Google LLC v. Sandpiper CDN LLC

1. Case Identification

• Case #: IPR2025-01010
• Patent #: 10,057,322
• Filed: May 23, 2025
• Petitioner(s): Google LLC
• Patent Owner(s): Sandpiper CDN, LLC
• Challenged Claims: 1-15

2. Patent Overview

• Title: Network Address Resolution
• Brief Description: The ’322 patent discloses systems for selecting an edge server within a Content Delivery Network (CDN). The invention centers on an origin server that communicates with a CDN via an application programming interface (API) to obtain a direct link to a specific CDN node, which is then provided to a client device for direct content retrieval.

3. Grounds for Unpatentability

Ground 1: Obviousness/Anticipation over Verma - Claims 11, 12, 14, 15 are obvious over Verma; Claim 15 is anticipated by or obvious over Verma.

• Prior Art Relied Upon: Verma (Application # 2009/0132640).
• Core Argument for this Ground:
  • Prior Art Mapping: Petitioner argued that Verma discloses all elements of the challenged claims. Verma teaches a "content provider" (origin system) that communicates with a CDN "broker" (API) to optimize content delivery. Upon receiving a request for a content file (e.g., a URL), the content provider forwards the request, including "delay tolerance information" (a delivery parameter), to the broker. The broker then generates a "re-written URL" (modified embedded resource) that provides a direct link to a specific "cache server" (CDN node) and "returns" this modified URL to the content provider. The provider then delivers the webpage containing this modified URL to the client, which uses it to fetch content directly from the designated cache server.
  • Motivation to Combine (for §103): To the extent Verma does not explicitly disclose embedding the resource URL within a webpage, Petitioner argued a POSITA would find it obvious to do so. Embedding URLs was the standard, conventional method for providing clickable links, and doing so would enable Verma’s system to function as intended in a predictable manner.
  • Expectation of Success (for §103): A POSITA would have reasonably expected success, as embedding and rewriting URLs for CDN optimization were well-known techniques used to improve web performance. The outcome of applying these conventional methods to Verma's system was entirely predictable.

Ground 2: Obviousness over Verma and Raciborski - Claims 11, 12, 14, 15 are obvious over Verma in view of Raciborski.

• Prior Art Relied Upon: Verma (Application # 2009/0132640) and Raciborski (Application # 2007/0067424).
• Core Argument for this Ground:
  • Prior Art Mapping: Petitioner asserted that Raciborski supplements Verma by explicitly teaching a "content originator" sending a "bid message" to a CDN via an "analysis interface" (API). This message includes the "full object location" (embedded resource) and information on whether "delivery can be delayed" (an associated parameter). Raciborski further discloses replacing an original URI with a rewritten URI that is then embedded in a webpage for delivery to a client, clarifying any potential ambiguity in Verma.
  • Motivation to Combine: A POSITA would combine Verma and Raciborski because both are analogous art addressing the common goal of improving CDN performance. Raciborski's explicit disclosure of passing the full object location and delivery parameters to the CDN API provides a clear and obvious blueprint for implementing Verma's more general description of a broker that rewrites URLs based on delivery criteria.
  • Expectation of Success: Success was expected because the systems disclosed in both references are highly similar, comprising the same core components (origin system, API, CDN) and objectives. Integrating Raciborski’s explicit parameter-passing mechanism into Verma's framework represented a predictable combination of known elements.

Ground 3: Obviousness over Verma and Lewin - Claims 1, 8, 10, 13 are obvious over Verma in view of Lewin.

• Prior Art Relied Upon: Verma (Application # 2009/0132640) and Lewin (Application # 2014/0052811).
• Core Argument for this Ground:
  • Prior Art Mapping: Petitioner argued that Lewin teaches the "tag" limitation of claim 1 by disclosing the use of HTML tags (e.g., <esi:include>) to specify URLs for dynamic content assembly. A server "parses the HTML page looking for tags," which constitutes "identification of the tag" upon receiving a request for the content resource (the HTML page). Lewin's teaching of conditional tags that serve different content based on a user's geographic location further addresses the "unique tag" limitation of dependent claim 8.
  • Motivation to Combine: A POSITA would combine Lewin's tag-based system with Verma's CDN architecture to provide a standard, efficient method for identifying which embedded resources require modification. Using HTML tags to flag URLs for server-side processing was a well-known, conventional web development practice and a natural way to implement Verma’s URL rewriting function.
  • Expectation of Success: A POSITA would have expected success because incorporating tag-based URL identification into Verma’s system was a routine application of conventional web technologies. The integration would predictably allow Verma’s broker to identify and modify the correct URLs.
• Additional Grounds: Petitioner asserted further obviousness challenges for dependent claims. Ground 4 added Raciborski to the Verma/Lewin combination to teach using a client IP address as a delivery parameter. Ground 5 added Drai (Application # 2009/0172167) to teach using content popularity and geographic information as delivery parameters.

4. Arguments Regarding Discretionary Denial

• Petitioner argued that discretionary denial under both §314(a) and §325(d) is unwarranted.
• Fintiv Factors: The factors weigh in favor of institution. The co-pending district court trial is scheduled for late 2026, which is concurrent with the anticipated Final Written Decision deadline. Further, discovery is in its early stages, and Petitioner stipulated that it will not pursue in court the same grounds raised in the IPR.
• §325(d) Factors: The primary prior art references relied upon in the petition (Verma, Raciborski, Lewin, and Drai) were not applied or considered during the original prosecution. Therefore, the petition raises substantial new questions of patentability that the USPTO has not previously addressed.

5. Relief Requested

• Petitioner requests institution of an inter partes review and cancellation of claims 1-15 of Patent 10,057,322 as unpatentable.