Advanced Micro Devices, Inc.

Patent Trials and Appeals Board

Feb 16, 2022

2021005007 (P.T.A.B. Feb. 16, 2022)

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. 16/219,820 12/13/2018 Skyler Jonathon Saleh AMD-180447-US-ORG1 1200 25310 7590 02/16/2022 Volpe Koenig DEPT. AMD 30 SOUTH 17TH STREET -18TH FLOOR PHILADELPHIA, PA 19103 EXAMINER HOANG, PETER ART UNIT PAPER NUMBER 2616 NOTIFICATION DATE DELIVERY MODE 02/16/2022 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): eoffice@volpe-koenig.com PTOL-90A (Rev. 04/07) UNITED STATES PATENT AND TRADEMARK OFFICE BEFORE THE PATENT TRIAL AND APPEAL BOARD Ex parte SKYLER JONATHON SALEH and RUIJIN WU Appeal 2021-005007 Application 16/219,820 Technology Center 2600 Before JEAN R. HOMERE, CARL W. WHITEHEAD JR., and BRADLEY W. BAUMEISTER, Administrative Patent Judges. Opinion for the Board filed by Administrative Patent Judge JEAN R. HOMERE. Concurring opinion filed by Administrative Patent Judge BRADLEY W. BAUMEISTER. HOMERE, Administrative Patent Judge. DECISION ON APPEAL Appeal 2021-005007 Application 16/219,820 2 I. STATEMENT OF THE CASE1 Pursuant to 35 U.S.C. § 134(a), Appellant2 appeals from the Examiner’s decision to reject claims 1-20. Claims App. We have jurisdiction under 35 U.S.C. § 6(b). An oral hearing was held in this appeal on February 4, 2022. A transcript of the oral hearing will be entered into the record in due course. We affirm. II. CLAIMED SUBJECT MATTER According to Appellant, the claimed subject matter relates to a graphics rendering technique for casting simulated rays of light onto vertices of a triangle to thereby perform a ray-triangle intersection test that produces watertight results. In particular, the technique involves transforming (i.e., projecting and translating) the coordinates of the triangle (a.k.a., barycentric coordinates) into the viewspace of the ray such that the origin of the triangle is at the origin of the ray. Spec. ¶ 10. More particular, the Specification states the following: Described herein is a technique for performing ray-triangle intersection test in a manner that produces watertight results. The technique involves translating the coordinates of the triangle such that the origin is at the origin of the ray. The technique involves projecting the coordinate system into the viewspace of the ray. The technique 1 We refer to the Specification, filed Dec. 13, 2018 (“Spec.”); Final Office Action, mailed Oct. 23, 2020 (“Final Act.”); Appeal Brief, filed Apr. 14, 2021 (“Appeal Br.”); Examiner’s Answer, mailed June 17, 2021 (“Ans.”); and Reply Brief, filed Aug. 17, 2021 (“Reply Br.”). 2 “Appellant” refers to “Applicant” as defined in 37 C.F.R. § 1.42(a). Appellant identifies the real party in interest as Advanced Micro Devices, Inc., Appeal Br. 3. Appeal 2021-005007 Application 16/219,820 3 then involves calculating barycentric coordinates and interpolating the barycentric coordinates to get a time of intersect. The signs of the barycentric coordinates indicate whether a hit occurs. The above calculations are performed with a non-directed floating point rounding mode to provide watertightness. A non- directed rounding mode is one in which the mantissa of a rounded number is rounded in a manner that is not dependent on the sign of the number. Spec. ¶ 10 (emphasis added). The ray-triangle test involves asking whether the ray hits the triangle and also the time to hit the triangle (time from ray origin to point of intersection). Conceptually, the ray-triangle test involves projecting the triangle into the viewspace of the ray so that it is possible to perform a simpler test similar to testing for coverage in two-dimensional rasterization of a triangle as is commonly performed in graphics processing pipelines. More specifically, projecting the triangle into the viewspace of the ray transforms the coordinate system so that the ray points downwards in the z direction and the x and y components of the ray are 0 (although in some modifications, the ray may point upwards in the z direction, or in the positive or negative x or y directions, with the components in the other two axes being zero). The vertices of the triangle are transformed into this coordinate system. Such a transform allows the test for intersection to be made by simply asking whether the x, y coordinates of the ray fall within the triangle defined by the x, y coordinates of the vertices of the triangle, which is the rasterization operation described above. Spec. ¶ 31. Appeal 2021-005007 Application 16/219,820 4 Figures 5 and 6, reproduced below, are useful for understanding the claimed subject matter: Figure 5 above depicts the transformation of the coordinates of triangle 504 projected into the viewspace of ray 502 so that their origins coincide or intersect at (00). That is, the projection of ray 502 and triangle 504 in coordinate system 500 is transformed in coordinate system 510 such that ray 512 is shown pointing in the -z direction and triangle 514 is repositioned after the transformation. Spec. ¶ 33. Figure 6 illustrates the ray intersection test as a rasterization operation. Specifically, vertices