Datasphere LLC v. Equinix Inc

I. Executive Summary and Procedural Information

• Parties & Counsel:
  • Plaintiff: Datasphere, LLC (Georgia)
  • Defendant: Equinix, Inc. (Delaware)
  • Plaintiff's Counsel: Folio Law Group PLLC
• Case Identification: 2:26-cv-00075, E.D. Tex., 05/04/2026
• Venue Allegations: Venue is based on Defendant Equinix allegedly having a regular and established place of business in the district, specifically its DA7 Data Center in Plano, Texas, where acts of infringement are alleged to occur.
• Core Dispute: Plaintiff alleges that Defendant's data center infrastructure and interconnection services infringe four U.S. patents related to data storage management, optical transport flow control, and packet-switched network routing.
• Technical Context: The technologies at issue concern foundational data center operations, including managing I/O between servers and storage, extending high-speed networks over long distances, and routing traffic efficiently, which are all critical to modern cloud computing and internet services.
• Key Procedural History: The complaint is an Amended Complaint for Patent Infringement. The complaint does not mention any prior litigation, inter partes review proceedings, or licensing history related to the Asserted Patents.

Case Timeline

Date	Event
2002-09-27	'172 Patent Priority Date
2004-02-24	'239 Patent Priority Date
2004-09-08	'778 Patent Priority Date
2005-10-07	'381 Patent Priority Date
2007-04-10	'172 Patent Issued
2009-05-05	'239 Patent Issued
2009-12-29	'381 Patent Issued
2010-08-24	'778 Patent Issued
2026-05-04	Amended Complaint Filed

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

U.S. Patent No. 7,640,381 - Input/Output Decoupling System Method Having a Cache for Exchanging Data Between Non-Volatile Storage and Plurality of Clients Having Asynchronous Transfers

• Patent Identification: U.S. Patent No. 7,640,381, "Input/Output Decoupling System Method Having a Cache for Exchanging Data Between Non-Volatile Storage and Plurality of Clients Having Asynchronous Transfers," issued December 29, 2009 (the "'381 Patent"). Compl. ¶16

The Invention Explained

• Problem Addressed: The patent addresses the performance bottleneck created by slow secondary storage devices (like disk drives) compared to fast host processors. While caching helps, systems with multiple concurrent clients can lead to cache "thrashing," where data is prematurely flushed and re-loaded, degrading performance. '381 Patent, col. 1:25-2:31
• The Patented Solution: The invention proposes an "I/O accelerator" that sits between multiple host computers (clients) and the storage system. This accelerator uses a large cache to decouple the fast data transfers to the clients from the slower, asynchronous data transfers from storage. The system manages the data rate to each client based on the amount of data available in that client's dedicated buffer space within the cache, allowing for tailored performance for each client independent of the underlying storage access. '381 Patent, abstract '381 Patent, col. 3:5-3:20 '381 Patent, FIG. 2
• Technical Importance: This architecture is designed to smooth I/O performance and prevent bottlenecks in multi-client, high-demand environments typical of enterprise storage systems and data centers. '381 Patent, col. 2:60-3:4

Key Claims at a Glance

• The complaint asserts at least independent claim 25. Compl. ¶45
• The essential elements of independent claim 25 are:
  • An apparatus, comprising:
  • a storage interface configured to communicate with a storage system having a non-volatile storage component;
  • a host interface configured to communicate with a plurality of clients;
  • a cache configured to cache communications exchanged between the non-volatile storage component and the plurality of clients, the cache configured to associate a buffer with each client; and
  • circuitry configured to: identify a data rate for transferring data from the apparatus, over the host interface, to one of the plurality of clients, wherein the identified data rate is based on an amount of data existing in the buffer for the receiving client and decoupled from asynchronous transfers over the storage interface. Compl. ¶46
• The complaint reserves the right to assert other claims of the '381 Patent. Compl. ¶45

U.S. Patent No. 7,782,778 - Apparatus and Method for Fibre Channel Distance Extension Embedded Within an Optical Transport System

• Patent Identification: U.S. Patent No. 7,782,778, "Apparatus and Method for Fibre Channel Distance Extension Embedded Within an Optical Transport System," issued August 24, 2010 (the "'778 Patent"). Compl. ¶20

