1:26-cv-00799
DB USA Core Corp v. Intellectual Ventures I LLC
I. Executive Summary and Procedural Information
- Parties & Counsel:
- Plaintiff: DB USA Core Corporation (New Jersey)
- Defendant: Intellectual Ventures I LLC and Intellectual Ventures II LLC (Delaware)
- Plaintiff's Counsel: Skadden, Arps, Slate, Meagher & Flom LLP
- Case Identification: 1:26-cv-00799, D. Del., 07/02/2026
- Venue Allegations: Venue is alleged to be proper in the District of Delaware because both Defendants are limited liability companies organized under the laws of Delaware and therefore reside in the district for venue purposes.
- Jurisdiction: The complaint pleads a declaratory-judgment patent controversy under federal patent jurisdiction and alleges personal jurisdiction over both Intellectual Ventures I LLC and Intellectual Ventures II LLC in Delaware.
- Core Dispute: Plaintiff seeks a declaratory judgment of non-infringement against Defendant, arguing that its use of five common open-source software platforms does not infringe five patents that Defendant has asserted against it.
- Technical Context: The technologies involve foundational aspects of large-scale, distributed enterprise computing, including load balancing, virtual networking, data caching, asynchronous messaging, and failure recovery.
- Key Procedural History: Defendant sent Plaintiff a formal notice letter on May 13, 2024, accusing it of infringement. The complaint notes that Defendant has engaged in a "sustained campaign of litigation" against other financial services companies based on the same open-source platforms and patents. Two of the asserted patents, U.S. Patent Nos. 7,949,785 and 8,407,722, are noted as having expired prior to the filing of the complaint. In parallel, the source record reflects inter partes review proceedings filed at the PTAB against asserted patents, including IPR2025-00987 challenging the '722 Patent (U.S. Patent No. 8,407,722).
Case Timeline
| Date | Event |
|---|---|
| 2001-12-14 | '722 Patent Priority Date |
| 2002-03-13 | '582 Patent Priority Date |
| 2003-03-31 | '785 Patent Priority Date |
| 2004-08-17 | '635 Patent Priority Date |
| 2004-12-30 | '844 Patent Priority Date |
| 2007-08-14 | '582 Patent Issue Date |
| 2011-05-24 | '785 Patent Issue Date |
| 2012-12-11 | '844 Patent Issue Date |
| 2013-03-26 | '722 Patent Issue Date |
| 2015-05-26 | '635 Patent Issue Date |
| 2024-05-13 | Defendant sent formal notice letter to Plaintiff |
| 2025-01-14 | '722 Patent Expiration Date |
| 2025-02-13 | '785 Patent Expiration Date |
| 2026-07-02 | Complaint Filing Date |
II. Technology and Patent(s)-in-Suit Analysis
U.S. Patent No. 7,257,582 - "Load balancing with shared data," issued August 14, 2007
The Invention Explained
- Problem Addressed: The patent addresses the problem of inefficient workload distribution in environments with multiple, potentially heterogeneous computer systems, where one computer might be heavily loaded while others are idle ʼ582 Patent, col. 1:28-39
- The Patented Solution: The invention proposes a method to parallelize a computing process, such as a data sort, across multiple processors. It works by "logically subdividing" a pre-existing input file into partitions without actually reading the file's contents first ʼ582 Patent, abstract Descriptions of these logical partitions are made available via a central "control file," and a plurality of "subtask processors" then claim and process these partitions on a "first come first serve basis" ʼ582 Patent, col. 2:1-2 '582 Patent, Fig. 4
- Technical Importance: The technology provided a method for parallelizing large batch-processing tasks in multi-computer environments with shared storage, aiming to improve efficiency and processor utilization.
Key Claims at a Glance
- The complaint appears to focus on limitations found in independent method claim 1 Compl. ¶38
- Key elements of independent claim 1 include:
- (a) automatically determining file allocation and logically subdividing records of said input file into a plurality of partitions;
- (b) distributing descriptions of all of said partitions to each of a plurality of subtask processors;
- (c) simultaneously executing at least a respective one of the subtasks... in each of at least some of said processors on a respective one of the partitions;
- (d) thereafter repeating step (c) in at least some of the subtask processors each with another unprocessed partition on a first-come/first-served basis; and
- (e) generating at least one output combining all of the subtask outputs.
