Cypresswood Solutions LLC v. Google LLC

I. Executive Summary and Procedural Information

• Parties & Counsel:
  • Plaintiff: Cypresswood Solutions LLC (Texas)
  • Defendant: Google LLC (Delaware)
  • Plaintiff's Counsel: Russ August & Kabat; Miller Fair Henry, PLLC
• Case Identification: Cypresswood Solutions LLC v. Google LLC, 7:26-cv-00279, W.D. Tex., 07/24/2026
• Venue Allegations: Venue is based on Defendant's alleged residence, acts of infringement, and regular and established place of business within the Western District of Texas, including the operation of Google Cloud servers and employment of relevant engineering teams at its Austin campus.
• Core Dispute: Plaintiff alleges that Defendant's Google Cloud infrastructure, including its Colossus and Hyperdisk storage systems, infringes four patents related to distributed data storage, RAID implementation, and cloud computing management.
• Technical Context: The technology concerns hyperconverged infrastructure and software-defined datacenters, specifically methods for distributing RAID (Redundant Array of Independent Disks) functionality across multiple networked storage nodes to improve performance and resilience.
• Key Procedural History: The complaint alleges that Defendant had pre-suit knowledge of the asserted patent families, citing to Google's own patents which reference patents developed by Plaintiff's predecessor-in-interest, Pivot3, Inc.

Case Timeline

Date	Event
2008-06-06	Priority Date for '909 and '624 Patents
2008-06-30	Priority Date for '750 and '821 Patents
2012-01-03	'909 Patent Issued
2012-07-10	'750 Patent Issued
2012-08-07	'624 Patent Issued
c. 2013-07	Alleged Knowledge Date for '909, '750, and '624 Patents
2015-07-21	'821 Patent Issued
c. 2015-07	Alleged Knowledge Date for '821 Patent
2026-07-24	Complaint Filing Date

II. Technology and Patent(s)-in-Suit Analysis

U.S. Patent No. 8,090,909 - Method and System for Distributed RAID Implementation

• Patent Identification: U.S. Patent No. 8,090,909, issued January 3, 2012.

The Invention Explained

• Problem Addressed: Traditional RAID systems often rely on a single, centralized server to control and manage the disk array, creating a performance bottleneck and a single point of failure. Furthermore, the initial RAID level configuration is often inflexible. '909 Patent, col. 2:5-23
• The Patented Solution: The invention proposes a "distributed RAID system" composed of multiple "data banks" (networked computers with their own storage). Each data bank runs a similar RAID application, and they coordinate to distribute both data and RAID-related calculations (e.g., parity data) across the entire system. This avoids a central controller by distributing control functions, allowing for the creation of virtual storage "volumes" with specific RAID levels that span multiple data banks. '909 Patent, abstract '909 Patent, col. 2:26-54
• Technical Importance: This approach provided a architectural foundation for scalable, resilient, and high-performance storage systems built from commodity hardware, a concept central to modern cloud and hyperconverged infrastructure. '909 Patent, col. 3:1-5

Key Claims at a Glance

• The complaint asserts method claims, including dependent claim 4 Compl. ¶25, which relies on independent system claim 1.
• The essential elements of independent claim 1 include:
  • A system with one or more hosts and a plurality of "data bank computers."
  • The segments of a storage "volume" are stored across the data banks according to a "data bank order," which is a "permutation" of the plurality of data banks.
  • Redundancy data (e.g., for a RAID level) is stored in "redundancy segments" that do not reside on the same data bank as the corresponding data segment.
  • A data bank computer receives a command, determines which data bank stores the target segment based on the data bank order, and either executes the command locally or forwards it to the correct remote data bank.
• The complaint reserves the right to assert other claims.

U.S. Patent No. 8,219,750 - Method and System for Execution of Applications in Conjunction with Distributed RAID

• Patent Identification: U.S. Patent No. 8,219,750, issued July 10, 2012.