The Invention Explained

• Problem Addressed: Extending Fibre Channel (FC) based Storage Area Networks (SANs) over wide geographical distances is challenging. Standard FC protocols have inherent distance limitations due to their flow-control mechanisms ("buffer credits"). Using public IP or SONET networks to bridge this distance introduces performance degradation, security vulnerabilities, and increased costs. '778 Patent, col. 1:30-2:14
• The Patented Solution: The invention embeds the distance-extension functionality directly into an optical transport system. It aggregates multiple native FC data streams at an "ingress terminal," transports them over a long-distance optical link, and de-aggregates them at an "egress terminal." The system uses large memory buffers and a proprietary flow-control mechanism between the ingress and egress terminals to manage congestion and prevent data loss over the long-haul link, thereby extending the SAN's reach transparently. '778 Patent, abstract '778 Patent, col. 3:15-4:2 '778 Patent, FIG. 10
• Technical Importance: This technology enables reliable, high-performance, and secure long-distance SAN connectivity, which is critical for enterprise applications like disaster recovery and business continuity between geographically separate data centers. '778 Patent, col. 1:62-66

Key Claims at a Glance

• The complaint asserts at least independent claim 25. Compl. ¶72
• The essential elements of independent claim 25 are:
  • A method for controlling a flow of data over an optical transport system, the method comprising:
  • determining whether a memory in an egress terminal is available to store the data received by the egress terminal based at least in part on a data amount transmitted to the egress terminal;
  • allowing transmission of the data from a first network to an ingress terminal only when the memory is available to store the data;
  • receiving the data from the first network; and
  • transmitting the data over the optical transport system from the ingress terminal to the egress terminal. Compl. ¶73
• The complaint reserves the right to assert other claims of the '778 Patent. Compl. ¶72

U.S. Patent No. 7,529,239 - Distributed Dynamic Routing

• Patent Identification: U.S. Patent No. 7,529,239, "Distributed Dynamic Routing," issued May 5, 2009 (the "'239 Patent"). Compl. ¶24
• Technology Synopsis: The patent describes a method for routing packets through an internal network connecting multiple external networks. To solve the problem of inefficient, multi-hop routing within the internal network '239 Patent, col. 2:1-29, the invention proposes a system where a source "edge device" determines the final "target edge device" for an incoming packet and generates a transmission packet for direct, Layer 2 delivery to that target, minimizing internal routing decisions '239 Patent, abstract
• Asserted Claims: The complaint asserts at least independent claim 1. Compl. ¶62
• Accused Features: The accused features are Equinix's Fabric Cloud Router, IP-WAN, and Equinix Fabric services, which are alleged to comprise an internal packet-switched network with edge devices that perform distributed routing to forward packets between external networks. (Compl. ¶¶41; Compl. ¶62).

U.S. Patent No. 7,203,172 - Internodal Routing Engine For A Next Generation Network

• Patent Identification: U.S. Patent No. 7,203,172, "Internodal Routing Engine For A Next Generation Network," issued April 10, 2007 (the "'172 Patent"). Compl. ¶28
• Technology Synopsis: The patent addresses cost-effective call routing in hybrid networks that use both packet-switched (e.g., VoIP) and traditional circuit-switched (PSTN) infrastructure. The described problem is that inefficient routing can lead to excessive use of the more expensive PSTN. '172 Patent, col. 2:15-28 The patented solution is a routing engine within a softswitch that uses geographic location information to identify the softswitch in the provider's network that is physically nearest to the call's destination, and then routes the call to that nearest softswitch over the packet network to minimize PSTN usage. '172 Patent, abstract
• Asserted Claims: The complaint asserts at least independent claim 1. Compl. ¶81 Compl. ¶85
• Accused Features: The complaint alleges inducement of infringement, where Equinix's customers (VoIP Providers) use Equinix's data center and interconnection platforms (including Equinix Fabric) to deploy their own softswitches and call-routing infrastructure, which allegedly perform the patented method. Compl. ¶¶82-84 Compl. ¶97

III. The Accused Instrumentality

• Product Identification: The accused instrumentalities are a suite of data center infrastructure and interconnection services provided by Equinix, including Equinix Network Edge, Equinix Managed Private Cloud, Equinix Metal, Equinix Fabric (including E-Line and Data Center Interconnect), Equinix Metro Connect, Equinix Fabric Cloud Router, and IP-WAN. Compl. ¶¶39-42
• Functionality and Market Context: The complaint alleges these services collectively provide the foundational compute, storage, virtualization, and networking infrastructure used by enterprises and cloud providers. Compl. ¶37 For example, Equinix Network Edge provides virtualized network functions on demand Compl. ¶48, Equinix Metal offers bare-metal server infrastructure Compl. ¶50, and Equinix Fabric provides a global, software-defined network to interconnect digital infrastructure Compl. ¶75 These services are positioned as core components for building and scaling modern hybrid and multi-cloud applications. Compl. ¶57

