IMAGINATION TECHNOLOGIES LIMITEDDownload PDFPatent Trials and Appeals BoardOct 19, 20212020004960 (P.T.A.B. Oct. 19, 2021) Copy Citation 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. 13/725,424 12/21/2012 Morrie Berglas 2645-0160US01 1001 104840 7590 10/19/2021 Imagination Technologies c/o Potomac Law Group PLLC 8229 Boone Boulevard Suite 430 Vienna, VA 22182 EXAMINER RICHER, AARON M ART UNIT PAPER NUMBER 2612 NOTIFICATION DATE DELIVERY MODE 10/19/2021 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): Eofficeaction@appcoll.com Patents@potomaclaw.com vdeluca@potomaclaw.com PTOL-90A (Rev. 04/07) UNITED STATES PATENT AND TRADEMARK OFFICE ____________________ BEFORE THE PATENT TRIAL AND APPEAL BOARD ____________________ Ex parte MORRIE BERGLAS ________________________ Appeal 2020-004960 Application 13/725,424 Technology Center 2600 ____________________ Before ALLEN R. MacDONALD, DAVID M. KOHUT, and HUNG H. BUI, Administrative Patent Judges. BUI, Administrative Patent Judge. DECISION ON APPEAL Appellant1 seeks our review under 35 U.S.C. § 134(a) of the Examiner’s Final Rejection of claims 5–25, all the pending claims pending. Appeal Br. 16–21 (Claims App.). We have jurisdiction under 35 U.S.C. § 6(b). We affirm.2 1 Appellant refers to “Applicant” as defined in 37 C.F.R. § 1.42. Appellant identifies the real party in interest as Imagination Technologies Limited. Appeal Br. 1. 2 Our Decision refers to Appellant’s Appeal Brief filed December 16, 2019 (“Appeal Br.”); Reply Brief filed June 17, 2020 (“Reply Br.”); Examiner’s Appeal 2020-004960 Application 13/725,424 2 STATEMENT OF THE CASE Appellant’s Invention Appellant’s Admitted Prior Art (AAPA) graphics techniques for applying texture maps to a frame of polygon to render a 3-D image require a blending unit to perform multiple “read-modify-write” operations, i.e., read pixels from a frame buffer, modify these pixels by applying texture data in a blending operation, and then write them back to the frame buffer. Spec. 1:24–32; 5:17. However, such operations incur extra cost of storage and lead to “a performance penalty because of the limited bandwidth for the ‘read modify write’ process.” Spec. 3:3–4:5. Appellant’s invention seeks to provide pixel blending buffer 32, shown in Figure 3, on a graphics chip to (1) enable portions of a frame buffer to be accessed on a polygon-by-polygon basis and (2) enable multiple textures to be accessed simultaneously in a single blending operation in order to avoid the need to perform “read-modify-write” operations on the frame buffer. Spec. 4:6–16, 5:14–18. Figure 3 shows a modified 3-D pixel pipeline in a graphics chip, as reproduced below with additional markings for illustration. Answer mailed April 17, 2020 (“Ans.”); Final Office Action mailed August 15, 2019 (“Final Act.”); and original Specification filed December 21, 2012 (“Spec.”). Appeal 2020-004960 Application 13/725,424 3 Appellant’s Figure 3 depicts a modified 3-D pixel pipeline in a graphics chip where blend buffer 32 having read and write ports 34, 36 is arranged between texture read unit 4 and blend operations unit 8 to avoid the need to perform a “read-modify-write” operation on the frame buffer. Spec. 5:14– 18. This way, “the blend[ing] operations can be performed as many times as desired on the data held in the blend buffer [32] using feedback loop X which takes data directly from the blend buffer 8 to the write ports 34.” Spec. 5:18–21. As such, “current contents of the pixel data in the blend buffer [32] are written to the frame buffer 10 in a single [write only] Appeal 2020-004960 Application 13/725,424 4 operation [as opposed to multiple read-modify-write operations on the frame buffer as required by prior art techniques].” Spec. 9:12–14. Representative Claim Claims 5, 13, and 20 are independent. Representative claim 5 is reproduced below with the disputed limitation in italics: 5. A 3-D graphics system, comprising a graphics chip, the 3- D graphics system comprising: a frame buffer; a blend buffer, provided in a memory distinct from a memory in which the frame buffer exists, the blend buffer comprising a plurality of registers, a first write port, a second write port and at least one read port, wherein registers of the plurality of registers are assignable to pixels of one