DCT
2:25-cv-00664
Sandpiper CDN LLC v. Microsoft Corp
I. Executive Summary and Procedural Information
- Parties & Counsel:
- Plaintiff: Sandpiper CDN, LLC (Delaware)
- Defendant: Microsoft Corporation (Delaware)
- Plaintiff’s Counsel: Boies Schiller Flexner LLP
- Case Identification: Sandpiper CDN, LLC v. Microsoft Corporation, 2:25-cv-00664, E.D. Tex., 10/30/2025
- Venue Allegations: Plaintiff alleges venue is proper in the Eastern District of Texas based on Defendant’s regular and established places of business within the district, including corporate offices, retail operations within Best Buy stores, data servers at data centers, and a confirmed "point of presence (POP) location" for its Azure Content Delivery Network in Plano, Texas.
- Core Dispute: Plaintiff alleges that Defendant’s content delivery network (CDN) products and services, including Microsoft Azure CDN and related cloud infrastructure, infringe six patents related to foundational CDN technology for routing, caching, and delivering internet content.
- Technical Context: The technology at issue is Content Delivery Networks (CDNs), a critical infrastructure for the modern internet that accelerates the delivery of web content, streaming video, and software updates to end-users by distributing it on servers closer to the user's geographic location.
- Key Procedural History: The complaint details a complex corporate history, alleging the patents originated with CDN pioneer Sandpiper Networks in the 1990s and passed through a series of acquisitions (Digital Island, Savvis, Level 3) to the current Plaintiff. It alleges that Microsoft, a major customer of former patent owners, attempted to acquire the patent portfolio in 2006 but was outbid. Plaintiff also notes its recent, related lawsuits against Google and Comcast and alleges Microsoft was aware of this litigation through its membership in the defensive patent aggregator RPX.
Case Timeline
| Date | Event |
|---|---|
| 1996-05-24 | Sandpiper Networks develops streaming infrastructure |
| 1998-02-10 | '903 Patent Priority Date |
| 1998-09-11 | Sandpiper Networks' first paying customer begins service |
| 1999-12-01 | Sandpiper Networks merges with Digital Island |
| 2002-02-14 | '466 Patent Priority Date |
| 2006-12-01 | Level 3 announces acquisition of Savvis CDN business/patents |
| 2008-04-04 | '692 Patent Priority Date |
| 2010-02-01 | Microsoft Azure CDN launched |
| 2011-12-14 | '053 Patent Priority Date |
| 2012-12-13 | '883 Patent Priority Date |
| 2013-07-02 | '903 Patent Issued |
| 2014-01-01 | Azure Resource Manager introduced |
| 2014-12-15 | '173 Patent Priority Date |
| 2014-12-30 | '466 Patent Issued |
| 2016-03-31 | Azure Service Fabric becomes generally available |
| 2016-09-27 | '053 Patent Issued |
| 2017-08-01 | '883 Patent Issued |
| 2017-09-12 | '692 Patent Issued |
| 2018-05-01 | Microsoft announces its own CDN |
| 2019-11-04 | Azure Arc announced |
| 2020-06-30 | '173 Patent Issued |
| 2025-10-30 | Complaint Filing Date |
II. Technology and Patent(s)-in-Suit Analysis
U.S. Patent No. 8,478,903 - "Shared Content Delivery Infrastructure"
- Issued: July 2, 2013
The Invention Explained
- Problem Addressed: The patent’s background describes how the rapid growth of the internet created performance issues, such as slow content delivery, due to network congestion and the distance between users and origin servers ('903 Patent, col. 1:27-31).
- The Patented Solution: The invention proposes a system to "off-load processing of requests" by replicating resources from multiple content providers' origin servers onto a network of shared "repeater servers" (i.e., CDN caches) ('903 Patent, col. 2:62-65). Requests from end-users are directed to these shared repeater servers using "alias names," which allows a single server infrastructure to deliver content on behalf of multiple, distinct websites or content providers ('903 Patent, col. 4:37-40).
- Technical Importance: This architecture was a foundational approach to solving the internet's early scaling problems, enabling faster and more reliable content access by distributing the load from central origin servers to a geographically dispersed network. (Compl. ¶¶ 21-22).
Key Claims at a Glance
- The complaint asserts at least independent claim 28 (Compl. ¶112).
- Claim 28 (System):
- A content delivery system with at least one shared repeater server for replicating resources associated with a plurality of origin servers.
- The system associates a first resource on a first origin server with a first alias name and a second resource on a second origin server with a second, distinct alias name.
- Requests for the first resource are directed to the repeater server based on the first alias name, and requests for the second resource are directed to the repeater server based on the second alias name.