IV. Analysis of Infringement Allegations

'381 Patent Infringement Allegations

Claim Element (from Independent Claim 25)	Alleged Infringing Functionality	Complaint Citation	Patent Citation
[A] a storage interface configured to communicate with a storage system having a non-volatile storage component;	The accused services utilize servers with storage interfaces (e.g., PCIe, NVMe, SATA, SAS) that communicate with non-volatile storage components like SSDs and NVMe flash storage.	¶¶52-54	col. 4:5-10
[B] a host interface configured to communicate with a plurality of clients;	The accused services utilize network interfaces (e.g., NICs) and virtual interfaces to communicate with multiple clients, including customer systems, cloud providers, and other network endpoints, via technologies like Equinix Fabric.	¶¶56-58	col. 4:17-20
[C] a cache configured to cache communications exchanged between the non-volatile storage component and the plurality of clients, the cache configured to associate a buffer with each client;	The servers, hypervisors, and operating systems within the accused services are alleged to use various caches (e.g., page caches, socket buffers, I/O buffers) that associate buffers with respective client connections.	¶¶60-62	col. 5:21-25
[D] circuitry configured to: identify a data rate for transferring data...wherein the identified data rate is based on an amount of data existing in the buffer for the receiving client and decoupled from asynchronous transfers over the storage interface.	Processors, controllers, and OS/hypervisor logic are alleged to identify and enforce data rates through bandwidth reservation and traffic policing. This is allegedly based on buffer data (e.g., TCP socket buffers, NIC queues) and decoupled from storage I/O, which proceeds separately via mechanisms like page cache writeback.	¶¶64-66	col. 6:49-55

• Identified Points of Contention:
  • Scope Questions: A potential issue is whether the collection of hardware (servers, NICs) and general-purpose software (hypervisors, operating systems) that comprise Equinix's services constitutes the singular, integrated "apparatus" described in the patent.
  • Technical Questions: A key question is whether standard operating system mechanisms for network traffic management (e.g., TCP flow control, traffic policing) perform the specific function of "identify[ing] a data rate ... based on an amount of data existing in the buffer for the receiving client" as required by the claim, or if the claim requires a more specialized, purpose-built function as described in the '381 patent's detailed embodiments.

'778 Patent Infringement Allegations

Claim Element (from Independent Claim 25)	Alleged Infringing Functionality	Complaint Citation	Patent Citation
[A] determining whether a memory in an egress terminal is available...based at least in part on a data amount transmitted to the egress terminal;	The accused Equinix Fabric and Metro Connect services are alleged to use Ethernet flow control, where an egress terminal with full receive buffers sends PAUSE frames to halt transmission. The determination of memory availability is thus based on buffer occupancy, which is a function of the amount of data transmitted to and received by the egress terminal.	¶¶78-79	col. 6:1-5
[B] allowing transmission of the data from a first network to an ingress terminal only when the memory is available to store the data;	The ingress terminal allegedly allows data to enter the transport path only when the downstream egress terminal has memory available, as signaled by the absence of PAUSE frames, consistent with standard Ethernet flow-control mechanisms.	¶¶81-82	col. 5:66-68
[C] receiving the data from the first network; and	The accused services receive customer data at an ingress terminal, such as an Equinix Fabric port or Metro Connect endpoint.	¶¶84-85	col. 5:46-49
[D] transmitting the data over the optical transport system from the ingress terminal to the egress terminal.	The accused services transmit customer data from the ingress terminal to the egress terminal over Equinix's underlying optical transport infrastructure, including optical links and switches.	¶¶87-88	col. 5:20-24

• Identified Points of Contention:
  • Scope Questions: The '778 Patent is titled "Apparatus and Method for Fibre Channel Distance Extension" and its specification is heavily focused on solving problems specific to the Fibre Channel (FC) protocol. The accused services are primarily Ethernet-based. Compl. ¶75 This raises the question of whether the method claims, though written in general terms, are implicitly limited by the specification to an FC environment.
  • Technical Questions: The patent describes an internal proprietary flow control algorithm. Compl. ¶72 The complaint alleges infringement via standard Ethernet flow control (e.g., IEEE 802.3x PAUSE frames). (Compl. ¶78; Compl. ¶79). The analysis may turn on whether the operation of standard Ethernet PAUSE frames is functionally equivalent to the claimed step of "determining whether a memory...is available...based at least in part on a data amount transmitted to the egress terminal."