or more primitives; a texture read unit on the graphics chip coupled for reading from a texture memory and coupled to the first write port of the blend buffer, the texture read unit operable to receive calculated texture coordinates from a texture coordinate calculator, the texture coordinates for pixels of a polygon, and to use the received calculated texture coordinates in reading texture data for a plurality of textures from the texture memory and to store the texture data for the plurality of textures in the blend buffer in respective registers through the first write port; and a blending unit on the graphics chip coupled to the second write port of the blend buffer, and to the at least one read port of the blend buffer, the blending unit operable to perform a plurality of texturing passes on the polygon using data read from the blend buffer, wherein each pass of the plurality of texturing passes comprises walking the polygon, the walk comprising reading texture data, from a plurality of the registers, for each pixel of the polygon, producing a blending output, and storing the blending output of each pass in the plurality of registers of the blend buffer, through the second write port of the blend buffer, and only after completing the plurality of texturing passes, the blending unit is operable to read the pixels from the blend buffer Appeal 2020-004960 Application 13/725,424 5 [and] to write the pixels read from the blend buffer to the frame buffer only once, after all texturing passes have been performed as a single write-only transaction on the frame buffer, such that pixels written to the frame buffer by the blending unit in said single write-only transaction are not thereafter read by the blending unit. Appeal Br. 16 (Claims App.) (underlying added for emphasis). EVIDENCE CONSIDERED Koegel US 5,036,456 July 30, 1991 Eckart et al. (“Eckart”) US 5,574,847 Nov. 12, 1996 McCarthy et al. (“McCarthy”) US 5,666,509 Sept. 9, 1997 Vangemert et al. (“Vangemert”) US 6,973,561 B1 Dec. 6, 2005 Drebin et al. (“Drebin”) US 7,034,828 B1 Apr. 25, 2006 Mang et al. (“Mang”) US 7,111,156 B1 Sept. 19, 2006 Van Hook US 7,847,803 B1 Dec. 7, 2010 Baker et al. (“Baker”) US 2002/0140703 A1 Oct. 3, 2002 Jarvis US 2003/0014614 A1 Jan. 16, 2003 EXAMINER’S REJECTIONS (1) Claims 5, 8–10, 12–15, and 17–20 stand rejected under 35 U.S.C. § 103(a) as obvious over the teachings of Baker and Drebin. Final Act. 7–15. (2) Claims 6 and 7 stand rejected under 35 U.S.C. § 103(a) as obvious over the teachings of Baker, Drebin, and Vangemert. Final Act. 15– 16. Appeal 2020-004960 Application 13/725,424 6 (3) Claim 11 stands rejected under 35 U.S.C. § 103(a) as obvious over the teachings of Baker, Drebin, and Mang. Final Act. 16–17. (4) Claims 16 and 24 stand rejected under 35 U.S.C. § 103(a) as obvious over the teachings of Baker, Drebin, and Van Hook. Final Act. 17– 18. (5) Claim 21 stands rejected under 35 U.S.C. § 103(a) as obvious over the teachings of Baker, Drebin, and Koegel. Final Act. 18–19. (6) Claim 22 stands rejected under 35 U.S.C. § 103(a) as obvious over the teachings of Baker, Drebin, and McCarthy. Final Act. 19. (7) Claim 23 stands rejected under 35 U.S.C. § 103(a) as obvious over the teachings of Baker, Drebin, and Jarvis. Final Act. 20. (8) Claim 25 stands rejected under 35 U.S.C. § 103(a) as obvious over the teachings of Baker, Drebin, and Eckart. Final Act. 20–21. ANALYSIS At the outset, we note that the instant application has a complicated prosecution history, including its continuation of U.S. Application No. 11/725,354, filed on March 19, 2007, which was a continuation of U.S. Application No. 11/188,259, filed July 22, 2005 (abandoned), which was a continuation of U.S. Application No. 10/310,120, filed December 4, 2002 (abandoned). Spec. 1:3–5. Claims 5–25 were previously rejected under 35 U.S.C. § 103(a) as obvious over the teachings of Baker and Drebin and then appealed to the Board, via Appeal 2018-006827. The Board affirmed the Examiner’s final rejection of claims 5–25 based on the Examiner’s combination of Baker and Drebin. See Decision dated June 4, 2019. Appeal 2020-004960 Application 13/725,424 7 In response to the Board’s Decision, Appellant amended independent claims 5, 13, and 20 to (1) clarify that data is written to the frame