- The repeater server utilizes a table to correlate itself to the plurality of origin servers.
U.S. Patent No. 8,924,466 - "Server handoff in content delivery network"
- Issued: December 30, 2014
The Invention Explained
- Problem Addressed: The patent addresses the impracticality of storing all content objects (e.g., large video files) at all edge servers in a CDN due to storage limitations. This creates a need for intelligent object replication and a system to handle requests for content not located at the initial edge server. ('466 Patent, col. 1:52-2:4).
- The Patented Solution: The invention discloses a multi-tier "handoff" method. A client request is first sent to an edge server based on network conditions. If that server does not have the content, the client is redirected to a second server, such as a parent server deeper in the network. If the second server also lacks the content, the client is directed to another server, creating a hierarchical lookup process to find and deliver the object. ('466 Patent, Abstract; Claim 1).
- Technical Importance: This tiered architecture allows CDNs to efficiently manage vast content libraries by balancing the performance benefits of edge caching for popular content with the cost-effectiveness of centralized storage for less-frequently accessed "long-tail" content. (Compl. ¶84).
Key Claims at a Glance
- The complaint asserts at least independent claim 1 (Compl. ¶117).
- Claim 1 (Method):
- Providing a plurality of edge servers and parent servers distinct from an origin server.
- Directing a client request to a first edge server based on network traffic conditions, regardless of whether that server has the object.
- If the first edge server has the object, serving it.
- Otherwise, redirecting the client to a second server (e.g., a parent server) to handle the request.
- If the second server does not have the object, directing the client to another server in the CDN.
U.S. Patent No. 9,456,053 - "Content delivery network"
- Issued: September 27, 2016
- Technology Synopsis: This patent addresses the technical challenge of keeping configuration data updated across a scaled, capacity-constrained CDN (Compl. ¶87). The patented solution involves cache servers obtaining global configuration data from a central "control core," selectively updating this data, and retrieving customer-specific configuration information to deliver tailored content based on the combined data ('053 Patent, Abstract; Compl. ¶89).
- Asserted Claims: At least Claim 27 (Compl. ¶122).
- Accused Features: The complaint alleges that Azure CDN cache servers obtain global configuration data (e.g., routing policies, caching rules) from a control core, identified as Azure Resource Manager, and combine it with customer-specific rule sets to serve content. (Compl. ¶¶ 123-125).
U.S. Patent No. 9,722,883 - "Responsibility-based peering"
- Issued: August 1, 2017
- Technology Synopsis: This patent addresses the problem of managing content libraries that are too large for a single cache server by configuring nodes in a "peer group" to handle discrete responsibilities (Compl. ¶91). The solution uses configurable peering policies to map requests to nodes based on their capabilities and capacity, utilizing consistent hashing to minimize disruption when nodes fail or capacities fluctuate ('883 Patent, Abstract; Compl. ¶¶ 92-93).
- Asserted Claims: At least Claim 1 (Compl. ¶127).
- Accused Features: Plaintiff accuses Microsoft's Azure Service Fabric, which is described as a platform for managing scalable microservices where clusters of virtual machines assume discrete responsibilities to collaboratively process requests. (Compl. ¶¶ 127-128).
U.S. Patent No. 9,762,692 - "Handling long-tail content in a content delivery network (CDN)"
- Issued: September 12, 2017
- Technology Synopsis: This patent addresses the technical issues related to serving "long-tail" content that can dynamically become popular or fade into obscurity (Compl. ¶95). The invention describes a tiered server system where a first-tier server determines if a resource is popular. If it is popular, the server obtains and serves it. If not, the requester is directed to a second, distinct tier of servers that serves the less-popular resource. ('692 Patent, Abstract; Compl. ¶96).
- Asserted Claims: At least Claim 1 (Compl. ¶133).
- Accused Features: The complaint accuses Azure CDN's use of a tiered server system where first-tier servers determine content popularity (e.g., via cache eviction policies) and, if the content is deemed not popular, direct the client to a second tier of servers for fulfillment. (Compl. ¶¶ 134-135).
U.S. Patent No. 10,701,173 - "Caching in a content delivery framework"
- Issued: June 30, 2020
- Technology Synopsis: This patent addresses the costly process of invalidating cached data when caching policies change (Compl. ¶100). The solution employs "late-binding," where a CDN node, upon receiving a request, dynamically determines the current caching policy and evaluates the locally cached content against it. This allows the node to serve the content if valid or retrieve a fresh version if stale, applying up-to-date rules without forcing a network-wide cache invalidation. ('173 Patent, Abstract; Compl. ¶101).
- Asserted Claims: At least Claim 1 (Compl. ¶137).