A, B, and Appeal 2021-005007 Application 16/219,820 5 C define triangle 514 and vertex T is the origin of the ray 512. Testing for whether the ray 512 intersects the triangle 514 is performed by testing whether vertex T is within triangle ABC. Spec. ¶ 34. Claims 1, 10, and 19 are independent. Claim 1, with disputed limitations emphasized, is illustrative: 1. A method for detecting a hit between a ray and a triangle, the method comprising: projecting, into a viewspace of the ray, vertices of the triangle, by transforming the vertices of the triangle and a vertex representative of a direction of the ray, into a coordinate system in which the ray direction has x and y components of 0 and each of the vertices and the ray have z components that are unmodified by the coordinate transformation unit; determining barycentric coordinates that describe the location of the point of intersection of the ray relative to the vertices of the triangle in two-dimensional space, wherein determining the barycentric coordinates is performed using a non-directed rounding mode; and interpolating the barycentric coordinates to generate a numerator and a denominator for a time of intersection of the ray with the triangle. Appeal Br. 14 (emphasis added). III. REFERENCES The Examiner relies upon the following references as evidence.3 Name Reference Date Maiyuran US 2016/0189327 A1 June 30, 2016 Laine US 2020/0051314 A1 Feb. 13, 2020 3 All reference citations are to the first named inventor only. Appeal 2021-005007 Application 16/219,820 6 Woop Watertight Ray Triangle Intersection, Journal of Computer Graphics Techniques, Vol. 2, No.1 2013 IV. REJECTIONS The Examiner rejects claims 1-20 as follows: Claims 1-3, 5-12, and 14-20 stand rejected under 35 U.S.C. § 103 as unpatentable over the combined teachings of Woop and Laine. Final Act. 5- 10. Claims 4 and 13 stand rejected under 35 U.S.C. § 103 as unpatentable over the combined teachings of Woop, Laine, and Maiyuran. Id. at 10-11. V. ANALYSIS We consider Appellant’s arguments seriatim, as they are presented in the Appeal Brief, pages 7-13 and the Reply Brief, pages 1-18.4 We are unpersuaded by Appellant’s contentions. Except as otherwise indicated herein below, we adopt as our own the findings and reasons set forth in the Final Action, and the Examiner’s Answer in response to Appellant’s Appeal Brief.5 Final Act. 3-12; Ans. 3-17. However, we highlight and address specific arguments and findings for emphasis as follows. 4 We have considered in this Decision only those arguments Appellant actually raised in the Briefs. Arguments not made are forfeited. See 37 C.F.R. § 41.37(c)(1)(iv) (2017). 5 See ICON Health and Fitness, Inc. v. Strava, Inc., 849 F.3d 1034, 1042 (Fed. Cir. 2017) (“As an initial matter, the PTAB was authorized to incorporate the Examiner’s findings.”); see also In re Brana, 51 F.3d 1560, 1564 n.13 (Fed. Cir. 1995) (upholding the PTAB’s findings, although it “did Appeal 2021-005007 Application 16/219,820 7 Appellant argues that the Examiner errs in finding that the combined teachings of Woop and Laine render claim 1 obvious. Appeal Br. 7-10. In particular, Appellant argues that because Laine does not teach using a non- directed rounding mode for determining (rounding) barycentric coordinates, as recited in independent claim 1, it does not cure the admitted deficiencies of Woop. Id. According to Appellant, although Laine discloses rounding intersection results, they are not barycentric coordinates; nor does the disclosure of rounding a number to zero indicates using a non-directed rounding mode. Id. at 8-9 (citing Laine ¶ 201). Therefore, Appellant submits that because the cited disclosure of Laine does not indicate that a mode should be used to round to zero all values that are sufficiently close to zero, it does not teach using a non-directed mode for rounding the barycentric coordinates. Id. In response to the Examiner’s findings and conclusions in the Examiner’s Answer, Appellant further submits that although Woop discloses transforming a ray by calculating transformed triangle vertices relative to the ray origin point at ray-triangle intersection time, and then calculating scaled barycentric coordinates to finish the intersection, Woop does not teach using a non-directed rounding mode to round the barycentric coordinates. Reply Br. 3 (citing Woop 68, 1st para, section 3). Appellant then reiterates that Laine does not cure the noted deficiencies of Woop because Laine does not specify how the mantissa is rounded. Id. at 3-14 (citing Spec. ¶¶ 10, 52-54). not expressly make any independent factual determinations or legal conclusions,” because it had expressly adopted the examiner’s findings). Appeal 2021-005007 Application 16/219,820 8 Appellant’s arguments are not persuasive of reversible Examiner error. As an initial matter, we note that the disputed claim language “non- directed rounding mode” must be given its broadest reasonable interpretation consistent with Appellant’s disclosure, as explained in Morris: [T]he PTO applies to the verbiage of the proposed claims the broadest reasonable meaning