PTAB

IPR2017-01225

LG Electronics Inc v. ATI Technologies ULC

Key Events
Petition
petition

1. Case Identification

2. Patent Overview

  • Title: Graphics Processing Architecture with Unified Shader
  • Brief Description: The ’454 patent describes a graphics processing architecture that utilizes a single "unified shader" to perform both vertex and pixel operations, replacing the traditional approach of using separate, dedicated shaders for each operation type.

3. Grounds for Unpatentability

Ground 1: Claims 2-11 are anticipated by Lindholm under 35 U.S.C. §102.

  • Prior Art Relied Upon: Lindholm (Patent 7,038,685).
  • Core Argument for this Ground:
    • Prior Art Mapping: Petitioner argued that Lindholm discloses every limitation of the challenged claims. The "Execution Pipeline 240" in Lindholm was asserted to be the claimed "unified shader" because it is a programmable pipeline that processes both vertex samples from a Vertex Input Buffer and pixel samples from a Pixel Input Buffer. Lindholm's "Execution Unit 470" was mapped to the claimed "processor unit," its "Register File 350" to the "general purpose register block" and "shared resources," and its combination of an Instruction Scheduler, Dispatcher, and Cache to the claimed "sequencer." Petitioner contended that Lindholm’s resource-based thread allocation priority scheme, which manages processing based on resource availability in the Register File, meets the functional limitations of claims 3, 4, and 5 regarding switching between pixel and vertex operations based on available resources.

Ground 2: Claims 2-11 are obvious over Lindholm, alone or in combination with OpenGL, under 35 U.S.C. §103.

  • Prior Art Relied Upon: Lindholm (Patent 7,038,685) and OpenGL (OpenGL Graphics System: A Specification, Version 1.4).
  • Core Argument for this Ground:
    • Prior Art Mapping: Petitioner presented this ground as an alternative, arguing that if Lindholm's "Execution Unit 470" was found not to explicitly execute instructions to generate pixel color or vertex position data (as recited in claims 2, 8, and 9), it would have been obvious to implement this functionality. Lindholm teaches performing operations like texture mapping, shading, and blending, which a person of ordinary skill in the art (POSA) would understand inherently involve generating pixel color and vertex data. Furthermore, OpenGL, a well-known graphics standard, expressly disclosed generating pixel color via texture maps and generating transformed vertex position and appearance data in its processing pipeline.
    • Motivation to Combine: A POSITA would combine Lindholm with the teachings of OpenGL to improve processor utilization, enhance the flexibility of processing different data types, apply realistic texture features, and simplify graphics manipulation by using a standard, well-known API.
    • Expectation of Success: Because both Lindholm and OpenGL operate in the same field of 3D graphics processing, a POSITA would have had a high expectation of success in applying OpenGL's standard operations to Lindholm's flexible architecture, achieving predictable results.

Ground 3: Claims 2 and 11 are anticipated by Stuttard under 35 U.S.C. §102.

  • Prior Art Relied Upon: Stuttard (WO 00/62182).
  • Core Argument for this Ground:
    • Prior Art Mapping: Petitioner argued that Stuttard’s graphics processing system anticipates the key features of claims 2 and 11. Stuttard's "processing core 10," which includes processing elements that perform both geometry processing on vertex data and subsequent operations on pixel data, was identified as the claimed "unified shader." The "PE memory unit 1061c" and "PE register file 1061b" were mapped to the "general purpose register block." The "array controller 104," which directs the operation of the processing block, was asserted to be the claimed "sequencer," and the "instruction launcher 1041" and "instruction table 1042" were mapped to the "instruction store." Petitioner further contended that Stuttard’s multithreading system, which halts and switches between active threads, inherently performs operations at "various degrees of completion" as required by claim 11.
  • Additional Grounds: Petitioner asserted additional obviousness challenges based on Stuttard alone, arguing that any missing elements for claims 3-10 would have been obvious modifications to improve processing efficiency and resource utilization, consistent with Stuttard's stated goals.

4. Arguments Regarding Discretionary Denial

  • Petitioner argued that the Patent Owner is collaterally estopped from asserting that it is entitled to a priority date that would antedate the Lindholm reference. Petitioner stated this exact issue was fully litigated and resolved against the Patent Owner in IPR2015-00325 and IPR2015-00326, which involved the parent of the ’454 patent. Because the Board previously rejected the Patent Owner's arguments to "swear behind" Lindholm in a related case, Petitioner asserted the Patent Owner should be precluded from re-litigating the issue, thus confirming Lindholm's status as valid prior art against the challenged claims.

5. Relief Requested

  • Petitioner requests institution of an inter partes review and cancellation of claims 2-11 of the ’454 patent as unpatentable.