buffer only once, after all texturing passes have been performed as a single write-only transaction on the frame buffer, and (2) add a new negative limitation: “such that pixels written to the frame buffer by the blending unit in said single write-only transaction are not thereafter read by the blending unit.” Appeal Br. 16 (Claims App.) (Emphasis added). However, we note that the newly added phrase is not described or supported by Appellant’s Specification. Nor do we find “a reason to exclude the relevant limitation” from Appellant’s Specification. See Santarus, Inc. v. Par Pharm., Inc., 694 F.3d 1344, 1351 (Fed. Cir. 2012). Because there is no basis in Appellant’s Specification for adding the negative limitation, such a negative limitation should not be accorded any patentable weight. Moreover, even if the negative limitation were to be accorded patentable weight (which we will), an ordinarily skilled artisan would understand that, when all texturing passes have been performed as a single write-only transaction on the frame buffer, pixel data written to the frame buffer in said single write-only transaction is no longer needed to be read by the blending unit in the manner recited in Appellant’s claims 5, 13, and 20. Claims 5, 13, and 20 Turning now to the substance of the obviousness rejection, the Examiner finds the combination of Baker and Drebin teaches or suggests all limitations of Appellant’s amended claims 1, 13, and 20. Final Act. 7–15. For example, the Examiner finds Baker teaches Appellant’s claimed “3-D Appeal 2020-004960 Application 13/725,424 8 graphics system,” shown in Figure 5A, including (1) a frame buffer, (2) a blend buffer, (3) a texture read unit, and (4) a blending unit operable to perform the disputed limitation: the blending unit is operable to read the pixels from the blend buffer [and] to write the pixels read from the blend buffer to the frame buffer only once, after all texturing passes have been performed as a single write-only transaction on the frame buffer, such that pixels written to the frame buffer by the blending unit in said single write-only transaction are not thereafter read by the blending unit. Final Act. 8–9 (citing Baker ¶¶ 9, 26–27, 66–70, Fig. 5A). Baker’s Figure 5A is reproduced below with additional markings for illustration. Appeal 2020-004960 Application 13/725,424 9 Baker’s Figure 5A shows a 3-D pixel pipeline in a graphics device 80 (shown in Figure 3), including (1) frame buffers 98 inclusive of (1a) destination frame buffer 98a to store pixel data, and (1b) temporary frame buffer 98b; (2) texture unit 92 to apply texture coordinate values from texture cache 94 in multiple texturing passes; and (3) blending units 93 to blend texture values for the pixel that is stored in temporary frame buffer Appeal 2020-004960 Application 13/725,424 10 98b with the texture values for the pixel obtained in the current texture pass in order to generate composite texture values that include texture values obtained from all texturing passes to be stored in temporary frame buffer 98b. Baker ¶¶ 60, 66–70. As recognized by the Examiner, “Baker does not [expressly] disclose that the blend buffer is provided in a memory distinct from a memory in which the frame buffer exists.” Final Act. 9. However, the Examiner relies on Drebin for “storing retrieved texture data in registers of a blending unit that acts as a blend buffer” to support the conclusion of obviousness. Id. at 9–10 (citing Drebin 3:17–25, 10:16–27, Figs. 5 and 8). Appellant does not challenge the Examiner’s rationale to incorporate Drebin into Baker to arrive at Appellant’s invention. Instead, Appellant contends neither Baker nor Drebin teaches the disputed limitation of claims 5, 13, and 20: the blending unit is operable to read the pixels from the blend buffer [and] to write the pixels read from the blend buffer to the frame buffer only once, after all texturing passes have been performed as a single write-only transaction on the frame buffer, such that pixels written to the frame buffer by the blending unit in said single write-only transaction are not thereafter read by the blending unit. Appeal Br. 7–14; Reply Br. 2–5. In particular, Appellant argues: Baker teaches that data in a destination frame buffer is modified to blend texture maps, to