in interest as QUALCOMM
Incorporated. App. Br. 3.
Appeal2017-005248
Application 13/298,051
STATEMENT OF THE CASE
The Invention
Appellants' invention relates to graphics processing, and specifically,
to techniques for tessellation when using a tile-based rendering architecture.
Spec. i-f 1. "Tessellation" is a process by "graphics primitives are divided
into smaller graphics primitives, thus allowing for finer detail in the
displayed image." Id. i-f 23; see also Title; Abst. Appellants' claimed
processes and systems involve "skipping" tessellation of obscured
primitives, i.e., those hidden from view when displayed, to increase
efficiency of graphics processing. Id. at i-fi-12, 6-8.
Claims 1, 8, 14, 19, 25, 32, 38, and 43 are independent. Claim 1 is
illustrative of the invention and the subject matter of the appeal, and reads as
follows (with disputed limitation emphasized):
Claim 1: A method for tessellation in a tile-based graphics
rendering pipeline by a graphics processor, the method
compnsmg:
generating an input visibility stream for each tile of a frame
before rendering the frame, the input visibility stream indicating
whether or not an input primitive is visible in each tile when
rendered;
generating an output visibility stream for each tile of the frame
before rendering the frame, the output visibility stream including
an indication of whether or not an output primitive is visible in
the tile based on a depth test,
wherein the output primitive is determined to be visible if
any pixel produced by rasterizing the output primitive is
determined to be visible, and
2
Appeal2017-005248
Application 13/298,051
wherein the output primitive is produced by tessellating
the input primitive,
wherein the output visibility stream comprises a series of
bits that indicate for every output primitive whether it
contributes to the rendered tile or not and the input
visibility stream comprises a series of bits that indicate for
every input primitive whether it contributes to the
rendered tile or not; and
tessellating the input primitives to create the output primitives
for a current tile based on the input visibility stream and the
output visibility stream,
wherein tessellating the input primitives comprises:
skipping tessellation for the input primitives
indicated by the input visibility stream as not
contributing to the current tile; and
performing tessellation, based on the output
visibility stream, for input primitives indicated by
the input visibility stream as contributing to the
current tile,
wherein tessellation is skipped for a portion of the input
primitives for which corresponding output primitives are
indicated as not contributing to the current tile by the
output visibility stream.
App. Br. 32-33 (Claims App.) (emphasis added).
The Rejections on Appeal
Claims 1, 2, 4---6, 14--17, 25, 26, 28-30, and 38--41 stand rejected
under pre-AIA 35 U.S.C. § 103(a) as unpatentable over Howson
(US 2011/0267346 Al; Nov. 3, 2011) and Yu et al. (US 2012/0212488 Al;
Aug. 23, 2012) ("Yu"). Final Act. (Apr. 19, 2016) 7--43.
3
Appeal2017-005248
Application 13/298,051
Claims 8-13, 19-22, 24, 32-34, 36, 37, and 43--48 stand rejected
under pre-AIA 35 U.S.C. § 103(a) as unpatentable over Howson, Diard et al.
(US 7,388,581 Bl; June 17, 2008), and Liao et al. (US 2005/0057564 Al
Mar. 17, 2005). Final Act. 43-72.
ANALYSIS
We have reviewed the Examiner's rejection in light of the arguments
raised in the Briefs, on the record before us. For the reasons set forth below,
we do not sustain the Examiner's rejections.
Claim 1
Appellants argue the Examiner erred in finding the prior art teaches or
suggests "generating ... an indication of whether or not an output primitive
is visible ... wherein the output primitive is produced by tessellating [an]
input primitive, ... [and] wherein tessellation is skipped for a portion of the
input primitives for which corresponding output primitives are indicated as
not contributing to the current tile," as recited in claim 1. App. Br. 13-15;
Reply Br. 5---6. Specifically, Appellants contend, "to the extent that Howson
teaches ... indications of visibility, Howson does so with reference to
tessellated primitives, not ... [primitives] prior to the tessellation." App.
Br. 15 (orig. emph.); see also Reply Br. 5---6.
The Examiner finds Howson teaches or suggests the foregoing
disputed limitation. Ans. 6 (citing Howson Figs. 6-8, i-fi-122-24, 41--43,
4
Appeal2017-005248
Application 13/298,051
45). 2 Specifically, the Examiner finds "Howson teaches a step (hide surface
remove) to remove the invisible primitives and keep visible primitives
before the te[ss]ellation (see Fig 6, par 0042, HSR, par 0022, par 0045), then
provide a tessellation step to the input visible primitives (see Fig 7, par 0043,
par 0024, par 0045))." Ans. 6. Appellants' arguments, however, persuade
us the Examiner has erred.
Howson' s hidden surface removal system distinguishes between
obscured and visible primitives, and then removes the obscured primitives
from processing. Howson iii! 41--43. The record before us, however, does
not indicate that Howson's removal of obscured primitives avoids or skips
any tessellating of a primitive into smaller primitives, as recited in
Appellants' claim 1. Rather, as Appellants argue (see supra 4), Howson's
system fully tessellates all primitives and then removes any obscured
resulting primitives from further processing. Howson ii 44; see also id. at
ii 24 ("[H]idden surface removal ... is performed ... for the tessellated
primitives.").
