VMware Inc v. WSOU Investments LLC

1. Case Identification

• Case #: IPR2021-00572
• Patent #: 9,164,800
• Filed: February 22, 2021
• Petitioner(s): VMware, Inc., Dell Technologies Inc., Dell Inc., and EMC Corporation
• Patent Owner(s): WSOU Investments, LLC d/b/a Brazos Licensing and Development
• Challenged Claims: 1-2, 6-9, 13-14, 18-21

2. Patent Overview

• Title: Cloud Controller and Method for Intelligently Placing Compute Nodes
• Brief Description: The ’800 patent discloses a system for reducing network latency in distributed computing environments, such as cloud data centers. The invention addresses this by intelligently assigning "compute nodes" (e.g., virtual machines) to "data nodes" (e.g., storage servers) that are geographically or network-wise "closer," treating the task as an "assignment problem" to be solved by an algorithm based on a selected assignment objective.

3. Grounds for Unpatentability

Ground 1: Obviousness over Queck - Claims 1-2, 6-9, 13-14, and 18-21 are obvious over Queck.

• Prior Art Relied Upon: Queck (EP Patent No. 1978445).
• Core Argument for this Ground:
  • Prior Art Mapping: Petitioner argued that Queck, which describes distributing data and task instances in a grid environment, teaches all limitations of the challenged claims. Queck’s "central node" functions as the claimed "cloud controller," and its "processing nodes" and "data nodes" are analogous to the claimed "compute nodes" and "data nodes." The central node assigns tasks to nodes to optimize performance, considering factors like latency. Petitioner contended that the claimed "edges" between nodes are taught by Queck’s disclosure of a "latency matrix," where the matrix positions represent possible connections and their associated costs (latencies).
  • Motivation to Combine: The petition asserted that Queck discloses multiple "assignment objectives," such as optimizing performance for a single task instance versus optimizing the average performance across multiple instances. A POSITA would have found it obvious to configure Queck's system to select between these objectives based on the characteristics of the task being deployed (e.g., a latency-sensitive streaming task versus a batch data processing task). This modification would be a simple design choice to improve the system's flexibility and overall efficiency, which are known goals in the art.
  • Expectation of Success: A POSITA would have had a high expectation of success in implementing this selection logic, as it involved applying known programming techniques to a system with already-defined optimization goals.

Ground 2: Obviousness over Queck and Singh - Claims 1-2, 6-9, 13-14, and 18-21 are obvious over Queck in view of Singh.

• Prior Art Relied Upon: Queck (EP Patent No. 1978445) and Singh (IBM Research Report, Apr. 18, 2007).
• Core Argument for this Ground:
  • Prior Art Mapping: This ground was presented as an alternative, arguing that if Queck is found not to teach selecting from multiple assignment objectives, Singh provides the missing element. Singh discloses a framework called "SPARK" for placing application CPU and data resources in a virtualized data center. Critically, Singh teaches that its framework is not tied to a single cost function and can accommodate various "user-controllable cost or distance functions" to achieve specific, administrator-defined objectives, such as satisfying "hard latency requirements" or application-specific preferences.
  • Motivation to Combine: A POSITA would combine Singh’s flexible, versatile framework for defining assignment objectives with Queck’s grid computing system. The motivation, as stated in Singh, is that such flexibility is a practical necessity in real-world data centers to adapt to changing policies and application needs. Combining the two would have been the application of a known technique (Singh’s flexible cost functions) to a known system (Queck) to achieve the predictable result of improved adaptability and performance.
  • Expectation of Success: Success would be expected because Singh provides the necessary algorithms and framework that a POSITA could readily integrate into Queck's system to manage and minimize any of the cost functions Singh describes.

Ground 3: Obviousness over Li and Singh - Claims 1, 8-9, 13, and 20-21 are obvious over Li in view of Singh.

• Prior Art Relied Upon: Li (Patent 8,972,983) and Singh (IBM Research Report, Apr. 18, 2007).
• Core Argument for this Ground:
  • Prior Art Mapping: Petitioner argued that Li discloses a cloud computing infrastructure with a "cloud manager" or "director" that functions as the claimed cloud controller. Li’s system intelligently assigns jobs to virtual machines (compute nodes) based on the location of their required data blocks (data nodes) to minimize access times. Li explicitly teaches modeling this as a graph problem with nodes, edges, and associated costs (e.g., "virtual machine closeness"), directly mapping to the core elements of the independent claims.
  • Motivation to Combine: A POSITA would have been motivated to incorporate Singh's framework of multiple, user-controllable cost functions into Li’s system to enhance its flexibility. This combination would allow Li’s director to select assignment objectives based on factors beyond just proximity, such as real-time network load or application priority, which Singh teaches is a valuable capability for improving overall network performance.
  • Expectation of Success: The integration would have been straightforward for a POSITA, as Singh's framework is designed for versatility and provides the algorithms necessary to implement various cost-based placement strategies within a system like Li’s.

4. Arguments Regarding Discretionary Denial

• Petitioner argued that the PTAB should not exercise discretionary denial under 35 U.S.C. §314(a) based on Fintiv factors. The district court trial schedule is highly uncertain, as the ’800 patent is one of twelve patents in parallel litigation and is scheduled to be the ninth trial, making the earliest trial date of May 2, 2022, unrealistic. Furthermore, Petitioner filed a stipulation agreeing not to pursue in the district court any invalidity grounds that were raised or could have been reasonably raised in the IPR, ensuring the IPR is a "true alternative" to litigation and mitigating concerns of duplicative efforts.

5. Relief Requested

• Petitioner requested the institution of an inter partes review and the cancellation of claims 1-2, 6-9, 13-14, and 18-21 of the ’800 patent as unpatentable.