blend pixel data with results of texturing passes, and to perform alpha blending. That is, the writing of data to the frame buffer in Baker is not performed “as a single write-only operation” as recited in the present claims. Appeal 2020-004960 Application 13/725,424 11 Appeal Br. 8. According to Appellant, Paragraph [0068] [of Baker] describes that pixel data is stored in the destination frame buffer 98a. In a first pass, texture values are stored in a temporary frame buffer 98b. At [0069] Baker describes that in a second pass texture values from the temporary frame buffer are blended with new texture values and the blend result is stored in the temporary frame buffer 98b. At [0070], a third pass is described producing a new blended result. This blended result is not stored in the temporary frame buffer, but is stated to be “optionally” blended with the pixel data in the destination frame buffer 98a, as the PTAB emphasized at page 12 of the Decision.” Id. at 9 (emphasis added). In other words, Appellant argues: [i]n Baker, results from one or more texturing passes are blended with the destination pixel data stored in the destination frame buffer and used to control the pixels on the display device when a frame of the display device is refreshed. Thus, in Baker, pixels are read from the destination frame buffer, blended, and then re- stored in a “read modify-write” operation. Reply Br. 5 (emphasis added). Likewise, Appellant argues Drebin does not teach the disputed limitation of claims 5, 13, and 20 because (1) “Figure 5 of Drebin shows arrows both to and from the frame buffer 702 and a blend box 700b (which carries out frame buffer pixel blending - see col. 8, line 43), indicating a blend operation on the frame buffer” and (2) “figure 15 shows arrows both to and from the embedded frame buffer 702 and the blend box 700b [indicating] that the blending results of the final blending operation are output for blending with frame buffer 702 contents.” Final Act. 13–14. Appeal 2020-004960 Application 13/725,424 12 We do not find Appellant’s arguments persuasive of Examiner error. Instead, we find the Examiner has provided a comprehensive response to Appellant’s arguments, supported by a preponderance of evidence. Ans. 21– 28. Therefore, we adopt the Examiner’s findings and explanations provided therein. Id. For additional emphasis, we note that obviousness is the question of law, based upon factual inquiries set forth in the Graham case: (1) the scope and content of the prior art; (2) the differences between the prior art and the claims at issue; (3) the level of ordinary skill in the pertinent art; and (4) objective evidence of non-obviousness, if any. Graham v. John Deere Co., 383 U.S. 1, 17–18 (1966). “[H]owever, the analysis need not seek out precise teachings directed to the specific subject matter of the challenged claim, for a court can take account of the inferences and creative steps that a person of ordinary skill in the art would employ.” KSR Int’l Co. v. Teleflex Inc., 550 U.S. 398, 418 (2007). In this regard, “[a] person of ordinary skill is also a person of ordinary creativity, not an automaton.” Id. at 421. As correctly recognized by the Examiner, On the final pass, blending occurs that blends the values from the temporary frame buffer/blend buffer with current texture values and the result is stored in the destination frame buffer. This blending does not occur again because the result is sent to display, and so the result is written only once. Ans. 22. As shown in Figure 5A, Baker’s blending units 93 is configured to blend the composite texture value for the pixel that is stored in temporary frame buffer 98b (a.k.a., “blend buffer”) with texture values for the pixel Appeal 2020-004960 Application 13/725,424 13 obtained in the current texture passes, and then after all texture passes have been performed, stores the results as a single write-only transaction on destination frame buffer 98a. Baker ¶ 70. As an option, Baker also teaches that such results could also be blended with destination pixel data stored in destination frame buffer 98a in order to further control the color luminous intensity values (e.g., R, G, B) when a frame of a display device is refreshed. Baker ¶ 70. However, Baker’s additional blending is only an