of the words in their ordinary usage as they would be understood by one of ordinary skill in the art, taking into account whatever enlightenment by way of definitions or otherwise that may be afforded by the written description contained in the applicant’s specification. In re Morris, 127 F.3d 1048, 1054 (Fed. Cir. 1997); see also In re Zletz, 893 F.2d 319, 321 (Fed. Cir. 1989) (stating that “claims must be interpreted as broadly as their terms reasonably allow.”). Our reviewing court states, “the proper … construction is not just the broadest construction, but rather the broadest reasonable construction in light of the specification.” In re Man Mach. Interface Techs. LLC, 822 F.3d 1282, 1287 (Fed. Cir. 2016). Our reviewing court further states, “the ‘ordinary meaning’ of a claim term is its meaning to the ordinary artisan after reading the entire patent.” Phillips v. AWH Corp., 415 F.3d 1303, 1321 (Fed. Cir. 2005) (en banc). In this case, Appellant’s Specification recites in relevant parts the following: A non-directed rounding mode is one in which the mantissa of a rounded number is rounded in a manner that is not dependent on the sign of the number. Spec. ¶ 10. There are several different rounding modes: Appeal 2021-005007 Application 16/219,820 9 that round to zero (RTZ), round to nearest even (RTNE), round to positive infinity (RTP), and round to negative infinity (RTN). RTZ and RTNE are both non-directed rounding modes and RTP and RTN are both directed rounding modes. The “directedness” of the rounding mode means that the manner in which the magnitude of the mantissa is rounded depends on the sign of the floating point number. Id. ¶ 54 (emphasis added). The cited portions of Appellant’s Specification indicate round to zero (RTZ), and round to nearest even (RTNE) are both non-directed rounding modes in which the mantissa of a rounded number is rounded in a manner independent of the sign of the number. Whereas the Specification indicates that round to positive infinity (RTP), and round to negative infinity (RTN) are both directed rounding modes in which the magnitude of the mantissa is rounded depends on the sign of the floating point number. Accordingly, we construe “non-directed rounding mode” to refer to rounding the mantissa of a rounded number independently of its sign. As correctly noted by the Examiner, and undisputed by Appellant, Woop discloses transforming a ray by calculating transformed triangle vertices relative to the ray origin point at ray-triangle intersection time, and then calculating scaled barycentric coordinates to finish the intersection. Ans. 10-11; Reply Br. 3 (citing Woop 68, 1st para, section 3). Further, the Examiner relies upon Laine’s disclosure of the following: Intersection results may be inaccurately determined at or near an edge or vertex of a triangle because small results are often computed at these locations and rounding in floating point arithmetic operations can produce insufficient precision to make accurate spatial determinations. For example, computations for a ray intersection near an edge or vertex will produce a small positive or negative value, which may get rounded to zero during Appeal 2021-005007 Application 16/219,820 10 the computations by the limited precision computation hardware. To address this issue, example non-limiting embodiments use fused floating-point operation units configured to combine multiple floating-point arithmetic operations to efficiently achieve higher numerical precision. The fused floating-point operation units may be implemented in hardware included in the TTU and in some example embodiments, in the Ray Triangle Test (RTT) block of the TTU (see e.g., block 720 in FIG. 9). Ans. 11 (citing Laine ¶ 201). We agree with the Examiner that Laine’s rounding to zero the floating point of the ray intersection near a vertex uses a non-directed mode because, pursuant to our claim interpretation above consistent with the Specification, the rounding is performed irrespective of the sign of the floating point. Id. at Ans. 11. We further agree with the Examiner that Appellant’s argument that Laine’s disclosed rounding is not performed using a non-rounding mode because it does not round all numbers that are substantially close zero is not commensurate with the scope of the claim, nor is it supported by the Specification. Ans. 9. It suffices that Laine’s disclosed mantissa of the number is rounded irrespective of the sign thereof. Further, we agree with the Examiner that Woop’s disclosure of transforming a ray by calculating transformed triangle vertices relative to the ray origin point at ray-triangle intersection time calculating scaled barycentric coordinates to finish the intersection, and then rounding to a nearby floating point number the transformed vertices teaches or suggests rounding the barycentric coordinates. Id. at 10-11 (citing Woop 68-70). We find the Examiner’s proposed combination of the cited teachings of Woop and Laine is no more than a simple arrangement of old elements with each performing the same function it had been known to perform, Appeal 2021-005007 Application 