Only the tessellation occurs in Howson's pre-rasterization phase
("geometry processing") of generating "primitive indices for the tessellated
primitives." Id. at ii 22; Reply Br. 6. Specifically, input primitives
("patches") are provided to shading unit 805 for vertex and hull shading.
Howson ii 44. The resulting data is passed to domain tessellation unit 825
and connectivity tessellation unit 830, which respectively generate domain
2 The Examiner appears to cite (at one point or another in the Final Action
and/or Answer) the entirety of Howson's disclosure, other than the
background and paragraphs 26, 39, and 40 of the summary and detailed
description. Final Act. 7-16; Ans. 2-10.
5
Appeal2017-005248
Application 13/298,051
points and tessellated primitive indices. Id.; see also id. at i-f 7
("Connectivity tessellation determines how the resulting 'Domain' points are
combined or connected to produce[] tessellated primitives."). The primitive
indices are then cached (in cache 815). Id. at i-f 44.
The identification of obscured primitives and their removal from
processing occur in two ensuing rasterization phases of Howson. Id. at i-fi-1
23- 24. The tessellated domain points are twice fed back to the shading unit
805 for this identification and removal. Id.; Reply Br. 6. The shading unit
805 first performs only position processing (in the first rasterization phase)
and a hidden surface removal unit 840 uses the position data to identify the
obscured primitives. Howson i-f 24. The shading unit 805 then performs the
attribute processing (in the second rasterization phase) for only visible
primitives, thus avoiding this further processing for obscured primitives. Id.
at ,-r 25.
We cannot discern where, in Howson's processes as described above,
the Examiner finds tessellation is skipped for the obscured primitives. We
observe that, in the above-described two phases concerning obscured
primitives (i.e., rasterization phases), the domain tessellation unit 825
recalculates the domain points for the shading unit 805. Id. at i-fi-124--25.
There is no indication, however, that the recalculating subdivides a primitive
into smaller primitives (and thus no indication that the recalculating for only
visible primitives skips tessellating obscured primitives). Howson describes
the recalculating by the domain tessellation unit 825 as merely generating
"u,v values." Id. at i-f 45. Howson does not even mention the connection
tessellation unit 830 with respect to the recalculating-much less describe it
as again "produc[ing] tessellated primitives" (id. at i-f 7).
6
Appeal2017-005248
Application 13/298,051
In sum, the Examiner does not explain how Howson's removal of
obscured primitives from processing would teach or suggest avoiding
tessellation of a primitive into smaller primitives. Accordingly, we are
persuaded by Appellants' arguments that the Examiner has not
demonstrated, on this record, that Howson teaches or suggests the disputed
claim limitation.
Moreover, the Examiner's interpretation of "tessellation" is unclear,
on the record before us. Appellants' Specification (unlike Howson's use of
the term "tessellation") expressly distinguishes "tessellation" itself from the
"additional processing stages" of the graphics processing pipeline. Id. at
i-f 23. Giving "tessellation" its broadest reasonable interpretation in view of
the Specification, we interpret Appellants' claimed "tessellation" to mean
"[t]he basic function of the tessellator 38[, which] is to tessellate (i.e.,
subdivide) input primitives (e.g., patches) of a 3D scene into multiple output
primitives." Id. at i-f 42; see also id. at i-fi-147 ("tessellator input (i.e., patches)
and output (i.e., triangles, quads, or isolines) primitives"), i-f 48 ("tessellation
input primitives (patches) and ... tessellation output primitives (triangles)"),
i-f 61 ("tessellation is completed" by the subdividing process of Figures 6-7);
accord tessellate. (2010). In S. M. H. Collin (Ed.), Dictionary of Computing
(6th ed.) London, UK: Bloomsbury (defining "tessellate" as "to reduce a
complex shape into a collection of simple shapes, often triangles"), available
at https :// search.credoreference.com/ content/ entry/ acbcomp/tessellate/
O?institutionid=743 (last visited Mar. 22, 2018).
If the Examiner has based the conclusion of obviousness upon a
broader interpretation of "tessellation" (which, again, is unclear on the
record before us), that interpretation is erroneous. See In re Power
7
Appeal2017-005248
Application 13/298,051
Integrations, Inc., No. 2017-1304, 2018 WL 1370551, at *6 (Fed. Cir. Mar.
19, 2018) ("The correct inquiry in giving a claim term its broadest
reasonable interpretation in light of the specification is not whether the
specification proscribes or precludes some broad reading of the claim term
adopted by the examiner. Instead, a proper claim construction analysis
endeavors to assign a meaning to a disputed claim term that corresponds
with how the inventor describes his invention in the specification."
(quotation and citation omitted)).
For the foregoing reasons, we do not sustain the obviousness rejection
of claim 1.
Remaining Claims
Independent claims 8, 14, 19, 25, 32, 38, and 43 include the foregoing
disputed limitations of claim 1, as discussed above, and the Examiner relies
on Howson as teaching or suggesting the disputed limitations. For the same
reasons as claim 1, we do not sustain the obviousness rejections of claims 8,
14, 19, 25, 32, 38, and 43. We also not sustain the rejections of the
remaining claims, all of which are dependent and thus include the limitations
of their base claims.
DECISION
We reverse the Examiner's decision rejecting claims 1, 2, 4---6, 8-17,
19-22, 24--26, 28-30, 32-34, 36-41, and 43--48.
REVERSED
8