option and, as such, is not necessary to perform multiple rendering passes and to blend data from multiple texture maps that is applied to a polygon so as to model a 3-D image. Accordingly, we agree with the Examiner that “[i]f the option for blending data in the destination buffer is not exercised, the data resulting from texture blends is written with no read-modify step.” Ans. 24. For these reasons, we agree with the Examiner that: both [1] the prior art method described in section 0018 and [2] Baker’s method described in section 0070 support the concept of the single write-only transaction such that pixels written to the frame buffer by the blending unit in said single write-only transaction are not thereafter read by the blending unit, since section 0018 teaches the resulting blended pixels are finally written to the frame buffer and section 0070 teaches allowing for blending with the destination frame buffer data only being performed “optionally.” Ans. 22. Separately, we note Appellant’s claims 5, 13, and 20 do not distinguish over Baker alone. For example, Baker’s 3-D pixel pipeline, shown in Figure 5A, is provided on graphics device 80, shown in Figure 3, as reproduced below: Appeal 2020-004960 Application 13/725,424 14 Baker’s Figure 3 depicts graphics device 60, including pixel setup module 64, texture module 66, buffer module 68, and blending module 69 for performing multiple rendering passes and blending texture data from all the passes to be stored in buffer module 68 so as to model a 3-D image. Baker ¶¶ 33, 56–60. Baker’s frame buffers 98, shown in Figure 5A, including (1) destination frame buffer 98a and (2) temporary frame buffer 98b, are part of buffer module 68, shown in Figure 3. Similarly, Baker’s texture unit 92 and blending units 93, shown in Figure 5A, are part of texture module 68 and blending module 69, shown in Figure 3. Baker ¶¶ 59, 60. We understand that Baker’s storage element 98b, shown in Figure 5A, is labeled or designated as “temporary frame buffer” because both storage elements 98a and 98b are part of Baker’s frame buffers 98. However, Baker’s frame buffers 98 are also part of Baker’s buffer module 68, shown in Figure 3. As such, Baker’s storage element 98b could also have been Appeal 2020-004960 Application 13/725,424 15 labeled as the claimed “blend buffer” as well. Nevertheless, regardless of its label, an ordinarily skilled artisan would understand that Baker’s “temporary frame buffer 98b” serves the same function as Appellant’s claimed “blend buffer” to store the blended result from blending units 93, shown in Figure 5A, or blending module 69, shown in Figure 3. As the Supreme Court has stated, obviousness requires an “expansive and flexible” approach that asks whether the claimed improvement is more than a “predictable variation” of “prior art elements according to their established functions.” KSR, 550 U.S. at 415, 417. Here, in contrast, Appellant’s arguments rigidly focus on a narrow reading of individual prior art references, including Baker, without considering a skilled artisan’s “creatively[] and common sense.” Randall Mfg. v. Rea, 733 F.3d 1355, 1362 (Fed. Cir. 2013). Baker does not and need not expressly label its storage element 98b as Appellant’s claimed “blend buffer” in order to support a finding of obviousness. More importantly, there are only two identified, predictable storage spaces for Appellant’s claimed “blend buffer.” For example, in a graphics device shown in Baker’s Figure 3, Appellant’s claimed “blend buffer” could be part of (1) buffer module 68 or (2) blending module 69. Given the teachings of Baker’s Figures 3 and 5A, we find the use of Appellant’s claimed “blend buffer” as part of Baker’s buffer module 68 to store the blended result from blending units 93, shown in Figure 5A, or blending module 69, shown in Figure 3 would have been obvious to those ordinarily skilled artisans. In other words, a skilled artisan would have had only two choices regarding how to arrange Appellant’s claimed “blend buffer” as part Appeal 2020-004960 Application 13/725,424 16 of (1) buffer module 68 or (2) blending module 69. When two equally viable options are available, as here, then, without more, either one would seem to have been obvious. Likewise, there are only two identified, predictable