- The complaint seeks a declaration of non-infringement of "any valid claim" Compl. ¶37
U.S. Patent No. 7,949,785 - "Secure virtual community network system," issued May 24, 2011
The Invention Explained
- Problem Addressed: The patent addresses the challenge of enabling secure communication between devices across the internet, particularly when devices are located behind firewalls or Network Address Translation (NAT) devices and thus have private, non-routable IP addresses ʼ785 Patent, col. 2:20-29. It also notes the problem of mobile devices moving between networks ʼ785 Patent, col. 2:31-44.
- The Patented Solution: The invention describes a "virtual community network" (VCN) where member devices can communicate securely as if on a private LAN '785 Patent, abstract. A central "VCN Manager" registers devices and assigns each a unique "virtual network address" that is not directly routable on the public internet. A "Route Director" entity helps route encapsulated packets between members. The VCN Manager also functions as a specialized DNS server that, in response to a query, returns the public IP address of the appropriate Route Director along with the private and virtual IP addresses of the destination device '785 Patent, col. 13:35-54.
- Technical Importance: The technology provided a centralized architecture for creating a Virtual Private Network (VPN) capable of connecting endpoints with dynamic or private IP addresses, a significant challenge for peer-to-peer communications at the time.
Key Claims at a Glance
- The complaint appears to contest limitations found in independent system claim 1 Compl. ¶43
- Key elements of independent claim 1 include:
- a virtual network manager... configured to register devices in a virtual network that is defined by a domain name;
- each device... being identified to the other devices by a virtual network address that is unique for each device and not directly routable via a public network;
- a route director... configured to communicate data between the devices... as encapsulated packets; and
- the virtual network manager further configured to receive a DNS request... and return a public network address of the route director, a private network address for the destination device, and the second virtual network address.
- The complaint seeks a declaration of non-infringement of "any valid claim" Compl. ¶42
U.S. Patent No. 8,332,844 - "Root image caching and indexing for block-level distributed application management," issued December 11, 2012
- Technology Synopsis: The patent describes a system for managing application environments in a distributed system. It aims to solve inefficiencies in creating and updating boot images for many compute nodes by using a shared, read-only "root image" and individual, writable "leaf images" that store only the changes or additions made by each node, operating at the data block level rather than the file level ’844 Patent, abstract ’844 Patent, col. 1:15-29
- Asserted Claims: The complaint does not specify claims but contests core architectural requirements Compl. ¶48
- Accused Features: The Docker platform is accused of infringement, with the complaint arguing that Docker's layered file-system, container architecture, and copy-on-write behavior do not satisfy the patent's block-level "root image" and "leaf image" limitations Compl. ¶¶46-48
U.S. Patent No. 8,407,722 - "Asynchronous messaging using a node specialization architecture in the dynamic routing network," issued March 26, 2013
- Technology Synopsis: The patent discloses a network for routing "update messages" that contain updates for properties of "live objects" at client devices. The system identifies a "category" for an update message and routes it to a "node" having a specialized "node type" based on a pre-defined mapping. The specialized node then forwards the update to the client that has registered for it ’722 Patent, abstract
- Asserted Claims: The complaint does not specify claims but contests key functional steps Compl. ¶53
- Accused Features: The Apache Kafka platform is accused of infringement. The complaint contends that Kafka, a distributed publish-subscribe platform with immutable records, does not route "update messages" to perform operations on a receiving device or employ the claimed "node specialization architecture" Compl. ¶¶51-54
U.S. Patent No. 9,043,635 - "Techniques for upstream failure detection and failure recovery," issued May 26, 2015
- Technology Synopsis: The patent describes a method for a client to detect an "upstream failure" between its content provider and its primary data center. The client identifies a "first data center that is closest," receives an application, and then determines a failure has occurred based on the "non-receipt of a recurring 'status message'." Upon failure detection, the application automatically connects to a second data center to receive content ’635 Patent, abstract
- Asserted Claims: The complaint does not specify claims but contests the core architectural and functional requirements Compl. ¶58
- Accused Features: The Apache Cassandra platform is accused of infringement. The complaint argues that Cassandra's masterless, peer-to-peer architecture does not have the claimed "client"/"data center" structure, does not identify a "closest" data center, and does not perform client-side failure detection based on a missed status message Compl. ¶¶56-58