- Accused Features: Plaintiff accuses Azure CDN's policy-based caching system, which offers caching rules that determine expiration periods. The complaint alleges that CDN nodes use these rules to determine whether it is acceptable to serve cached content or if a new version must be obtained. (Compl. ¶¶ 137-138).
III. The Accused Instrumentality
Product Identification
The complaint collectively refers to the accused products as the "Accused Microsoft Functionalities" (Compl. ¶110). These include Microsoft's Azure cloud computing platform services, specifically the Azure Content Delivery Network (including "Standard from Microsoft," "Standard from Edgio," and "Premium from Edgio"), Azure Front Door, Azure Service Fabric, Azure Arc, Azure Policy, and Azure Resource Manager. (Compl. ¶¶ 102, 104-108).
Functionality and Market Context
The accused products constitute a global CDN that leverages a distributed network of Points of Presence (PoPs) containing "EdgeNodes" or "Endpoints" to cache and deliver content closer to end-users (Compl. ¶103). The system employs a multi-tier architecture, including Edge POPs and "Regional Cache POPs," to manage content distribution (Compl. ¶117). Key technical features identified in the complaint include performance and streaming optimization, customizable caching rules, URL redirection, and load balancing (Compl. ¶¶ 103, 104). The complaint alleges that Microsoft's "Intelligent Cloud" segment, driven by these Azure services, is its largest source of profit and that the CDN is used for both internal Microsoft services (e.g., Windows updates, Xbox) and leased to external customers for a fee. (Compl. ¶¶ 65, 67, 69).
IV. Analysis of Infringement Allegations
8,478,903 Infringement Allegations
| Claim Element (from Independent Claim 28) | Alleged Infringing Functionality | Complaint Citation | Patent Citation |
|---|---|---|---|
| a content delivery system comprising at least one shared repeater server...for replicating resources associated with a plurality of origin servers | Microsoft Azure CDN comprises shared repeater servers (e.g., EdgeNodes or Endpoints) that replicate resources associated with the origin servers of multiple customers. | ¶112 | col. 4:37-40 |
| associating a first resource located on a first origin server with a first alias name...directing...requests...based at least in part on said first alias name | Microsoft Azure CDN provides instructions for users to create alias records (e.g., custom domains via CNAME DNS records) that associate a user's domain with an Azure CDN endpoint for request delivery. A Microsoft tutorial screenshot shows this process. (Compl. p. 34). | ¶113 | col. 2:62-65 |
| associating a second resource on a second origin server with a second alias name | The Azure CDN supports custom domain alias creation for multiple distinct CDN hosting clients that access content from various origin servers. The complaint provides a screenshot of customer logos. (Compl. p. 35). | ¶114 | col. 4:37-40 |
| utilizing a table to correlate said at least one shared repeater server to said plurality of origin servers | Microsoft Azure CDN utilizes a table to correlate its shared repeater servers (edge infrastructure) to the respective origin servers for its multiple CDN customers. | ¶115 | col. 2:54-65 |
- Identified Points of Contention:
- Scope Questions: A potential point of contention may be the term "shared repeater server." The defense may argue that modern cloud architectures, through virtualization or other means, create logically separate environments for each customer, raising the question of whether Microsoft’s EdgeNodes are "shared" in the manner contemplated by the patent. The interpretation of "table" may also be disputed, as Microsoft's implementation could be a distributed database or other mechanism rather than a simple table structure.
8,924,466 Infringement Allegations
| Claim Element (from Independent Claim 1) | Alleged Infringing Functionality | Complaint Citation | Patent Citation |
|---|---|---|---|
| providing a plurality of edge servers and a plurality of parent servers distinct from...at least one origin server | Microsoft Azure CDN comprises edge servers (Edge POPs) and parent servers (Regional Cache POPs) which are distinct from customer origin servers. A network map illustrates this tiered structure. (Compl. p. 40). | ¶117 | col. 1:52-54 |
| directing a client request for an object to a first edge server based at least in part on network traffic conditions, regardless of whether the first edge server has the requested object | Azure CDN's "Performance traffic-routing method" directs client requests to an endpoint (edge server) based on network latency, which is a measure of traffic conditions. A diagram depicts this latency-based routing. (Compl. p. 41). | ¶118 | col. 2:5-8 |
| if the first edge server has the requested object, serving the requested object to the client from the first edge server | If the first edge server to which the client is directed has a cached copy of the object, it is served directly from that server. | ¶119 | col. 2:9-10 |
| otherwise, redirecting the client to a second server in the content delivery network...to handle the request | If the edge server has a cache miss, the request is handled by a second server, such as a parent POP ("origin shield"). A table from Microsoft's documentation notes a "miss in the edge (child POP)" is "responded from origin shield (parent POP)." (Compl. p. 42). | ¶119 | col. 2:10-12 |
| if the requested object is not stored on the second server, directing the client to another server in the CDN...to serve the requested object | If the object is not on the second server (e.g., parent POP), the client is ultimately directed to another server, such as the origin server, to obtain the object. A diagram shows a workflow where a cache miss at the POP results in a request to the origin server. (Compl. p. 43). | ¶120 | col. 2:12-14 |
- Identified Points of Contention:
- Technical Questions: A significant technical question arises from the term "redirecting the client." The patent claim requires the client to be redirected. However, the complaint's own evidence (Compl. p. 42) suggests that in the case of a cache miss, the edge server may fetch the content from the parent server itself (a reverse-proxy action) rather than sending a redirect message (e.g., an HTTP 302) to the client. This raises the question of whether Microsoft’s server-side fetch mechanism meets the specific "redirecting the client" limitation.
