Google LLC v. Sandpiper CDN LLC

1. Case Identification

• Case #: IPR2025-00860
• Patent #: 10,924,573
• Filed: May 1, 2025
• Petitioner(s): Google LLC
• Patent Owner(s): Sandpiper CDN, LLC
• Challenged Claims: 1-20

2. Patent Overview

• Title: Content Delivery Network with Popularity Service
• Brief Description: The ’573 patent describes a content delivery network (CDN) that uses a hierarchical, multi-tier server architecture (e.g., edge, parent, and origin servers) to deliver content. The system incorporates a "popularity service" to determine how and where content is cached and served based on its popularity, with the goal of improving efficiency.

3. Grounds for Unpatentability

Ground 1: Anticipation by Newton-471 - Claims 1-20 are anticipated by Newton-471.

• Prior Art Relied Upon: Newton-471 (Application # 2016/0094471).
• Core Argument for this Ground:
  • Prior Art Mapping: Petitioner asserted that Newton-471, which published in 2016, discloses every limitation of the challenged claims. The specification of Newton-471 is nearly identical to that of the ’573 patent but crucially includes express disclosure and figures (e.g., Figure 8) detailing the "instruction to not cache" limitation. This limitation was central to the allowance of the ’573 patent claims. Petitioner contended that Newton-471 explicitly teaches a mid-tier server returning a "no-cache instruction" for unpopular content, which is then received by the edge server.
  • Key Aspects: The foundation of this ground is a challenge to the ’573 patent’s priority date. Petitioner argued that the "instruction to not cache" limitation was new matter introduced into the application that issued as the ’573 patent on its filing date of February 25, 2019. Because this limitation lacks written description support in the parent applications, the ’573 patent is not entitled to its claimed 2008 priority date. With a 2019 priority date, the 2016 publication of Newton-471 becomes anticipating prior art under 35 U.S.C. §102.

Ground 2: Obviousness over Dilley, Pai, and Wang - Claims 1-20 are obvious over Dilley in view of Pai and Wang.

• Prior Art Relied Upon: Dilley (Patent 7,133,905), Pai (Application # 2006/0271972), and Wang (Application # 2005/0198250).
• Core Argument for this Ground:
  • Prior Art Mapping: Petitioner argued the combination of these references teaches all limitations of the challenged claims. Dilley provided the foundational tiered CDN architecture, including edge, parent, and origin servers, and taught using metadata to manage content distribution. Pai supplied the specific teachings of a dynamic "popularity service" that uses popularity values and thresholds to redistribute content across a tiered network for efficiency. Wang supplied the critical "instruction to not cache" limitation, teaching the use of cache control attributes to identify content that should not be cached by lower-level servers to improve performance and avoid storing stale or unpopular content.
  • Motivation to Combine: A POSITA would combine Dilley and Pai to implement a known, dynamic popularity-based management system (Pai) within a standard tiered CDN (Dilley) to predictably improve content delivery efficiency. Dilley’s general teaching of using metadata would motivate a POSITA to look to specific, known implementations like Pai's popularity service. A POSITA would further incorporate Wang's teachings on non-cacheable content attributes into the Dilley/Pai system as a straightforward method to optimize storage and avoid caching unnecessary content, a well-known goal in CDN design.
  • Expectation of Success: A POSITA would have a reasonable expectation of success in combining these references. The combination involved applying known techniques (popularity-based distribution, no-cache instructions) to a conventional tiered CDN architecture to achieve the predictable results of improved performance, faster delivery speeds, and more efficient resource utilization.

4. Key Technical Contentions (Beyond Claim Construction)

• Priority Date Challenge: The central technical and legal contention of the petition is that the ’573 patent is not entitled to its claimed April 4, 2008 priority date. Petitioner argued the key limitation "receiving an instruction to not cache," which was added during prosecution to secure allowance, constitutes new matter that lacks written description support in any of the patent’s priority documents. Consequently, the effective priority date for all challenged claims is the application’s filing date of February 25, 2019, which renders several references, particularly Newton-471, available as prior art.

5. Arguments Regarding Discretionary Denial

• §314(a) (Fintiv Factors): Petitioner argued against discretionary denial under Fintiv, stating the factors strongly favor institution. The parallel district court litigation is in its early stages, with a trial date (September 2026) that is very close to the projected date for a Final Written Decision (November 2026). Furthermore, Petitioner stipulated that if IPR is instituted, it will not pursue the same grounds or any grounds that reasonably could have been raised in the district court action, mitigating concerns of duplicative efforts.
• §325(d) (Same or Substantially the Same Art/Arguments): Petitioner contended that denial under §325(d) is inappropriate. For Ground 1, the Examiner did not apply Newton-471 because it was not prior art under the incorrect, earlier priority date considered during prosecution. For Ground 2, the key combining references of Pai and Wang were never before the Examiner. While Dilley was listed in an IDS, there is no evidence the Examiner substantively considered its specific teachings in combination with Pai and Wang.

6. Relief Requested

• Petitioner requests institution of an inter partes review and cancellation of claims 1-20 of Patent 10,924,573 as unpatentable.