III. The Accused Instrumentality
Product Identification
- The accused instrumentalities are five widely used open-source software platforms: Apache Spark, Kubernetes, Docker, Apache Kafka, and Apache Cassandra Compl. ¶3 The complaint alleges infringement is based on Plaintiff's "ordinary use" of these platforms Compl. ¶3
Functionality and Market Context
- The complaint describes the accused instrumentalities as "widely adopted, third-party open-source software" created and maintained by independent communities and used by thousands of enterprises Compl. ¶3
- For the '582 Patent, the accused Apache Spark platform is described as a system that "distributes data and computational tasks to worker nodes" and "assigns work to executors through a centralized scheduler" Compl. ¶38
- For the '785 Patent, the accused Kubernetes platform is described as a "container-orchestration platform" that does not implement the patent's specific "virtual-network-manager and route-director architecture" Compl. ¶43
- No probative visual evidence provided in complaint.
IV. Analysis of Infringement Allegations
The complaint is for declaratory judgment of non-infringement and does not contain traditional infringement claim charts. The following tables summarize the infringement theories attributed to Defendant, as reconstructed from Plaintiff's non-infringement arguments.
'582 Patent Infringement Allegations
| Claim Element (from Independent Claim 1) | Alleged Infringing Functionality | Complaint Citation | Patent Citation |
|---|---|---|---|
| logically subdividing records of said input file into a plurality of partitions | Apache Spark's process of dividing a data processing job into smaller tasks. | ¶38 | col. 3:21-39 |
| distributing descriptions of all of said partitions to each of a plurality of subtask processors | Apache Spark's centralized scheduler assigning computational tasks to worker nodes. | ¶38 | col. 2:8-12 |
| ...repeating step (c) in at least some of the subtask processors each with another unprocessed partition on a first-come/first-served basis | Apache Spark's scheduler assigning work to executors. | ¶38 | col. 2:1-2 |
- Identified Points of Contention:
- Technical Question: A central question is whether Apache Spark's method of distributing computational tasks via a centralized scheduler is functionally equivalent to the patent's method of "distributing descriptions" of pre-defined data partitions for processors to claim on a "first-come/first-served basis" Compl. ¶38 The complaint suggests a fundamental architectural mismatch between Spark's dynamic task assignment and the patent's more static, pre-partitioned model.
'785 Patent Infringement Allegations
| Claim Element (from Independent Claim 1) | Alleged Infringing Functionality | Complaint Citation | Patent Citation |
|---|---|---|---|
| a virtual network manager implemented with a first device... configured to register devices in a virtual network | Kubernetes' control plane components (e.g., API server, etcd) managing the state of the cluster. | ¶43 | col. 11:41-47 |
| each device... being identified to the other devices by a virtual network address that is unique for each device and not directly routable via a public network | Kubernetes assigning internal, non-publicly routable IP addresses to pods within the cluster. | ¶43 | сol. 11:51-56 |
| a route director... configured to communicate data between the devices... as encapsulated packets | Kubernetes' networking layer (e.g., CNI plugins) managing traffic between pods, often using encapsulation. | ¶43 | col. 12:20-24 |
| the virtual network manager further configured to receive a DNS request... and return a public network address of the route director, a private network address... and the second virtual network address | Kubernetes' internal DNS service (e.g., CoreDNS) resolving service names to internal cluster IPs. | ¶43 | col. 13:35-54 |
- Identified Points of Contention:
- Scope Question: The dispute may turn on whether the distributed, consensus-based control plane of Kubernetes can be considered a "virtual network manager" and "route director" as those terms are used in the patent. The complaint alleges that Kubernetes "does not implement the claimed virtual-network-manager and route-director architecture" Compl. ¶43
- Technical Question: An issue for the court could be whether Kubernetes' DNS-based service discovery performs the specific function claimed: returning a combination of a public route director address, a private destination address, and a virtual destination address in response to a single DNS request Compl. ¶43
V. Key Claim Terms for Construction
For the '582 Patent:
- The Term: "distributing descriptions of all of said partitions"
- Context and Importance: This term is critical because the complaint's non-infringement argument hinges on the assertion that Apache Spark distributes computational tasks, not "descriptions of partitions" of a "preexisting input file" Compl. ¶38 The construction of this term will determine whether Spark's architecture falls within the claim's scope.