V. Key Claim Terms for Construction
"shared repeater server" ('903 Patent, Claim 28)
- Context and Importance: This term is the central component of the claimed system. Its construction is critical because the dispute may hinge on whether Microsoft's multi-tenant cloud infrastructure, which may use virtualization to segregate customer data and traffic, constitutes a "shared" server in the manner claimed by the patent.
- Intrinsic Evidence for Interpretation:
- Evidence for a Broader Interpretation: The specification describes a repeater server as one that "maintains a partial mirror of more than one origin server" ('903 Patent, col. 4:37-38). This language may support an interpretation where any single physical or logical server instance that caches content from multiple, distinct customers qualifies as "shared."
- Evidence for a Narrower Interpretation: The description of implementing "a distributed and/or coherent cache" ('903 Patent, col. 4:39) could be used to argue for a narrower definition that requires a specific, technically integrated caching architecture, which Microsoft might argue its system does not employ.
"redirecting the client" ('466 Patent, Claim 1)
- Context and Importance: This term defines the handoff mechanism between server tiers. Its construction is likely to be a dispositive issue for infringement of the ’466 Patent. Practitioners may focus on this term because there is a fundamental technical difference between a client-side redirect (where the client is instructed to make a new request) and a server-side fetch (where the first server retrieves content on the client's behalf).
- Intrinsic Evidence for Interpretation:
- Evidence for a Broader Interpretation: The patent's abstract states the "first server redirects the client request," which could be argued to broadly cover any action that causes the request to be fulfilled by a different server, regardless of the underlying network transaction. ('466 Patent, Abstract).
- Evidence for a Narrower Interpretation: The patent family's disclosure in the parent '903 patent explicitly discusses using a "REDIRECT message," which is a specific client-side instruction ('903 Patent, col. 8:34-39). This suggests the possibility that "redirecting the client" requires an action that directs the client itself, not just its request internally within the network.
VI. Other Allegations
- Indirect Infringement: The complaint alleges that Microsoft induces and contributes to infringement by operating its CDN and providing its customers with documentation, user interfaces, and APIs that instruct and enable them to use the accused functionalities in an infringing manner, such as configuring custom domains and caching rules. (Compl. ¶¶ 111, 116, 121, 126, 132, 136).
- Willful Infringement: The complaint alleges willful infringement based on both pre- and post-suit knowledge. It alleges pre-suit knowledge stemming from Microsoft's 2006 bid to acquire the patent portfolio from a predecessor, its status as a major customer of services that used the patented technology, its hiring of personnel from predecessors, and its citation of one of the patent applications during its own patent prosecution (Compl. ¶¶ 48, 53, 54, 61-62). It further alleges knowledge based on notices Microsoft allegedly received via its RPX membership regarding Plaintiff's recent lawsuits against other major technology companies on overlapping patents. (Compl. ¶¶ 57-60).
VII. Analyst’s Conclusion: Key Questions for the Case
- A core issue will be one of technical operation: does the Azure CDN’s method for handling a cache miss at an edge server—which the complaint’s evidence suggests is a server-side fetch from a parent tier—meet the '466 Patent’s claim requirement of "redirecting the client," or is there a fundamental mismatch between a reverse-proxy mechanism and the claimed client-side action?
- A second central question will be one of definitional scope: can the term "shared repeater server," rooted in the architecture of the 1990s internet, be construed to cover modern, multi-tenant cloud infrastructure where customer resources may be logically or virtually segregated on shared hardware?
- A key evidentiary question will focus on willfulness: does Plaintiff's narrative—alleging Microsoft's knowledge of the patents from a failed 2006 acquisition bid, its history as a customer, and its awareness of other litigation—provide sufficient evidence to establish the subjective intent required for a finding of willful infringement and potential enhanced damages?