blending operations available for Appellant’s claimed “blend unit.” For example, Appellant’s Admitted Prior Art (AAPA) graphics techniques require a blending unit 8, shown in Figure 1, to perform multiple “read-modify-write” operations, i.e., read pixel data from frame buffer 10, modify pixel data by applying texture data in a blending operation, and then write them back to frame buffer 10. Spec. 1:24–32. However, AAPA’s “read-modify-write” operations incur extra cost of storage and lead to “a performance penalty.” Spec. 3:3–4:5. Alternatively, such a blending unit, as shown in Baker’s Figure 1, could perform a single “write only” operation to frame buffer 19, shown in Baker’s Figure 1, where texture data from all texture passes for each pixel could be accessed and blended with pixel data prior to writing to frame buffer 19. When two equally viable options are available, as here, then, without more, either one would seem to have been obvious. In reply, Appellant raises several new arguments to explain as to why paragraphs 27, 66, and 70 of Baker do not support the Examiner’s finding that Baker teaches the disputed limitation of claims 5, 13, and 20. Reply Br. 2–5. For example, Appellant argues: the term “optionally” in paragraph 0070 of Baker would not be understood by a person of skill in the art having read Baker’s disclosure in its entirety as “teaching” that destination pixel data is not blended with texture map data. See Appeal Brief at pages 12 - 13. By contrast, the Answer has not presented any reasoned analysis of why Baker would be understood to teach what the Appeal 2020-004960 Application 13/725,424 17 examiner has alleged it teaches, but instead only states a conclusion based on an unsupported assertion. Reply Br. 3. We do not agree. First, it is well settled that mere lawyer’s arguments and conclusory statements, which are unsupported by factual evidence, are entitled to little probative value. In re Geisler, 116 F.3d 1465, 1470 (Fed. Cir. 1997); In re De Blauwe, 736 F.2d 699, 705 (Fed. Cir. 1984). Second, Baker teaches the blended results could “optionally” be blended by blending units 93, shown in Figure 5A, with destination pixel data stored in destination frame buffer 98a in order to further control the color luminous intensity values (e.g., R, G, B) when a frame of a display device is refreshed. Baker ¶ 70. However, Baker’s additional blending is only “optionally” and is not necessary to perform multiple rendering passes and to blend data from multiple texture maps that is applied to a polygon so as to model a 3-D image. For these reasons, Appellant has not persuaded us of Examiner error. Accordingly, we sustain the Examiner’s obviousness rejection of independent claims 5, 13, and 20, and their respective dependent claims 6– 12, 14–19, and 21–25 which Appellant does not argue separately. Appeal Br. 14. DECISION On the record before us, we AFFIRM the Examiner’s final rejection of (1) claims 5, 8–10, 12–15, and 17–20 under 35 U.S.C. § 103(a) as obvious over the teachings of Baker and Drebin; (2) claims 6 and 7 as Appeal 2020-004960 Application 13/725,424 18 obvious over the teachings of Baker, Drebin, and Vangemert; (3) claim 11 as obvious over the teachings of Baker, Drebin, and Mang; (4) claims 16 and 24 as obvious over the teachings of Baker, Drebin, and Van Hook; (5) claim 21 as obvious over the teachings of Baker, Drebin, and Koegel; (6) claim 22 as obvious over the teachings of Baker, Drebin, and McCarthy; (7) claim 23 as obvious over the teachings of Baker, Drebin, and Jarvis; (8) claim 25 as obvious over the teachings of Baker, Drebin, and Eckart. DECISION SUMMARY In summary: Claim(s) Rejected 35 U.S.C. § Reference(s)/Basis Affirmed Reversed 5, 8–10, 12–15, 17– 20 103 Baker, Drebin 5, 8–10, 12–15, 17– 20 6, 7 103 Baker, Drebin, Vangemert 6, 7 11 103 Baker, Drebin, Mang 11 16, 24 103 Baker, Drebin, Van Hook 16, 24 21 103 Baker, Drebin, Koegel 21 22 103 Baker, Drebin, McCarthy 22 23 103 Baker, Drebin, Jarvis 23 25 103 Baker, Drebin, Eckart 25 Appeal 2020-004960 Application 13/725,424 19 Claim(s) Rejected 35 U.S.C. § Reference(s)/Basis Affirmed Reversed Overall Outcome 5–25 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). See 37 C.F.R. § 41.50(f). AFFIRMED Copy with citationCopy as parenthetical citation