- Intrinsic Evidence for Interpretation:
- Evidence for a Broader Interpretation: The patent describes the use of a "control file" that contains the subdivision of the input, which "Each Sub Task finds" ʼ582 Patent, col. 4:1-2 Parties arguing for infringement may suggest that making this control file available constitutes "distributing descriptions."
- Evidence for a Narrower Interpretation: Claim 1 requires "distributing descriptions... to each of a plurality of subtask processors," which may suggest an active push or broadcast of the partition map to all processors, rather than processors pulling tasks from a central scheduler as alleged for Spark ʼ582 Patent, col. 6:51-54 Compl. ¶38
For the '785 Patent:
- The Term: "virtual network manager"
- Context and Importance: Plaintiff argues that Kubernetes, a "container-orchestration platform," does not implement the claimed architecture, including the "virtual network manager" Compl. ¶43 Practitioners may focus on whether the distributed components of the Kubernetes control plane can be mapped to this singular, logically centralized entity.
- Intrinsic Evidence for Interpretation:
- Evidence for a Broader Interpretation: The specification notes the VCN Manager can be a "central server or server cluster" '785 Patent, col. 9:66-10:1, which may support an argument that a distributed set of components acting in concert can constitute the claimed manager.
- Evidence for a Narrower Interpretation: The patent's figures and process flows consistently depict the VCN Manager (e.g., item 510 in Fig. 5B) as a distinct logical entity with which member agents explicitly communicate for registration and DNS services '785 Patent, col. 17:10-24; ’785 Patent, Fig. 12A. This could support a narrower construction requiring a more formally centralized role than what is present in Kubernetes.
VI. Other Allegations
- Indirect Infringement: The complaint seeks a declaratory judgment that Plaintiff "does not infringe, and has not infringed, directly or indirectly" any valid claim of the asserted patents Compl. ¶37 Compl. ¶42 Compl. ¶47 Compl. ¶52 Compl. ¶57 The complaint does not, however, provide specific factual arguments related to theories of indirect infringement, as its primary focus is on demonstrating a lack of direct infringement based on technical mismatches between the accused open-source platforms and the claim limitations.
- Willful Infringement: The complaint makes no allegations regarding willfulness.
VII. Analyst's Conclusion: Key Questions for the Case
This declaratory judgment action presents several core questions for the court, revolving around the application of patent claims from the early-to-mid 2000s to modern, widely adopted open-source technologies.
A central issue will be one of definitional scope: can claim terms rooted in the context of their time, such as "logically subdividing records of [a] preexisting input file" ('582 patent) or a centralized "virtual network manager" ('785 patent), be construed to read on the dynamic, on-the-fly data handling of Apache Spark and the distributed control plane architecture of Kubernetes?
A second key issue will be one of functional mismatch: does the technical operation of the accused platforms perform the specific steps required by the claims? For example, does Apache Kafka's publish-subscribe model for immutable logs perform the function of routing an "update message" to "update the property of the live object" as claimed by the '722 patent, or is there a fundamental difference in their technical purpose and operation?
Finally, the case raises an overarching question of architectural equivalence: the complaint consistently alleges that the accused platforms (Spark, Kubernetes, Docker, Kafka, Cassandra) are based on "fundamentally different architectures" than what is described and claimed in the patents-in-suit Compl. ¶5 A key question for the court will be to determine whether the high-level functions performed by the accused platforms are achieved in a manner that is structurally and operationally equivalent to the specific architectures mandated by the asserted claims.