The Invention Explained

• Problem Addressed: In traditional datacenters, application servers (hosts) and storage systems (RAID arrays) are separate physical devices, leading to increased hardware costs, power consumption, and network complexity. Consolidating applications onto the same hardware that runs the storage system presents challenges related to resource management and operating system compatibility. '750 Patent, col. 2:1-12
• The Patented Solution: The invention builds on the distributed RAID system by adding a "virtualization layer" on the data bank computers. This allows applications to run inside "virtual machines" directly on the data banks. The state and context of each virtual machine are themselves stored as a volume within the distributed RAID system, enabling fault tolerance; if one data bank fails, the virtual machine can be recovered and executed on another data bank in the system. '750 Patent, abstract '750 Patent, col. 2:15-36
• Technical Importance: This patent describes a core principle of hyperconverged infrastructure: the unification of compute and storage resources on a single, scalable platform managed by software. '750 Patent, col. 2:37-43

Key Claims at a Glance

• The complaint asserts method claims, including dependent claim 5 Compl. ¶36, which relies on independent system claim 1.
• The essential elements of independent claim 1 include:
  • A system with a plurality of "data bank computers."
  • Each data bank can execute a "first application program" which is a "first virtual machine" using a "virtualization layer."
  • A distributed RAID application is implemented to manage a "first volume," where commands for this volume are issued by the application program.
  • The "first virtual machine" itself corresponds to a "second volume" also stored across the plurality of data banks.
  • The system determines the correct data bank for a command related to the first volume and processes it, including handling write commands and associated redundancy data.
• The complaint reserves the right to assert other claims.

U.S. Patent No. 8,239,624 - Method and System for Data Migration in a Distributed RAID Implementation

• Patent Identification: U.S. Patent No. 8,239,624, issued August 7, 2012.
• Technology Synopsis: This patent addresses the problem of migrating data within a distributed RAID system when its configuration changes, such as when new storage is added or a RAID level is altered. It discloses a method for seamlessly migrating portions of a data volume from a source configuration to a new destination configuration, using temporary "source" and "destination" range entries to manage access during the transition and minimize downtime. '624 Patent, abstract '624 Patent, col. 2:25-38
• Asserted Claims: The complaint asserts method claims, including at least Claim 8. Compl. ¶47
• Accused Features: The accused features are the infrastructure and services of Google Cloud, including Colossus and Hyperdisk. Compl. ¶42

U.S. Patent No. 9,086,821 - Method and System for Execution of Applications in Conjunction With RAID

• Patent Identification: U.S. Patent No. 9,086,821, issued July 21, 2015.
• Technology Synopsis: This patent appears to be a continuation of the technology described in the '750 Patent, focusing on the execution of applications within virtual machines on a distributed RAID system. It further details the system and method for running an application program within a first virtual machine, which itself is stored as a second volume on the distributed RAID system, and processing commands issued by that application. '821 Patent, abstract '821 Patent, col. 2:25-45
• Asserted Claims: The complaint asserts method claims, including at least Claim 13. Compl. ¶58
• Accused Features: The accused features are the infrastructure and services of Google Cloud, including Colossus and Hyperdisk. Compl. ¶53

III. The Accused Instrumentality

Product Identification

The accused instrumentalities are Google Cloud services, specifically including the "Colossus" and "Hyperdisk" storage systems and the servers, relays, and other infrastructure used to operate them. Compl. ¶20 Compl. ¶31 Compl. ¶42 Compl. ¶53

Functionality and Market Context

The complaint alleges that Colossus and Hyperdisk are functionalities of Google Cloud services offered from numerous global locations, including network edge locations within the Western District of Texas. Compl. ¶17 The complaint identifies these as fundamental components of Google's cloud computing platform. Compl. ¶6 Compl. ¶17 The complaint does not provide specific technical details on the operation of Colossus or Hyperdisk beyond identifying them as the infringing systems.

IV. Analysis of Infringement Allegations

The complaint alleges that Defendant's Google Cloud, Colossus, and Hyperdisk systems practice the methods claimed in the Asserted Patents Compl. ¶24 Compl. ¶35 Compl. ¶46 Compl. ¶57 For each patent, the complaint refers to an external exhibit containing a descriptive chart purporting to show how the accused systems infringe a specific claim Compl. ¶25 Compl. ¶36 Compl. ¶47 Compl. ¶58 As these exhibits were not filed with the complaint, a detailed claim chart analysis is not possible. The narrative infringement theory presented in the complaint is conclusory, stating that Defendant's making, using, and selling of the Accused Instrumentalities infringes the asserted claims Compl. ¶24 Compl. ¶35

No probative visual evidence provided in complaint.

