Intel Corp v. VLSI Technology LLC

1. Case Identification

• Case #: IPR2020-00583
• Patent #: 7,606,983
• Filed: February 28, 2020
• Petitioner(s): Intel Corporation
• Patent Owner(s): VLSI Technology LLC
• Challenged Claims: 1-5, 7, 9, 11-14, and 16

2. Patent Overview

• Title: Transaction Ordering Policy for Digital Systems
• Brief Description: The ’983 patent describes a digital system with multiple processors that request access to a shared resource like memory. The patent’s purported invention is a transaction ordering policy where each access request itself includes embedded ordering indications, specifying whether the request must be performed in a sequential order and, if so, what that specific order is.

3. Grounds for Unpatentability

Ground 1: Obviousness over AAPA and Khare - Claims 1-3, 5, 7, 9, 11-12, 14, and 16 are obvious over AAPA in view of Khare.

• Prior Art Relied Upon: Applicant Admitted Prior Art ("AAPA") from Patent 7,606,983 and Khare (Application # 2003/0005167).
• Core Argument for this Ground:
  • Prior Art Mapping: Petitioner argued that AAPA discloses the fundamental architecture of the claimed system, including at least two processors ("requestors") and a memory controller that manages access requests to a shared memory. However, AAPA’s fixed ordering policies, such as first-in-first-out, created performance bottlenecks. Petitioner asserted that Khare remedies this deficiency by teaching a multi-processor system where access requests are sent as data packets containing "ordering semantics." These semantics explicitly provide the two key indications required by claim 1: (1) an indication of whether the request is ordered or unordered, and (2) if ordered, an indication of the specified order (e.g., forward-ordered, backward-ordered, sequentially-ordered). Khare's "ordering semantics" field, comprising ordering bits, directly maps onto the core limitation of the challenged claims.
  • Motivation to Combine: A POSITA would combine Khare's method of embedding ordering semantics into access requests with the system described in AAPA to solve the precise performance problems AAPA identified, such as inefficient pipelining and the risk of reading "stale" data. Khare was directed to improving performance in analogous multi-node architectures, providing a clear reason to apply its teachings to AAPA’s system.
  • Expectation of Success: A POSITA would have had a high expectation of success. The combination involved applying a known technique (Khare's flexible ordering data packets) to a known system (AAPA's architecture) to achieve the predictable result of improved transaction ordering and system performance. The modification amounted to adding data fields to existing access requests, a trivial implementation.

Ground 2: Obviousness over AAPA, Khare, and Weber - Claims 4 and 13 are obvious over AAPA in view of Khare and Weber.

• Prior Art Relied Upon: AAPA (admitted prior art from Patent 7,606,983), Khare (Application # 2003/0005167), and Weber (Patent 7,149,829).
• Core Argument for this Ground:
  • Prior Art Mapping: This ground incorporated the arguments from Ground 1 and added Weber to teach the additional limitation of claims 4 and 13: including in each access request "an indication of whether or not the access request is the last access request to which the specified order applies." Petitioner contended that Weber discloses this feature by teaching the use of an "end-of-epoch marker" that is attached to transactions. This marker explicitly identifies the last access request within a group of requests, thereby providing the claimed indication.
  • Motivation to Combine: A POSITA would have been motivated to add Weber's "end-of-epoch marker" to the system of AAPA and Khare to gain additional flexibility in processing access requests. Weber teaches that its method allows for variable-sized transaction groups and enables high-value quality-of-service features without increasing controller complexity, as the marking is handled by the upstream processors. These benefits would have been desirable improvements to the AAPA/Khare system.
  • Expectation of Success: The combination was a simple application of a known technique to improve a similar device. Modifying the access requests of AAPA/Khare to include the additional "end-of-epoch" bit taught by Weber was presented as a trivial modification for a POSITA, leading to the predictable result of more flexible transaction group management.

4. Key Claim Construction Positions

• "an indication of a/the specified order": This term, present in all independent claims, was the sole term disputed in parallel district court litigation. Petitioner noted that while it had previously argued for a construction requiring "a second, different indication," the district court adopted the plain and ordinary meaning. For the petition, Petitioner applied the court's construction but argued that the claims are obvious under either construction. Specifically, Petitioner argued that Khare discloses the limitation both via a single "ordering semantics" field and, alternatively, via a combination of different fields (e.g., StreamID/NodeID plus ordering semantics), thus meeting even the more stringent "second, different indication" construction.

5. Arguments Regarding Discretionary Denial

• §314(a) (Fintiv): Petitioner argued against discretionary denial based on the parallel district court case. It asserted that the petition was efficient, as it was filed after the district court's Markman ruling, narrowing the issues for the Board. Petitioner also argued that the trial date in the co-pending litigation was uncertain and that, due to case narrowing, there was no guarantee the district court would ever rule on the specific prior art combinations presented in the petition.
• §325(d): Petitioner argued that denial under §325(d) was unwarranted because the primary prior art references, AAPA and Khare, presented new and non-cumulative arguments not previously before the Patent Office. During prosecution, the claims were allowed over Weber. Petitioner contended that Khare is materially different from Weber because Khare, unlike Weber, explicitly teaches access requests that include both an indication of whether an access request is ordered and, if so, the specific type of order—the very features the Examiner found lacking in Weber.

6. Relief Requested

• Petitioner requested institution of an inter partes review and cancellation of claims 1-5, 7, 9, 11-14, and 16 of the ’983 patent as unpatentable.