No probative visual evidence provided in complaint.

V. Key Claim Terms for Construction

For the '381 Patent

• The Term: "circuitry configured to: identify a data rate ... based on an amount of data existing in the buffer for the receiving client" (from Claim 25)
• Context and Importance: This term is the core active function of the claimed invention. Its construction will determine whether the general-purpose traffic management features of a modern server operating system can be found to infringe, or if a more specialized, purpose-built mechanism is required. Practitioners may focus on this term because the plaintiff's infringement theory relies on mapping it to commodity server functions. Compl. ¶¶64-66
• Intrinsic Evidence for Interpretation:
  • Evidence for a Broader Interpretation: The patent describes the function of the "host manager" and "function manager" in high-level, functional terms, such as performing a "set-up operation" and "passing information." '381 Patent, col. 4:65-68 This language may support an interpretation that covers any combination of hardware and software components that achieve the stated outcome.
  • Evidence for a Narrower Interpretation: The specification consistently describes the invention as an "I/O accelerator" with distinct logical components like a "host manager," "function manager," and "buffer manager." '381 Patent, FIG. 2 '381 Patent, col. 4:60-68 This could support an argument that the term requires a more structured apparatus than a general-purpose server running a standard operating system.

For the '778 Patent

• The Term: "determining whether a memory in an egress terminal is available ... based at least in part on a data amount transmitted to the egress terminal" (from Claim 25)
• Context and Importance: The plaintiff alleges this step is met by standard Ethernet flow control. Compl. ¶78 The defendant may argue that this claim requires a more sophisticated accounting of data quantities than the simple "buffer full" signal provided by an Ethernet PAUSE frame.
• Intrinsic Evidence for Interpretation:
  • Evidence for a Broader Interpretation: The claim requires the determination to be "based at least in part on" the data amount. An argument could be made that a buffer becoming full is a direct consequence of the amount of data transmitted to it, thereby satisfying this broad language. The specification also states the ingress circuit "receives buffer memory status from the far-end," which could encompass a simple PAUSE signal. '778 Patent, col. 6:2-3
  • Evidence for a Narrower Interpretation: The specification describes a system that "ensures that the number of unacknowledged transmitted data frames is less than the buffer memory size." '778 Patent, col. 5:66-6:2 This language, along with the patent's deep focus on Fibre Channel's credit-based flow control, suggests a more active accounting mechanism than simply reacting to a binary "stop/go" signal from a downstream buffer.

VI. Other Allegations

• Indirect Infringement: The complaint alleges active inducement of infringement of the '172 Patent. The factual basis for this allegation is that Equinix provides its customers (VoIP Providers) with the necessary infrastructure (e.g., Equinix Fabric), services, and technical guidance (documentation, APIs) to build and operate distributed call-routing systems. Compl. ¶¶96-98 The complaint further alleges that Equinix promotes use cases like VoIP interconnection that "necessarily involve" performing the patented method. Compl. ¶99
• Willful Infringement: The complaint does not contain a standalone count for willful infringement. However, in its inducement allegations for the '172 Patent, the complaint alleges that Equinix had knowledge of the patent and of the infringing nature of the induced acts "at least as of the filing of the original complaint, and earlier," and continued to induce infringement despite this knowledge. Compl. ¶¶100-101 These allegations could form the basis for a later claim of willfulness.

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

1. Definitional Scope: A central issue for the '778 Patent will be whether claims from a patent rooted in solving problems specific to Fibre Channel (FC) can be construed to cover accused services based on standard Layer 2 Ethernet technology, or if the detailed specification implicitly limits the claim scope to the FC context.
2. Functional Equivalence: A key evidentiary question for both the '381 and '778 patents will be one of operational correspondence. Does the alleged infringing functionality-which relies on general-purpose OS features and standard networking protocols-perform the same specific, and arguably more specialized, functions recited in the claims and described in the patent specifications?
3. Indirect Liability: For the '172 Patent, the case raises the question of where to draw the line for inducement liability. Can a provider of foundational data center and interconnection services be held liable for its customers' alleged infringement when those customers build and operate their own distinct applications on the provider's platform? The outcome will likely depend on the specificity of Equinix's alleged instructions and its intent for customers to use its services in the allegedly infringing manner.