16/219,820 11 yielding no more than what one would expect from such an arrangement. See KSR Int’l Co. v. Teleflex Inc., 550 U.S. 398, 416 (2007). Therefore, the ordinarily skilled artisan, being “a person of ordinary creativity, not an automaton,” would have been able to fit the teachings of the cited references together like pieces of a puzzle to predictably result in a technique using a non-directed rounding mode for determining barycentric coordinates of a triangle that describe the location of the point of intersection of the ray relative to the vertices of the triangle in two-dimensional space. Id. at 420- 21. Because Appellant has not demonstrated that the Examiner’s proffered combination would have been “uniquely challenging or difficult for one of ordinary skill in the art,” we agree with the Examiner that the proposed modification would have been within the purview of the ordinarily skilled artisan. See Leapfrog Enters., Inc. v. Fisher-Price, Inc., 485 F.3d 1157, 1162 (Fed. Cir. 2007) (citing KSR, 550 U.S. at 418). Because we are not persuaded of Examiner error, we sustain the Examiner’s rejection of claim 1 as obvious over the combined teachings of Woop and Laine. Regarding the rejection of claims 2, 3, 5-12, and 14-20, Appellant either does not present separate patentability arguments or reiterates substantially the same arguments as those discussed above for the patentability of claim 1. As such, claims 2, 3, 5-12, and 14-20, fall therewith. See 37 C.F.R. § 41.37(c)(1)(iv). On this record, we affirm the Examiner’s rejections of claims 1-20. Appeal 2021-005007 Application 16/219,820 12 VII. DECISION SUMMARY In summary: Claim(s) Rejected 35 U.S.C. § Reference(s)/Basis Affirmed Reversed 1-3, 5-12, 14-20 103 Woop, Laine 1-3, 5-12, 14-20 4, 13 103 Woop, Laine, Maiyuran 4, 13 Overall Outcome 1-20 VIII. TIME PERIOD FOR RESPONSE 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). See 37 C.F.R. § 1.136(a)(1)(iv). AFFIRMED Appeal 2021-005007 Application 16/219,820 13 BAUMEISTER, Administrative Patent Judge, concurring. I agree with the Majority’s decision affirming the appealed obviousness rejections. I write separately for two reasons. I. As a threshold matter, I note that claim 1 recites a method that only consists of performing three mathematical operations: (1) a step of performing a mathematical transform of a geometric coordinate system; (2) a step of performing a mathematical analysis, including a rounding operation, to determine barycentric coordinates; and (3) a step of interpolating the barycentric coordinates that were determined in step (2). Claim 1 does not recite using these three math steps in conjunction with any post-solution activity or practical application, such as generating a picture. Furthermore, claim 1 does not recite any additional elements, such as computers or calculators for performing the three recited math steps. We do not address the question of whether claim 1 (or any of the other claims) satisfies the requirements of 35 U.S.C. § 101 or alternatively whether claim 1 is directed to an exception to patent-eligible subject matter without reciting significantly more. Our decision not to address this question sua sponte should not be interpreted as meaning that we find claim 1 to satisfy 35 U.S.C. § 101. Rather, we do not address this issue because it is not a ground of the Examiner’s rejection, and the issue is not before us on appeal. See ex parte Frye, 94 USPQ2d 1072, 1075 (BPAI 2010) (precedential) (The Board conducts a limited de novo review of the appealed rejections for error based upon the issues identified by Appellant, and in light of the arguments and evidence produced thereon.). Appeal 2021-005007 Application 16/219,820 14 II. Turning to the merits of the appeal, I would affirm the rejections of record because Woop, alone, either teaches or at least suggests every limitation of the disputed claims. The Examiner need not rely on the additional teachings of Laine to demonstrate that the appealed claims are unpatentable. See In re Meyer, 599 F.2d 1026, 1031 (CCPA 1979) (noting that obviousness rejections can be based on references that happen to anticipate the claimed subject matter). More specifically, Woop discloses that in performing the disclosed calculations, the computer follows an industry standard for rounding numbers, which standard entails rounding to a nearby floating-point number. Woop 70 (“The IEEE 754 floating-point standard [IEEE 1985] requires calculations (such as multiplications) to be internally executed with (principally) infinite precision and finally precisely rounded to a nearby floating-point number.”). That is, because a non-directed rounding mode entails rounding to the nearest number of desired digits independent of the sign of the number (Spec. ¶ 10)6, Woop’s disclosure of rounding to a nearby floating-point either teaches, or at least suggests, performing the disclosed rounding using a non-directed rounding mode. 6 Non-directed rounding is distinguished from directional rounding- rounding to negative infinity (or truncating) and rounding to positive infinity. Spec. ¶ 54.