• Identified Points of Contention:
  • Scope Questions: A central issue will be whether the architecture of Google's Colossus and Hyperdisk systems maps onto the specific "data bank computer" architecture required by the claims. The court may need to determine if Google's proprietary and likely complex internal systems, which are not described in the complaint, meet the claim limitation of discrete "data bank computers" that coordinate according to a "permutation" to store data and redundancy segments.
  • Technical Questions: For the '750 and '821 Patents, a key technical question will be whether Google's cloud platform practices the claimed method of running applications in a "virtual machine" where the virtual machine itself is stored as a distinct "volume" within the same distributed storage system that it utilizes. The evidence required to prove this specific two-level storage relationship (application-uses-storage-system, storage-system-stores-application) may be a significant point of dispute.

V. Key Claim Terms for Construction

For the '909 Patent

• The Term: "data bank computer"
• Context and Importance: This term is the fundamental building block of the claimed distributed system. Its definition will be critical to determining if Google's cloud infrastructure, which may be architected in numerous ways, falls within the scope of the claims. Practitioners may focus on this term because the infringement analysis hinges on whether Google's infrastructure is a collection of these specific units or something architecturally different.
• Intrinsic Evidence for Interpretation:
  • Evidence for a Broader Interpretation: The specification describes a data bank as a "computer operable to implement distributed RAID" and notes that "standard x86 based servers" may be used, suggesting the term is not limited to custom or highly specialized hardware. '909 Patent, col. 6:50-54 '909 Patent, col. 6:65-67
  • Evidence for a Narrower Interpretation: The figures and description consistently show a data bank as a discrete unit comprising a processor, data store, and a distributed RAID application, which coordinates with other data banks. '909 Patent, Fig. 2A '909 Patent, col. 6:49-55 This could support an interpretation that requires a collection of distinct, peer-like server nodes, which may or may not map to Google's architecture.

For the '750 Patent

• The Term: "virtual machine"
• Context and Importance: The invention's core is executing applications in "virtual machines" on the storage nodes themselves. The definition of "virtual machine" is therefore central. Google uses various containerization and virtualization technologies; whether these meet the claim definition will be a key issue.
• Intrinsic Evidence for Interpretation:
  • Evidence for a Broader Interpretation: The patent lists a wide variety of then-current virtualization technologies, including "Containers by Sun Microsystems, Linux-KVM, Linux VServer, Oracle VM, Virtual PC, Virtual Server by Microsoft, PowerVM by IBM Logical Domains by Sun Microsystems, VMware Server, etc. or any other type of virtualization or emulation machine application," suggesting the term was intended to be encompassing. '750 Patent, col. 7:45-53
  • Evidence for a Narrower Interpretation: The claims require that this "first virtual machine" also "corresponds to a second volume" stored on the data banks. '750 Patent, claim 1 This linkage might be used to argue for a narrower definition, requiring not just any virtualization, but one where the application's entire context is encapsulated and stored as a volume in the underlying distributed file system, a specific technical implementation.

VI. Other Allegations

• Indirect Infringement: The complaint alleges induced infringement, stating that Defendant encourages and instructs customers and vendors to use Google Cloud services in ways that directly infringe the Asserted Patents through "normal and customary use," documentation, and education. Compl. ¶27 Compl. ¶38 Compl. ¶49 Compl. ¶60
• Willful Infringement: Willfulness is alleged for all four patents. The basis for this allegation is Defendant's alleged pre-suit knowledge of the patents, purportedly demonstrated by citations to the patent family in Google's own patent filings, and alleged knowledge dating back to at least July 2013 for the '909, '750, and '624 Patents and July 2015 for the '821 Patent. Compl. ¶12 Compl. ¶26 Compl. ¶37 Compl. ¶48 Compl. ¶59

VII. Analyst's Conclusion: Key Questions for the Case

1. A core issue will be one of architectural mapping: can Plaintiff produce evidence, likely subject to significant discovery disputes, to show that Google's opaque and proprietary Colossus and Hyperdisk systems are organized as a "plurality of data bank computers" that manage data and redundancy according to a "permutation" as required by the claims, or does Google's architecture differ in a material way?
2. A second central question is one of definitional scope, particularly for the '750 and '821 patents: does Google's use of virtualization and container technologies for cloud applications meet the specific claim requirement of running a "virtual machine" that is itself stored as a "volume" in the underlying distributed storage system, or is there a technical distinction in how Google implements its compute and storage services?
3. An evidentiary question will be the strength of the willfulness allegation: does the citation to Plaintiff's patent family in Google's own patents establish the "egregious" level of conduct necessary for a finding of willfulness, or will it be viewed as routine patent prosecution practice insufficient to prove culpable intent?